Citation
City councils and public policy

Material Information

Title:
City councils and public policy
Creator:
Bradley, Robert Bernard, 1946-
Copyright Date:
1977
Language:
English
Physical Description:
xvii, 553 leaves : ; 28 cm.

Subjects

Subjects / Keywords:
Cities ( jstor )
City councils ( jstor )
City politics ( jstor )
Conservatism ( jstor )
Councils ( jstor )
High socioeconomic status ( jstor )
Political science ( jstor )
Public policy ( jstor )
Socioeconomic status ( jstor )
Socioeconomics ( jstor )
Community power -- Case studies ( lcsh )
Decision making -- Case studies ( lcsh )
Dissertations, Academic -- Political Science -- UF ( lcsh )
Municipal government -- Case studies ( lcsh )
Political Science thesis Ph. D ( lcsh )
Florida City ( local )
Genre:
Case studies ( lcsh )
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 539-552.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Robert Bernard Bradley.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
026306811 ( AlephBibNum )
04043728 ( OCLC )
AAX3592 ( NOTIS )

Downloads

This item has the following downloads:


Full Text











CITY COUNCILS AND PUBLIC POLICY


By
Robert Bernard Bradley























A Dissertation Presented to the Graduate Council of
The University of Florida
in Partial Fulfillment of the Requirements for the
Degree of Doctor of Philosophy







UNIVERSITY OF FLORIDA
1977














'4

"yr





















L
-^


















i














To the Memory of
My Parents


















--













-.S.












^-?














ACKNOWLEDGEMENTS

I began this project almost four years ago. During that time I have

benefited immensely from the insight and counsel of Dr. Alfred B. Clubok.

He is an extraordinary teacher and a valuable friend. He has been patient

and sensitive, critical and constructive. And in my association with him

I have learned more than I might easily say. For what he has given, for

what he is, I would like to extend my sincere appreciation.

I would also like to express my appreciation to Dr. Bert E. Swanson

who has patiently read through several versions of this dissertation and

whose comments improved the entire effort markedly. This dissertation also

benefited from the advice and commentary of Dr. Keith Legg, Dr. Eugene Witt-

kopf and Dr. Robert Ziller. Dr Legg, in particular, contributed more than

he might know since it was he along with Dr. James F. Morrison who first

convinced me that political science rather than physics might provide fas-

cinating avenues of intellectual development.

In a work such as this numerous other individuals provided comfort and

assistance along the way. In particular I would like to thank my good

friends Dr. Clem Bezold and Dr. Paul Peretz with whom I discussed aspects

of my argument and without whose counsel this work would be much the worse.

Throughout the entire time I was preoccupied with this exercise, I have

had the recourse to the advice of Dr. Carolyn Herrington. She has consis-

tently urged me to clarity though I be muddled, pushed me to precision when

I was vague, and counseled me to simplicity when I was roundabout. That I

have often failed to this good counsel is a blame I regretfully accept.














TABLE OF CONTENTS


ACKNOWLEDGEMENTS


LIST OF TABLES . . .


LIST OF FIGURES


ABSTRACT

CHAPTER


INTRODUCTION


The Problem and Its Importance . . . .
The Scope and Limits of the Research . . .
The Argument of the Dissertation . . .
Notes .

THE POLICY FOCUS, CITY COUNCILS AND EXCHANGE THEORY


Policy Models and the Policy Outcomes . .
Approaches to the Study of City Councils
An Exchange Perspective on Policy Making in City
Councils .
Summary .
Notes .


27
34

41
68
69


THE DEMOGRAPHIC APPROACH TO FLORIDA MUNICIPAL POLICY


The Units of Analysis . .
The Dependent Variable
The Independent Variables
The Intervening Variables
Specifying the Linkages
Methods and Analysis . .
A Commentary on the Findings
Notes .

CLUSTERING FLORIDA CITIES


Considering the Operational Environment .
The Establishment of Environment
The Reliability of the Factor Solution .
City Classification: Socioeconomic Environment
City Classification: Expenditure Types .
Summary . .
Notes .


S112
S116
S119
133
142
S156
157


S 1ii

S vii


xii


xv


TWO


THREE


FOUR


S . 80
. 83
. . 87
S . 90
S . 93
S . 97
. . 103
107

. . 112








FIVE FLORIDA COUNCIL MEMBERS AND THEIR COUNCILS . . 161

The Typical Florida Council Member . . . 162
Florida Council Members: Characteristics and Percep-
tions 167
The Characteristics of Florida City Councils .179
The City Councils of SES Group F . . . 195
Summary 206
Notes 209

SIX PERSONAL ATTRIBUTES, CONTEXTUAL FACTORS AND COUNCIL
RELATIONS 211

Members and the Form of Their Cognitive Structures .212
Filling in the Members' Cognitive Structure 223
Cognitive Structures and Council Relations . . 231
The Council Context 240
Council Relations in SES Group F . . . 249
Summary 261
Notes 264

SEVEN COUNCIL CONFLICT AND DECISION-MAKING MODES . . 266

Council Membership and the Prospect of Conflict 267
Exchange Theory and Council Conflict . . . 275
Council Conflicts 279
Decision-Making Modes . . . . . 288
The Correlates of Decision-Making Modes . . 296
Decision-Making Modes in Group F . . . 301
Decision-Making Modes and Conflict . . . 309
Summary 319
Notes 320
EIGHT DECISION-MAKING MODES AND POLICY . . . .323

The Importance of Issues 325
Conflict, Issues, and Decision Modes . . . 335
Issue Conflict Clusters and Decision-Making Modes .347
Summary 362
Notes 364
NINE POLICY ISSUE CONFLICT AND POLICY TYPES . . 366

Conflict Over Issues 366
Revealed Interest Versus Posited Interests . .378
An Examination of Council Policy . . . 390
An Examination of Policy Types . . . . 405
Summary 415
Notes 419


Page


CHAPTER








CHAPTER

TEN









ELEVEN


THE RESEARCH INSTRUMENT AND SCALE VARIABLES

Representativeness of the Sample . .
The Scale Variables . . . .
The Conservatism-Liberalism Scales
Group Influence Scales . . .
Policy Disagreement Scales . . .
Policy Agreement Scales . .....
Scales of Various Council Activities


Scales of Council Evaluation of Counc
Information Cue Scales . . .
Council Relations Scales
Scales of Problem Perceptions .
Notes .

ANALYSIS OF FINANCIAL DATA FOR FY 197

Notes .

DEMOGRAPHIC VARIABLES USED IN STUDY


POLITICS AND POLICY .

Policy Types and Expenditure Groups in the High SES
Cities .
Political Features of the High SES Cities and Their
Policies .
Summary .
Notes .

SALVAGING THE POLITICAL SYSTEM . . . .

The City Council in Policy Making . . .
The Implications for Policy Explanation
Policy Research and the Policy Problem.....
Notes . .


467

. . 476
478
. . 480
. . 484
489
490
492


il Performance





2 FOR CITIES


493
493
495
498
500

502


. . 508

. 509
513t


No tLes . . .

D VARIMAX ROTATION OF SEVENTY-THREE SOCIOECONOMIC
VARIABLES . .

Notes . . .

E CLUSTERING ISSUE DISAGREEMENTS . .

Notes .

F STATISTICS .

Box's M .
Wilks Lambda .
Notes .

LIST OF REFERENCES .

BIOGRAPHICAL SKETCH .


514

528

529

535

536

536
S537
538

539

553


Page

422


424

435
442
444

445

445
455
458
465


APPENDIX

A


M / *- -














LIST OF TABLES

Table Page

3-1 Alphabetical List of Cities in Study . . . 84

3-2 Variables Contained in the Fiscal Reporting and Uni-
form Accounting and Budgeting System for the State of
Florida 86

3-3 Classification of Variables Employed in Factorial
Ecology 89

3-4 Variables Relating to City Political Structure . 92

4-1 Factor Analysis of Expenditure Variables: Varimax
Rotated Factor Matrix 121

4-2 Factor Analysis of Percent of Total Expenditures by
Functional Area Variables: Varimax Rotated Factor
Matrix .. 123

4-3 Beta Weights of Regression of SES (Factor Scores) on
Expenditures and Percentage of Expenditures by Func-
tional Area (Factor Scores) with All SES Variables
Forced into the Equation . . . . . 129

4-4 Relative Depression of Multiple Correlation Coeffi-
cients 132

4-5 Socioeconomic Groupings Used Throughout Study: Group-
ings of Florida Cities Over 10,000 Population Based on
Discriminant Analysis of 73 SES Variables Reduced by
Varimax Factor Rotation into 12 Scores Per City (Pri-
marily 1970 Data) 135

4-6 Contiguity Analysis of Cities in the Socioeconomic
Groups 137

4-7 Summary Listing of Final Assignments by Discriminant
Analyses of Cities on Total and Relative Proportions
of Expenditures in Each Functional Category . 144

4-8 Grouping of Florida Cities Based on Discriminant
Analysis of Expenditure Variables . . . 147











4-9 Grouping of Florida Cities Based on Discriminant
Analysis of Percent of Total Expenditure Variables
by Functional Area 149

4-10 Discriminant Analysis of Percentage of Total Ex-
penditures by Functional Category Within Selected
SES Groups 152

5-1 Frequency Distributions of Member Response on
Scale Variables 169

5-2 Analysis of Variance On Scale Variables For Coun-
cil Means of SES Groups . . . . 182

5-3 Group Council on Council Relations Scales for
SES Groups: Normalized to Range 1-10 . . 194

6-1 Kendall's Tau-c Between Conservatism Scales and
Member Life Experiences . . . . 216

6-2 The Relationship Between Measures of Conserva-
tism and Occupation . . . . . 218

6-3 Kendall's Tau-c Between Measures of Personal
Goals and Life Experiences . . . 220

6-4 Kendall's Tau-c Between Measures of Conservatism
and Members' Goals 223

6-5 Council Relations Scales Used to Tap Each Element of
Eulau and Prewitt's Concept of "Triple Bonding" 232

6-6 Kendall's Tau-c Between Selected Measures and Coun-
cil Characteristics for All Individual Council
Members . . 233

6-7 Spearman's Rho Between Measures of Council Charac-
teristics 236

6-8 Kendall's Tau-c Between Selected Measures and Council
Characteristics for All Councils . . . 243

6-9 Kendall's Tau-c Between Selected Measures and Council
Characteristics For Councils of Group F . . 253

6-10 Kendall's Tau-c Between Measures of Council Character-
istics for High SES Cities of Group F . . 260

6-11 Kendall's Tau-c Between Measures of Predispositions and
Differences in Predispositions for Councils of High SES
Cities in Group F 260


viii


Page


Table








Table Page

7-1 Typology of Legislative Forms . . . . 270

7-2 Kendall's Tau-c Between Measure of Council Performance
and Council Characteristics (Relations) . . 278

7-3 Kendall's Tau-c Between Selected Measures of Council
Life and Total Council Conflict for All Councils . 282

7-4 Number of Councils Above and Below the Sample Mean on
Policy Disagreement (MODES) Scales . . . 291

7-5 Number of Councils Above Group F Mean on Only One of
the Policy Disagreement (MODES) Scales . . 291

7-6 Council Means on Decision-Making Mode Scales for SES
Groups: Normalized in Range from 1 to 10 . . 297

7-7 Kendall's Tau-c Between Selected Measures of Council
Life and Decision-Making Modes for All Councils . 299

7-8 Z-Scores of Council Means on Decision-Making Modes for
SES Group F Cities 303

7-9 Kendall's Tau-c Between Measures of Decision-Making
Modes and Conflict 310

7-10 Generalized Disagreement and Bargaining in SES Group F 312

7-11 Amicable Agreement and Generalized Disagreement in SES
Group F 313

7-12 Number of Decision-Making Modes Referenced Highly by
Each City Council in SES Group F . . . 314

8-1 Number of Council Members and Councils Citing Policy
Issues as Important. . 329

8-2 Policy or Funding Area Ranked by Proportion of Members
Indicating Council Disagreement . . . 339

8-3 A Comparison of Policy Issues According to Their Number
of Mentions and Their Propensity to Generate Council
Disagreement 341

8-4 Average Number of Issue Disagreements Per Member for
Each Council in SES Groups . . . . 344

8-5 Clusters of Council Issue Disagreements . . 348









8-6 Average Disagreement in Each Cluster Per Member
Per Issue for All Members of SES Group F .. 351

8-7 Average Proportion of Disagreement Per Issue Cluster
Per Member Per Issue for Group F Councils . . 352

8-8 Kendall's Tau-c Between Decision-Making Modes and Coun-
cil Divisions on Clusters for Group F Cities . . 354

8-9 Kendall's Tau-c Between Council Division Clusters and
Total Amount of Council Conflict on Issues for Coun-
cils of SES Group F 359

8-10 Kendall's Tau-c Between Member Problem Perception and
Council Division Clusters for Councils of SES Group F 361

9-1 Proportion of Council Members Indicating Disagreement,
Citizen Contacts and Group Actions by General Policy
Area . .. 371

9-2 Phi Associations Between Council Disagreement and
Types of Contacting by General Policy Area . . 375

9-3 Classification of Policy Issues into Four Types . 389

9-4 Z-Scores of Proportion of Council Members Indicating
Disagreement, Citizen Contacts and Group Actions by
General Policy Area . . . . . . 392

9-5 Difference Between Z-Scores for Proportions of Members
Indicating General Policy Area and Council Divided,
or Attracted Group and Citizen Activity . . 394

9-6 Z-Scores of Proportion of Council Members Indicating
Disagreement, Citizen Contact, and Group Action by
General Policy Area and Classified Policy Types 402

9-7 Proportion of Council Members Indicating Agreement With
Intent of Statement About Policies of Importance . 407

10-1 Average Proportion of Members Reporting Council Disagree-
ment, Citizen Contacts, and Group Actions by Policy Type
in Each Expenditure Group of High SES Cities (Group F) 426

10-2 SES Group F Council Members' Characterization of the
Political Aspects of Traditional Issues . . 430

10-3 Vector Analysis of Cross-Tabulation of Important Issues
Involving Council Disagreement and Group Conflict in
Cities of Group F 432


- ii LII


Table


Page









10-4 Average Proportion of Members Reporting Council
Disagreement, Citizen Contacts, and Group Actions
by Policy Type for All Council Members and Those in
SES Group F 437

10-5 Average Scores of Selected Indices in Each Expenditure
Group of High SES Cities (Group F) . . . 438

A-1 Interview Schedule 468

A-2 Response Rate by SES Group . . . . 475

A-3 Councils Represented in Each SES Group by Number of
Council Members Responding . . . . 476

A-4 Breakdown of All Members' Responses to Question:
What was your impression of this questionnaire? . 476

A-5 A Comparison of Sample and Population Characteristics
by Sex 477

A-6 Scales and Their Mnemonics . . . . 481

A-7 Reliability Coefficients for Scale Variables . 485

B-l General Information on Cities by Population Class:
1972 503

D-l Varimax Rotated Factor Matrix of Demographic
Variables 515

E-l Varimax Rotation of Issue Disagreement Areas Evaluated
by Members 534















xi



xi


I~.~


Page


Table









8-6 Average Disagreement in Each Cluster Per Member
Per Issue for All Members of SES Group F 351

8-7 Average Proportion of Disagreement Per Issue Cluster
Per Member Per Issue for Group F Councils . . 352

8-8 Kendall's Tau-c Between Decision-Making Modes and Coun-
cil Divisions on Clusters for Group F Cities . . 354

8-9 Kendall's Tau-c Between Council Division Clusters and
Total Amount of Council Conflict on Issues for Coun-
cils of SES Group F 359

8-10 Kendall's Tau-c Between Member Problem Perception and
Council Division Clusters for Councils of SES Group F 361

9-1 Proportion of Council Members Indicating Disagreement,
Citizen Contacts and Group Actions by General Policy
Area . .. . . 371

9-2 Phi Associations Between Council Disagreement and
Types of Contacting by General Policy Area . . 375

9-3 Classification of Policy Issues into Four Types . 389

9-4 Z-Scores of Proportion of Council Members Indicating
Disagreement, Citizen Contacts and Group Actions by
General Policy Area . . . . . . 392

9-5 Difference Between Z-Scores for Proportions of Members
Indicating General Policy Area and Council Divided,
or Attracted Group and Citizen Activity . . 394

9-6 Z-Scores of Proportion of Council Members Indicating
Disagreement, Citizen Contact, and Group Action by
General Policy Area and Classified Policy Types 402

9-7 Proportion of Council Members Indicating Agreement With
Intent of Statement About Policies of Importance . 407

10-1 Average Proportion of Members Reporting Council Disagree-
ment, Citizen Contacts, and Group Actions by Policy Type
in Each Expenditure Group of High SES Cities (Group F) 426

10-2 SES Group F Council Members' Characterization of the
Political Aspects of Traditional Issues . . 430

10-3 Vector Analysis of Cross-Tabulation of Important Issues
Involving Council Disagreement and Group Conflict in
Cities of Group F 432


- I r~,


Table


Page









10-4 Average Proportion of Members Reporting Council
Disagreement, Citizen Contacts, and Group Actions
by Policy Type for All Council Members and Those in
SES Group F 437

10-5 Average Scores of Selected Indices in Each Expenditure
Group of High SES Cities (Group F) . . . 438

A-1 Interview Schedule 468

A-2 Response Rate by SES Group . . . . 475

A-3 Councils Represented in Each SES Group by Number of
Council Members Responding . . . . 476

A-4 Breakdown of All Members' Responses to Question:
What was your impression of this questionnaire? . 476

A-5 A Comparison of Sample and Population Characteristics
by Sex 477

A-6 Scales and Their Mnemonics . . . . 481

A-7 Reliability Coefficients for Scale Variables . 485

B-1 General Information on Cities by Population Class:
1972 503

D-l Varimax Rotated Factor Matrix of Demographic
Variables 515

E-l Varimax Rotation of Issue Disagreement Areas Evaluated
by Members 534


Page


Table














LIST OF FIGURES

Figure Page

1-1 The Systems Approach to Policy Outcomes . . 14

1-2 Overview of the Major Conceptual Categories and Relation-
ships Investigated in the Micro-phase . . . 16

1-3 Overview of the Major Conceptual Categories, Analytical
Variables, and Relationships Investigated in the Micro-
phase of the Study 17

2-1 A Typology of Decision Organization . . . . 55

2-2 Decision Organization and Preference Articulation . 57

2-3 The Conditions Under Which Different Policy Types Can Be
Expected To Emerge From City Councils . . . 61

2-4 Modified Version of Salisbury and Heinz Typology . 65

3-1 The Demographic Perspective: Macro-phase Indicators . 80

4-1 Plot of First and Second Canonical Variables For Each I
City in Group: Based Upon Discriminant Function of SES
Variables 139

5-1 Mean Scores For All Council Members on Evaluation of Com-
munity Influence of Groups and Organizations . . 165

5-2 Mean Scores For All Council Members on Evaluation of the
Seriousness of Community Problems . . . . 166

5-3 Normalized Means on Conservation Scales For Councils of
SES Groups 186

5-4 Normalized Means on Group Influence Scales for Councils
of SES Groups . . 188

5-5 Normalized Means on Problem Perception Scales for Coun-
cils of SES Groups 191

5-6 Normalized Council Means on Conservatism Scales for .
Councils of SES Group F . . . . 197 ;

5-7 Normalized Council Means on Group Influence Scales for
SES Group F . . . 199

xii ':







Figure Page

5-8 Normalized Council Means on Problem Perception Scales
for SES Group F 202

5-9 Normalized Council Means on Council RElations Scales
for Councils of SES Group F . . . . . 204

6-1 Major Categories and Variables Treated in Chapter Six 213

6-2 Closed Networks of Bivariate Relationships Centering on
Measures of Problem Perceptions and Political Predispo-
sitions for All Members . . . . . 225

6-3 Closed Networks of Bivariate Relstionships Centering on
Measures of Council Relations for All Members . . 238

6-4 Closed Networks of Bivariate Relationships Centering on
Measures of Council Relations for All Councils (N = 61) 246

6-5 Closed Networks of Bivariate Relationships Centering on
Measures of Council Relations for the Councils of High
SES Cities (N = 12) 257

7-1 Major Categories and Variables Treated in Chapter Seven 268

7-2 Conflict in the SES Groups . . . . . 280

7-3 Closed Network of Bivariate Relationships Centering on
Measures of Council Relations and Total Council Conflict. 284

7-4 z-Scores of Council Means in SES Group F on CONFLICT . 287

7-5 Closed Networks of Bivariate Relationships Centering on
Decision-Making Modes for Councils of the High SES Cities
in Group F 305

7-6 Correlations of Decision-Making Modes and Conflict . 318

8-1 Major Categories and Variables Treated in Chapter Eight 324

8-2 Correlation of Decision-Making Modes With Average Disagree-
ment Over Issues in Group F (Kendall's Tau-C) .. . 346

9-1 Major Categories and Variables Treated in Chapter Nine 367

10-1 Systems Model of Policy Process: Micro-Phase Indicators. 423

E-l Plot of Stress Against Dimensionality in Multidimensional
Scaling Analysis of Issue Conflicts (MDSCAL Program) . 531

E-2 Plot of the Minimum Average Correlation Necessary for In-
clusion in Cluster Against the Number of Clusters Produced
at That Level of Correlation (CLUSTER Program) . . 531

xiii






Figure Page

E-3 Two-Dimensional Solution Scaling Issue Disagreements
(Stress = .270) .532


xiv


1~Ppr IE~P~ I~ IC- 1~- I' I q ~-~---~








Abstract of Dissertation Presented to the Graduate Council
in Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy
at the University of Florida


CITY COUNCILS AND PUBLIC POLICY

by
Robert Bernard Bradley

December 1977

Chairmen: Alfred B. Clubok
Major Department: Political Science

City councils are breakwaters for public policy. There is a strong

link between the political features of councils and their communities

and the types of policies which they pursue. So argues this disserta-

tion as it examines city councils and their handling of public policy

in cities throughout Florida.

In the last decade, research undertaken within the premises of the

demographic model of public policy has shown socioeconomic factors to

systematically account for municipal expenditures better than the struc-

tural features of municipal government. However, when 78 Florida cities

with population greater than 10,000 were grouped into six clusters, each

sharing relatively similar socioeconomic conditions, expenditure patterns

within several clusters were found to be different. Cities displayed

different expenditure profiles despite the similarity of their environ-

ments. This suggested the possibility that political forces might act

in a crucial way to determine public policy. City councils, it was

thought, might be crucibles where these political forces are given play.

To gauge the differences in the political situation of the many

councils throughout Florida, this study relied on responses of council

xv







members in 61 cities to a mail questionnaire and personal interviews.

These responses were investigated for the manner in which the personal

attributes of the members, their perceptions of significant contextual

factors and the characteristics of the councils to which they belong act

to alter each council's decision organization. City councils, it was

found, develop a repertoire of decision-making modes by which they handle

different policy choice situations. Seven different modes, falling into

two broad types, were investigated: 1) disagreement modes which include

generalized, particularized and idiosyncratic splits, and 2) agreement

modes which include bargaining, amicable agreement, consensual and infor-

mal agreement. Typically, the personal attributes of the members and es-

pecially their political and religious predispositions along with council

relations will help shape the mix of modes a council adopts. But the

specific modes a council adopts will evolve out of the kinds of issues

considered and the historical circumstance of the council.

Decision-making modes capture conceptually the orientation members

evidence toward matters which come before the council. The contextual

component of such orientations is great. Not every matter in a general

policy area will be resolved with the same decision mode. Nor are parti-

cular modes likely to be used for any given issue across a range of coun-

cils. Nonetheless, some decision modes are more permeable to certain is-

sues than others. And in any event, the referencing of modes appears to

be a highly complex and discriminating process.

In part, this complexity results from the impact of issues on each

council and their meaning within that council. A simple classification

scheme, such as that suggested by Lowi, appears inadequate to capture the

manner in which politics and policy relate across the range of issues

xvi


-- -^--- -- ---.-~- I -4LC- 1~----







which councils consider. But as the salience of the policies increases,

as they become more important in the perceptions of the council members,

policies can be expected to result in a predictable fashion from the pat-

tern of council decision making and demands placed on the council by the

community. In fact, the systematic policy differences that are observed

between cities appear related to systematic differences in the political

characteristics of councils and the communities which they serve.







Chairman


xvii














CHAPTER ONE
INTRODUCTION

In principle, municipal government stands at the very core of the

American political system. So conceived as to serve the smallest consti-

tuency in the nationwide areal division of powers, its place in the fede-

ral system has traditionally been viewed in purely functional terms.1

While Madison and others might have intended such an areal division as

part of the cure for the mischiefs of faction, legal doctrine and the vast

body of American political theory came to regard cities as entities with

limited jurisdiction, and what is worse, limited capacity to solve their

own problems. As such, the ability of municipal governments to exercise

powers in service of their constituencies was severely circumscribed. As

municipal corporations, cities often were not extended the inherent right

of municipal self-governance by the courts. Municipalities existed at the

pleasure of the states and frequently languished in their neglect. Gra-

dually, over the last 20 years or so, this position has given way to what
2
James Sundquist calls a "policy of deference."2 Increasingly, the con-

scious policy of the federal government, and state governments as well,

has been to defer to local judgement. Local political institutions have

acquired a new importance, if not always a new potency.

The Problem and Its Importance

Among the structural pillars of local government stands the city

council or commission. As the basic policy-making apparatus of the city,

the council is nominally charged with matters of taxation, appropriation,

tht


-___________- -- ,






and the supervision of city administration. In fact, however, there is

little concerning the social, political, and physical life of a city into

which its council does not or cannot inject itself. Everything from spon-

soring the local beauty queen to approving the course of genetic research

comes within its purview. For many of the most important issues facing

any municipality, councils serve as public crucibles. In theory at least,

they are used to amalgamate the demands of the citizenry, while charging

the community with commitments.

This description belies the place councils have found in the discus-

sion of urban problems and politics. For most scholars, the machinations

of local councils have been of little importance. The real politics of

the community and the real determinants of local public policy, it has

been argued, are seldom found in institutional bodies such as councils or

in the results of their formal sessions. Persuasively, and at length,

scholars and experienced observers have warned against the dangers of too

formalistic a view of local politics. They look instead to the manner in

which elite preferences are reflected in community choices, to the manner

in which community groups and organizations help shape local policy, and

to the manner in which the socioeconomic context of the city conditions

its policy alternatives.

However, the literature on urban processes has not always reflected

this view. During the first part of this century the importance of formal

structures was presupposed. Since the early 1950's a different perspective

on local government has held sway. During the early years of the century,

the hopes of city government reformers turned on the notion that institu-

tional change would result in better government; the problems with govern-

ment could be resolved in tinkering with the machinery. After the Second

World War, this optimistic view gave way to a period of reflection. Re-


I ~_~





forms appeared to have little effect. To determine why this was the case,

researchers re-examined their assumptions, and began anew to describe the

processes of city governance. For the most part, the workings of govern-

ment dominated the discussions. The concern for "good" government was left

to administrators and amateur philosophers.3

Still, the tradition of reform has not been entirely abandoned. In

the aftermath of the events of the late sixties, a new concern for the a-

bility of local governments to deal effectively with their problems has

arisen. Whether seen in the harsh pronouncements of the Unheavenly City

or in the criticism made of Banfield's iconoclasm, a "new reform' litera-

ture is upon us.4 And the thrust of the new reform movement concerns the

capacity of local government to satisfy the claims of its citizenry, es-

pecially minority and disadvantaged groups. This emphasis carries a re-

newed concern for the institutions of local government.

This interest is tempered by the research and concerns of the last

decade, however. There is the explicit recognition, amplified in the last

twenty years of research, that city hall does not itself have the capacity

to govern within the game as it is now played, nor is it an institution

which can be treated independently from the community of which it is a

part. In fact, policy initiatives may be more a function of the objective

resource capabilities of the community than the desires of local decision

makers. Politics itself, it has been argued, may have little impact on lo-

cal policy.5 But even where politics does have a role, public policy, it

is realized, is made in many sites throughout the community, oftentimes

outside of government. Bankers and brokers, planners and bureaucrats, pub-

lic and private--all interact to allocate the values in a municipality.

Municipal public policy is a complex result of many forces.

There are few studies, though, which can comprehend such complexity.







Indeed, this study will not attempt anything so ambitious. The perspec-

tive and research advanced here are aimed at capturing that part of pub-

lic policy which percolates through a city's council or commission. Clear-

ly not all of local policy will be covered. Instead, I have elected to

consider only that portion handled by local councils. Some scholars might

suggest this emphasis represents the triumph of form over substance, of

structure over behavior. This view, I think, is mistaken. Holistic con-

ceptions of political systems can illuminate stubborn problems, but are

difficult to document and research. In deciding to examine city councils

and their role in municipal policy formulation, I have chosen to view

councils as breakwaters for the policy disputes of a community; that is,

as points at which the political ramifications of choice situations may

become evident and highly visible. This constitutes not so much a rejec-

tion of theses on the significance of community forces in making policy,

but rather an appreciation of the ability of governmental units to selec-

tively take sustenance from their surroundings. Collective bodies have

the ability to defy their surroundings or to ratify the situation as it

obtains. For many segments of society, it is only through or with refer-

ence to collective decision-making bodies that recourse from systemic pro-

perties can be had. City councils as possible initiators of change or as

umbrellas of the status quo play an important role in policy formation

for any municipality. If politics can be important for the policies-

adoptedby a community, councils should be involved in the politics and

part of the policy determination. Research on the role of councils in the

formulation of municipal policy should illuminate the fashion in which

politics can impact on policy.

Part of the reason why city councils have attracted so few researchers

probably lies in their failure to appear dramatic. In fact, observing a







council is often an exercise in tedium; the flow of action is difficult

to capture, while the minutia of city governance appears all too evident.

"The casual or occasional visitor to the city council meetings is often

distressed at what he sees; major pieces of legislation passed by unani-

mous vote with little or no discussion, while an hour is spent over the

implications of an apparently minor request for zoning reclassification."6

As with most collective deliberative bodies, debate tends to drag on, and

decisive actions are infrequent. Decisions, when they occur, tend to ter-

minate with unanimous or near unanimous voting.7

In short, councils may appear to bear little resemblance to demo-

cratic bodies "designed to institutionalize conflicts and facilitate the

clarification, crystallization, and resolution of political differences."8

From this perspective, their policy-making role becomes problematical.

In fact, we might well question whether environmental factors or demogra-

phic characteristics do not really determine much of local public policy

as a good portion of the literature dealing with the subject appears to

suggest.9 Perhaps the political decisions made in council chambers have

little effect on local policy when compared to the impact of the charac-

teristics of the city itself. But if councils are breakwaters for the con-

sideration of local policy, then council politics should make a difference,

and in a way which depends on the issues which a council considers and the

politics that is played out among the members.

I have taken such expectations and distilled them into three broad

concerns which I will address in this study. First, at the aggregate le-

vel and for the councils under consideration, this study addresses prob-

lems of municipal expenditures and their relationship to the socioecono-

mic and political, structural characteristics of cities. My intent is to



P5!






examine the relationship between these factors in Florida cities and to

isolate major indicators of municipal dynamics which impact upon Florida

council members. This analysis is designed with a view to controlling,

in the subsequent analyses, those aspects of the council members' socio-

economic environment which appear to have the most importance for the

council's allocation of funds. The relevant caveat to be applied in such

analysis is, although one would expect environmental conditions to deter-

mine largely at what level expenditures will be set, politics is likely

to be pivotal in establishing the distribution of rewards and benefits at

that level. The pivotal role of politics is likely to be discovered in

similar socioeconomic situations. Council politics should create dif-

ferent expenditure priorities despite similar socioeconomic contexts.

The second major concern focuses on the relationship between indi-

vidual council members, their characteristics, the nature of the decision-

al structures in which they operate, and the environmental factors that

impinge upon them. The inquiry will center on the manner in which the po-

litical and social milieu of members help determine the prevalence of con-

flict or unanimity on the council, and the manner in which council rela-

tions and member characteristics shape the council's orientation toward

choice situations and the consideration of different policy.issues.

Finally, the investigation deals with the manner in which councils

and members are affected by the types of policy issues which they consid-

er. Do certain types of issues act as triggers which initiate conflict

or promote accord? How do the issues which councils handle affect the po-

litical methods by which councils settle such issues?

Stated briefly then, this study will examine various characteristics

of individual council members and salient aspects of their sociopolitical

environment in order to determine how the structure of decision within


L HI







the council is organized and to explain the presence of conflict or con-

sensus among council members. Additionally, different types of policy

issues will be investigated in order to determine if they are associated

with certain patterns of conflict or consensus among the members or with

other aspects of a council member's circumstance. Finally, various as-

pects of local councils will be treated to determine if the political

characteristics of a council make a difference for the types of policy

a council pursues. Do councils establish similar profiles of political

activity in similar socioeconomic settings? Or are there political dif-

ferences among councils with similar contexts which one might reasonably

associate with the peculiar policies of the councils?

The Scope and Limits of the Research

In an attempt to answer such questions, a questionnaire was sent in

the fall and winter of 1974-1975 to 441 council members throughout the

state of Florida asking their opinions on a series of items relating to

council activities. This included their perceptions and attitudes on va-

rious aspects of municipal politics and social issues (see Appendix A).

Every council member elected to office in a city with 10,000 or more re-

sidents was sent a form. These cities represent almost half of Florida's

total population (49.9%), over 60 percent of the total urban population

of the state, and almost 90 percent of all residents living in incorpo-

rated areas.10 Two waves of questionnaires were mailed out; the first in

November, the second in December. Of those asked to participate, 173 re-

sponded or 39.6 percent of the 437 members occupying council seats in this

time period.11 This rate, while itself low, compares favorably with the

return rates reported in other questionnaire-based, mail surveys of coun-

cil members.12 In addition to the mail-out questionnaire, however, face-







to-face interviews were conducted with council members in twelve cities

selected for their socioeconomic similarities. Among those cities, 81.6

percent of the council members were interviewed in sessions that lasted

from 45 minutes to an hour and one-half. In all, 61 cities were repre-

sented in the sample. In 51 of the cities, two or more members responded,

and in 25 cities, a majority of the council members are represented in the

results.

These results naturally limit the general applicability of the find-

ings, especially for those parts of the study which treat the entire sam-

ple of members. On the other hand, within the selected group of cities

where nearly all members responded, the results appear to bear consider-

able validity. The high response rate within this one group allows for

some check on the validity of the entire sample. While such inferences

must be treated with caution, a check of certain design variables in both

portions of the sample reveals the entire group of council members does

not differ significantly from the members who have been sampled more in-

tensively (see Appendix A). Of course, it is difficult to impute total

similarity on the basis of demographic patterns alone. Nonetheless, the

responses do represent a considerable amount of information on the city

politics and council life of communities throughout the state of Florida.

Statistical problems abound, but the end result promises to be worth the

effort.

This study is limited by more than its data base. Conceptions of

urban politics that draw their focus narrowly are likely to run afoul

of the open and dynamic nature of municipalities. My concern for coun-

cils issues from the expectation that they stand as the center of a swirl-

ing dynamic interwoven with a city's policy making. My own theoretical





and practical concerns would have led me to encompass additional elements

of the whole, but practical considerations required that I limit the prob-

lem. I am keenly aware of the caution issued by pioneers in the study

of legislative behavior:

The notion of the legislature as a decisional system tends
to limit the focus of inquiry more than seems appropriate
for the purpose of specifying just how legislature and po-
litical system are interdependent. . The decisional ap-
proach ignores the fact that a good deal of legislative be-
havior does not involve decisionmaking at all.13

Yet I believe the approach I have taken to city councils avoids many of

the pitfalls that characterized the decision-making approach per se.

Another limitation of this study should be noted; it does not come

to grips with one of the most significant aspects of institutional policy

making, namely, the issue of non-decision. It has long been realized

"that the most important part of the legislative decision process was

the decision about which decisions to consider."14 Every community has

a systematic agenda consisting of all "issues that are commonly perceived

by members of the political community as meriting public attention and as

involving matters within the legitimate jurisdiction of existing govern-

mental authority."15 Yet the systematic agenda will never be considered

as a whole for inclusion on the institutional agenda. Only certain ele-

ments merit such consideration as a result of the patterns of support and

opposition within the community. But the institutional agenda is impor-

tant not only for those individuals and groups able to control its con-

tent, but also for those segments of the community whose participation

it is able to influence. This is a vital issue in the consideration of

all public policies, however; it has been excluded from explicit consi-

deration in this study for practical reasons. Non-decisions require

that an elaborate set of assumptions be posited about a given situation,


-IC-~r~err I -r Ir







and that all stages of the decision-making process be monitored. This,

in turn, requires an enormous research effort far beyond the scope of this

project.

There is one further, major limitation to this study; that is, the

context in which it is cast. Surveys elicit responses from which a pic-

ture of a social situation can be painted. But the questions used to gen-

erate the responses, and the responses themselves stand as stony datum.

There is little way of assessing the remarks. Only tangentially, through

a member's own colleagues, can the situation be evaluated. Of course, in

those twelve instances where indepth field work was conducted, the overall

perspective can be augmented.- Even here, however, no interviewing of a

systematic nature was done in the community. Thus specific interpreta-

tions may neglect important keys to a more substantial understanding. My

only recourse is to fall back upon the questionnaire itself which was de-

signed so that important questions were re-enforced with internal checks

upon member consistency. Beyond this I must assume that the perceptions

of the members, their responses, are indicative of the situation as it

does obtain at one point in time.

The Argument of the Dissertation

In the next few pages, I will preview the shape of the argument de-

veloped in this study. For each chapter, I will present a brief sketch

of the approach used and the major findings.

The study of public policy has increasingly come to turn on the ques-

tion asked years ago by Harold Lasswell, "Who Gets What, When, and How?"16

In particular, recent research on public policy has addressed itself to

the "How" of public policy making and its link to "What" is decided upon.

Focusing upon the distinctions made in the systems approach, numerous re-

searchers have examined the relative importance of environmental variables






as opposed to political variables for policy when measured at the aggre-

age level.17 Others have suggested the solution lies in approaches at-

tentive to group dynamics, mass-elite interactions or community power.18

This study argues that the concern should be directed toward the concep-

tual issue of choice and its social manifestations. In fact, it argues

the conceptual underpinnings of a policy focus involves a subtle shift

in our image of political man; he is someone with a choice to make.

Municipal policy is formed by many different actors and in several

arenas. In this study, I will consider only the fashion in which the lo-

cal council helps shape policy through the choices it makes. Unlike many

studies of policy-making bodies and city councils in particular, I will

not rely heavily upon the conceptual apparatus associated with interac-

tion theory. In Chapter two, I argue that council actions can be consi-

dered in terms of exchange theoretical propositions. Borrowing from the

legacy of classical economics and primitive utilitarianism, I operate

from a series of assumptions:

1. While men do not seek to maximize profits, they always
seek to make some profit in their social transactions
with others.
2. While men are not perfectly rational, they engage in cal-
culations of costs and benefits in social transactions.
3. While some men do not have perfect information of all
available alternatives, they usually are aware of at least
some alternatives which form the basis for their assess-
ments of costs and benefits.
4. While there are always constraints on human activity, men
compete with each other in seeking to make profit in their
transactions.
5. While economic transactions are typified by material goals
in the economic marketplace, men also exchange other, non-
material commodities, such as sentiments and services of
various kinds.19

In addition, my conception of social organization also includes the no-

tion of strategic interaction. That is, the behavior of an invididual

in a group is conditioned and predicated upon expectations regarding the

behaviors of others in the group.


i~r


i

























z







Of course, the major difficulty posed by this framework is the con-

cept of "costs" and its measurement. While there is no general agreement

over the manner in which costs can be operationalized in empirical re-

search, the various branches of exchange theory appear to agree costs re-

fer ultimately to the internal calculations of value performed by the in-

dividual actor. Costs are the results of engaging in exchange processes

where an individual's preferences or goals are thwarted in some fashion.

Costs can be dealt with in terms of goals or preferences, but it is the

threat to goal fulfillment which is important. Thus, I have assumed costs

can be measured in the clash of members over issues, over personalities,

over styles. Costs are incurred in the manifestation of conflict.

The point of adopting the exchange framework is to present a means

of drawing the relationship between politics and public policy as deter-

mined in local councils. Specifically, I argue different types of policy

can be expected to emerge from the interaction of two kinds of costs in-

cumbent to council members: 1) those associated with the pattern of de-

mands impinging upon the members from sources external to the council,

and 2) those incurred by policy makers in the formation of decision coa-

litions among the members.20 For example, the more conflictual the po-

licy area, the more likely the policies involved are to be redistributive.

But this framework can also be used to investigate the ways in which coun-

cil members reach decisions. In particular, this approach can be used to

explain the considerable strains toward unanimity that exist in local

councils.

The exchange framework provides a means of animating the policy pro-

cess that is typically neglected in the systems view. This neglect is


C _~____






illuminated in Chapter three. At the conceptual level, the systems view

links environmental factors and the political system with policy outcomes

as shown in Figure 1-la. Researchers such as Dye and Hawkins, arguing

within this framework, have established the connection between socio-

economic factors and expenditures using the aggregate characteristics of

cities portrayed in census statistics.21 Yet their specification of the

systems model has been quite ambiguous. Without an explicit considera-

tion of the role played by policy-making institutions such as city coun-

cils, any number of hypotheses can link socioeconomic factors and various

expenditure measures. Nonetheless, their findings stand. Socioeconomic

factors appear to condition the pursuit of certain public policies. The

objective social and economic realities of a community can act to impose

constraints on a council member's choice. These factors can confound

the investigation of political behaviors and their link to policy formu-

lation. Thus I have conducted my research with an eye toward capturing

the link between politics and policy by engaging in a dual strategy. In

the macro-phase as shown in Figure l-lb, I examine the link between ag-

gregate measures of socioeconomic factors, certain structural aspects of

city government and a range of expenditure measures for the 78 cities

in Florida having over 10,000 population. In Chapter three, I pursue

this analysis to test whether Florida municipalities differ in these link-

ages from cities around the country and to display in explicit fashion

the strengths and weaknesses of this approach for our understanding of

public policy.

The lessons of the macro-phase were incorporated into the research

design investigating the manner in which the political features of city

councils and their contexts affect municipal policy and expenditures.


~I Crl ~


-S7-J


ISI1 L- IC_lldl P.-
















a. Conceptual Level


SES FACTORS


STRUCTURAL ASPECTS OF
CITY GOVERNMENT

--I-


EXPENDITURES


b. INDICATORS: Macro-phase


****[CONTROLLED]*****



POLITICAL FEATURES OF CLASSIFIED POLICY AND
COUNCIL AND COMMUNITY EXPENDITURE TYPES


c. INDICATORS: Micro-phase


Figure 1-1. The Systems Approach to Policy Outcomes


r''


EXPENDITURES


|






Chapter four initiates the micro-phase of the research. As shown in Fi-

gure 1-lc, socioeconomic factors were controlled in order to highlight

the relationship between political features and policy. This was done

by grouping the 78 Florida cities into clusters with similar socio-

economic characteristics. Six different group were generated. And

while council members were surveyed in each of the SES clusters, ex-

tensive interviewing was conducted in the cities ranked highest accor-

ding to the socioeconomic status of their citizens. These cities, it

will be shown, exhibit different expenditure patterns, despite their

relatively similar socioeconomic milieu. Extensive interviewing, using

a questionnaire incorporating exchange theoretical notions, helps estab-

lish why this might be.

Figure 1-2 represents an overview of the major conceptual categories

and relationships investigated in Chapters five through nine. These fea-

tures are used to develop the micro-phase inquiry pursued in this study.

In particular, it shows the manner in which policy choices are conditioned

by on-going political machinations in the council and the community and

act as mediators of such political events. Figure 1-3 provides a glimpse

of the analytical variables subsumed within each of the major conceptual

categories together with the relationships to be examined. These vari-

ables and the labelled relationships will be referenced in the discussion

which follows.

Chapter five begins a detailed investigation of council members and

councils throughout Florida aimed at forming an empirical foundation for

the exchange assumptions made in Chapter two. This chapter provides a

description of the characteristics and opinions of the individual coun-

cil members and their councils. It presents their responses on the many

I''














































d.




POLICY ISSUES





e. 1




POLITICAL FEATURES
OF
COUNCIL AND COMMUNITY




Figure 1-2. Overview of the Major Conceptual Categories and
Relationships Investigated in the Micro-phase








[PERSONAL ATTRIBUTES]


[COUNCIL CHARACTERISTICS]

DECISION ORGANIZATION
DECISION ORGANIZATION


Figure 1-3. Overview of the Major Conceptual Categories, Analytical
Variables, and Relationships Investigated in the Micro-
phase of the Study


[CONTEXTUAL FACTORS]











[DECISION


c ORGANIZATION]
ORGANIZATION]


POLICY ISSUE
CONFLICT


e.


N BY GROUPS
AL CITIZENS
e.
ISSUES:
COSTS


e.


e.


S e.


Figure 1-3. Continued


AGREEMENT MODES


A, .


DISAGREEMENT MODES


PARTICIPATION
AND INDIVIDUAL
ON POLICY
EXTERNAL


POLITICAL BEHAVIOR OF
COUNCIL MEMBERS ON POLICY
ISSUES: INTERNAL COSTS


POLICY TYPES:

1. Traditional Services
2. Amenities
3. Promoting Economic Growth
4. Arbitrating Between
Conflicting Interests

EXPENDITURE TYPES:

1. Service Expenditures
2. Service/Newer Public
Works Expenditures
3. Newer Public Works
Expenditures







scales and measures tapping members' personal attributes, perceptions of

contextual factors and council characteristics. It also considers the

manner in which councils with similar socioeconomic environments differ

according to such features.

In the wake of this analysis, council members emerge as conserva-

tives, politically and religiously, whose perceptions of local problems

are dominated by growth-related issues. As with their colleagues across

the United States, Florida council members are predominantly white, male,

businessmen in their forties who tend to be better educated than the rest

of the population. However, their predispositions, their perceptions of

community problems, their reliance on information sources (cues), their

evaluation of influential groups, and the types of relations which obtain

among members on the same council do not appear related, in any systema-

tic fashion, to the socioeconomic milieu of the councils themselves. The

ability of environmental conditions to constrain the attitudes or beha-

viors of city councils appears limited when examined in this fashion.

The univariate analysis presented in Chapter five underwrites the

data search that marks the next stage of the analysis. Among the high

SES cities of Florida, for example, the Boca Raton council evidences

consistent strains of polarization together with rather liberal poli-

tical predispositions. Contrariwise, Cocoa Beach is both conservative

and relatively cohesive. The coincidence of several such measures in

the univariate analysis prompted an examination of the cognitive struc-

tures of the council members and the manner in which these structures

predicate council characteristics.

In Chapter six, links a and b of Figure 1-3 are examined through

a systematic investigation of relationships between the many analytical




te: Y







variables. For example, liberal members are shown to be sensitive to

human resource problems in their community, to perceive people-oriented

groups as important in community decision making, and to be receptive

to information from citizens. In this case, personal attributes are

seen to be associated with certain contextual factors. Further, the

associational measures used in Chapter six help establish the manner in

which certain council member attributes and predispositions favor the

evolution of forms of interaction among members. Predispositions ap-

pear to condition the manner in which a council's members take sustenance

from groups within the community. Conservatives, for example, are much

more likely to attend to the influence of business groups than are po-

litical liberals. Religious conservatives, on the other hand, evidence

a completely different set of tendencies. The thrust of the argument

initiated in Chapter six is that given sets of empirically defined pre-

dispositions and established patterns of council relations may condition

the appearance of conflict among the council members. And if the ex-

change theoretical notions developed in Chapter two are correct, they

should also condition the way in which costs are apportioned to various

types of policies.

In fact, the predispositions of council members and the relations

obtaining among council members do appear to prejudice the manifestation

of conflict in the council. This association, together with the other

relationships specified in link c of Figure 1-3, is investigated in Ghap-

ter seven. Not surprisingly, conflictual relations among the members

such as polarization, stratification and the perception of community

pressure are associated with high levels of council conflict. However,

the most decisive political predisposition biasing conflictual situations






appears to be the degree to which the members evidence religious conser-

vatism. The greater the religious conservatism of the council, the less

conflictual relations will be and the less conflict will surface over

matters of policy. Among high SES cities, issue conflict is most likely

where the differences among the members on religious conservatism are

large.

The amount of conflict which surfaces in a council over matters of

policy or between the council and the city manager may affect the manner

in which the council chooses to decide on matters of policy. No council

faces every issue which comes before it in exactly the same fashion. In

Chapter seven, I argue councils are likely to develop and use a reper-

toire of modes to resolve policy choice situations.

If a council has established a mix of disagreement and agreement

modes by which it organizes itself to resolve most of the issues before

it, then issues which generate considerable costs should evoke disagree-

ment modes. This is the argument pursued in Chapter eight; it examines

the link between decision organization and policy issue conflict shown

in Figure 1-3, relationship d. And it does appear to capture a bit of

the empirical situation. However, councils are frequently able to han-

dle controversial issues within the context of agreement modes which rely

upon bargaining schemes or decision tacts which call upon all members to

work toward amicable agreement. On the other hand, controversial poli-

cies appear to fracture decision schemes based on interpersonal relations

as the foundation for consent.

The analysis presented in Chapter eight demonstrates that ongoing

political procedures developed to handle issue choices can survive even

the most controversial situations. But it also suggests that certain


I lL I- LL~I ILL ~L I







political arrangements will be unable to withstand the costs of certain

issue decisions. For example, certain policies regardless of the con-

flict they generate, do not engage generalized patterns of member disa-

greement. Members appear to act as though consensus is necessary in cer-

tain policy areas. Similarly, bargaining is typically excluded from is-

sues which generate large amounts of conflict and which attract the at-

tention of significant community groups. Exchange theoretical concep-

tions suggest members will have highly differentiated views of a given

policy; certain policies may have components which act to exclude their

consideration by a given decision mode.

Policies not only evoke decision modes differentially, they also

generate different amounts of decision-making costs and act to galvanize

community groups or citizen actions in different ways. The exposition

developed in Chapter nine examines these relationships, labelled e in

Figure 1-3. It demonstrates that amenities such as public welfare and

community development projects constitute a set of policies quite likely

to provoke council disagreement and community actions. Traditional mu-

nicipal policies such as the provision of police and fire protection

are unlikely to attract such attention.

The analysis developed in chapter nine makes it clear that not

every council is capable of sustaining the political costs necessary

for the pursuit of amenities. Cities which share relatively similar

socioeconomic characteristics may pursue quite different policy priori-

ties. This is exactly the relationship depicted in Figure l-lc and exa-

mined in Chapter ten. The findings buttress the conclusions of Chapter

four. Not only do cities with similar environments spend their monies

differently, they also vary systematically in the substantive policies


---






they pursue and in the most important issues which surface in their com-

munities. The politics of the cities differ and have different effects.

Not surprisingly, liberal, people-oriented councils where the religious

differences among the members are large are more likely to pursue po-

licies which provide for amenities than conservative, business-oriented

councils.

By focusing upon the empirically defined preferences and interests

of council members, it is possible to illuminate the fashion in which

councils help determine the policies of local government. The political

features of the local council are related to the policy priorities of

the community. These political features remain important even when the

socioeconomic context of the council remains relatively constant. But

just as the policy choices of a community are conditioned by the ongoing

political features of the council, so, too, these features may be medi-

ated by the policies considered by the council. As certain policy choices

make their way to the council chamber, their consideration can alter the

the way in which the council does its business. In the end, the formula-

tion and consideration of policy by local policy-making bodies remains

an immensely complex process.












Notes

1 Arthur Maas, "Division of Powers: An Areal Analysis," A Theory of
.1 Government, ed. Arthur Maas (Glencoe, Ill.: Free Press, 1959), pp.


2 James Sundquisc, Making Federalism Work: A Study of Program Coordina-
I at cthe Communicy Level (Washington, D.C.: Brookings Institution, 1969),
'4 7.

3 rlattheu Holden, "The Politics of Urbanization," People and Politics
Jrban Socievt, Vol. VI: Urban Affairs Annual Review, ed. Harlan Hahn
erly iHills: Sage, 1972), pp. 557-602.

Ibid.

5 See Thomas Dye, Understanding Public Policy (Englewood Cliffs, N.J.:
ncice-Hall, 1972) for a discussion of a number of models of policy analysis.
Chapter eleen, Dye concentrates on what I term the demographic model.

6 Robert Horlan, "Life on the City Council: Realities of Legislative
icics," Capitol, Courthouse and City Hall, ed. Robert Morlan (4th ed;
con: Houghcon-Mifflin, 1972), p. 213.

7 Heinz Eulau, "Logic of Rationality in Unanimous Decision-Making,"
ional Decisions, Vol. VII, Nomos, ed. Carl Friedrich (New York: Atherton,
7), pp. 26-54.

S Ibid., p. 29.

9 For a complete survey of this literature, see Brett Hawkins, Politics
Urban Policies (Indianapolis: Bobbs-Merrill, 1971).

10 Calculated from statistics provided in U.S. Department of Commerce,
eau of Census, General Social and Economic Characteristics: Florida, 1970
sus of Population, PC(1)-C11.

11 Responses from four cities indicated four different council seats
e vacant either through death or resignation. The list of council members
obtained using Florida League of Cities, Officials of Florida Municipalities:
4 (Tallahassee, Florida: Florida League of Cities, 1974).

12 The response rates from mail questionnaires are often quite low.
ever, the problem is especially acute when the respondents are not pro-
;sionals of some sort. See Delbert Miller, Handbook of Research Design
SSocial Measurement (New York: McKay, 1964), p. 77. For example, Benedict
ider had a response rate of 40 percent among council members in cities of
r 10,000. See Benedict Kweder, "Beliefs, Attitudes, and Policy Preferences:


~P ~IC~-~ I rl I ----r I r r







Views of Some North Carolina Mayors, Councilmen, and City Managers,"
:. (unpublished), University of North Carolina at Chapel Hill, 1972,
87. Lyman Kellstedt had a response rate of 34 percent among Philadelphia
Lcil members. See Lyman Kellstedt, "Precinct Committeemen in the Phila-
,hia Metropolitan Area: An Analysis of Roles,: Diss. (unpublished), Uni-
;ity of Illinois, 1965. Ronald Matheny's response was just 29.3 percent.
Ronald Matheny, "Role Conflict: A Quantitative Analysis of the Urban
Islator," prepared for delivery at American Political Science Association
:ing (Chicago: August 29-September 2, 1974). Only when the researcher is
Sto get the help of governmental officials does the response rate exceed
percent See Ross Robson and James Sheffield, Jr., "Effects of Population
i on Problem Perceptions and Policy Preferences of Local Government Offi-
Ls," prepared for delivery at Annual Convention of the Southern Political
mnce Association (New Orleans, Louisiana: November 7-9, 1974). Additional
)rmation can be found in Eli Cox, III, Thomas Anderson and David Fulcher,
appraising Mail Survey Response Rates," Journal of Marketing Research,
.II (June, 1974), pp. 413-417, and Herbert Blumberg, Carolyn Fuller and
L Hare, "Response Rates in Postal Surveys," Public Opinion Quarterly, Vol.
JIII (March, 1974), pp. 113-123.

13 John C. Wahlke, Heinz Eulau, William Buchanan and Leroy Ferguson,
Legislative System: Explorations in Legislative Behavior (New York:
ey, 1962), p. 378.

14 Raymond A. Bauer, Ithiel de Sola Pool and Lewis Dexter, American
iness and Public Policy: Politics of Foreign Trade (New York: Atherton,
3), p. 404.

15 Roger Cobb and Charles Elder, Participation in American Politics:
Dynamics of Agenda-Building (Boston: Allyn and Bacon, 1972), p. 58.

16 Harold Lasswell, Politics: Who Gets What, When, How (New York:
raw-Hill, 1936).

17 Hawkins, op. cit.

18 Dye, op. cit.

19 Jonathan Turner, The Structure of Sociological Theory (Homewood,
Dorsey, 1974), p. 212.

20 This approach is adopted from that presented in James Buchanan and
don Tullock, The Calculus of Consent: Logical Foundations of Constitutional
ocracy (Ann Arbor: University of Michigan, 1962).

21 Terry Clark, Community Power and Policy Outputs: A Review of Urban
each (Beverly Hills: Sage, 1973).













CHAPTER TWO
THE POLICY FOCUS, CITY COUNCILS AND EXCHANGE THEORY

"The city is manifestly a complicated thing."I No less is local

government. It stands as an intricate artifice serving several intents.

At base though, local government is charged with performing several po-

licy functions.2 However, municipal public policy issues from many

sources within local government and is conditioned by a welter of commu-

nity forces. In this chapter, I am concerned with sketching a framework

whereby the policy machinations of local councils might be considered.

My focus is narrow. I deal only with city councils and the way in

which they shape policy. But the policies with which councils deal range

across a broad spectrum of concerns. The scope of these matters is not

easily apprehended by many of the perspectives commonly used to analyze

public policy. Moreover, many of these policy perspectives do not faci-

litate investigation of the policy process, and it is in this process

that the council fits. The policy focus I present here turns on the no-

tion of choice and flows from exchange theoretical considerations. This

view seeks to incorporate a regard for the manner in which facets of the

council decision-making process interact with the pressures placed on

the council by forces in the community and thereby shape different poli-

cies. It also recognizes that policies themselves take on meanings

within a political context and act to condition the politics of their

consideration.





Policy Models and the Policy Focus

While it is true that "policy is not a new concern of political

science," it is nonetheless the case that the last decade has witnessed

a renewed concern with and interest in matters of public policy.3 Cer-

tainly the explosive social events of the sixties and the great number

of federal initiatives in social legislation contributed to this re-awa-

kening. But the phenomenon can also be traced to ongoing discussions be-

tween scholars and a changing view of politics and public policy within

the literature of political science.

"Once upon a time," Charles O. Jones has written, "V.O. Key reflec-

ted on some of the possible consequences of one-party factionalism in

the South. He (Key) was particularly interested in speculating about

the effect of 'disorganized politics' or 'loose factionalism' on the

match between the 'haves' and the 'have-nots.'"4 At the state level,

Key argued, the match usually came down to matters of taxation and ex-

penditure. Viewing politics as "the conflict between those who have and

those who have less," Key speculated that for the South of the late for-

ties an understanding of governmental action was to be found in the vary-

ing shades of political organization within the states. Key searched

the factional arrangements of the southern states for the basis of both

their differences and their similarities.5

However, Key's interest in such arrangements stemmed from a larger

concern with the way in which southern policy preferences over the is-

sue of race helped perpetuate one-party systems. The need to portray a

single front in national political forums such as Congress helped estab-

lish a reliance on a single party for national articulation of policy.

Key felt that "southern political regionalism derives basically from the


~P------;--"-aFlplp-r I~SII~-C- J~II~ILcll~-4~Y I~-WPsP11~6L






influence of the Negro," and he concerned himself with demonstrating the

manner in which this influence affected politics in general and party

politics in particular.6 Key assumed two-party politics might signal a

change in policy, but he recognized changes would not come automatically

with a move to two-party politics.7 Key was intent upon sketching the

broad outlines of policy movement. As a result, his portrayal of the

determinants of public policy avoids a rigidly deterministic conception

of the consequences of organization for policy. Instead, Key's approach

to the consequences of politics for public policy involved a very com-

plex and subtle view of the interplay of forces working within the sou-

thern states.

It is not surprising, of course, that political scientists such as

Key would undertake studies involving the relationship between politics

and policy, and ultimately conclude politics did influence policy. In-

deed, such findings almost seemed axiomatic; the study of politics was

predicated upon its importance for policy. However, not all researchers

have documented this importance.

The level and nature of public expenditure, or the form of public

policy, had always been a relevant concern for economists. And interest-

ingly enough, when such economists attempted to account for the differ-

ences in the levels of spending of the various states, political varia-

bles were seldom included in their models. Economists such as Solomon

Fabricant and Harvey Brazer found certain aspects of the social and eco-

nomic environments of the states were the most important determinants of

spending levels. Factors such as urbanization, income level, and indus-

trialization were viewed as important to the exclusion of political vari-

ables. "In brief, the political scientists, considering only political





variables, found these to be determinants of policy, while the economists,

using only economic variables, found economics to be crucial."8

In an attempt to reconcile the difference between the two perspec-

tives, Dawson and Robinson set out to "discover the relationships among

the extent of inter-party competition, the presence of certain economic

factors, and the extent of nine public welfare policies, using the Ame-

rican States as the units for investigation."9 Dawson and Robinson re-

lied on the well-known model of the political system framed in the work

of David Easton.

At heart, this conception of politics argues

that certain environmental conditions, primarily socio-
economic in nature, combine with certain characteristics
of the political system to produce public policy . .
Policy, in this view, is the outcome of activity or inter-
action among external conditions, political system and
political process. 10

As employed by many scholars, this model was designed to bridge a gap in

previous studies. It was intended to include both political variables

and environmental variables, and to be generally inclusive of a wide

range of political and social situations. Moreover, this model allowed

for a systematic and precise examination of the effects of each of the

variables on measures of public policy through multivariate techniques.

Indeed the synthesis embodied in the model did seem to focus square-

ly upon the determinants of public policy. The Dawson-Robinson study,

like those of Thomas Dye which followed, was able to conclude the role

of political processes in the determination of public policy had been

overdrawn. Dye, for example, suggested that "political science has been

guilty of viewing political life as a closed system."11 The thrust of


ilio Errli-rC L -t-e JI




30


Dye's findings pointed toward the relative importance of certain external

conditions in the formulations of public policy. Such conditions, he ar-

gued, were more influential in shaping policy than had been suggested in

the writings of many political scientists. However, many political sci-

entists demurred in the wake of such findings. New tacts were taken to

demonstrate the pre-eminence of politics. New variables were introduced

into the model; expenditures were criticized as poor indicators of policy;

additional linkages were specified.12 At best, the results of these ef-

forts were ambiguous. They never seemed to catch the importance of po-

litics for policy which appeared all too obvious to political scientists.

The poignant observations of.scholars such as Key were by-passed in the

refinements of the new approach.

While the faults of the analysis offered by Dye and others have, by

this time, been repeatedly dealt with in the literature of political sci-

ence, one point often goes unnoticed.13 Language both informs thought

and betrays it. The synthesis constructed by Dye and accepted by many

others as the basis for discussion about matters of policy purported to

deal with the same questions raised in previous considerations of the

topic. Heuristically, the simplified systems model promised to clarify

the discussion of public policy and to highlight prominent relationships.

In fact, this seemingly benign simplification significantly altered the

political and economic images embedded in the conceptions of public po-

licy which had preceded it. The terminology persisted, but acted to

serve different ends. The ramifications of the "systems" model used by

Dye and many others entailed a synthesis that was mainly illusion, based

upon a common vocabulary but offering quite different hypotheses about r

the nature of the influences shaping public policy. 0


.





In most of the discussions that animated the debate over the syn-

thesis suggested by Dye, the concern focused upon the appropriateness

and importance of either political or environmental variables. Hypothe-

ses were offered that talked of the determinants of public policy, but

the form of the hypotheses was typically ignored. The particular form

of the systems model assumed during the debate cast explanation of po-

licy in terms of two alternatives, environmental determinism as opposed

to free-will environmentalism.14 Authors such as Fenton, Lineberry, Fow-

ler, and Clarke who contested the evidence pointing to the importance of

environmental considerations in the determination of outputs responded

with metatheoretical statements about the mediating role of political

processes.15 But, in fact, it was a sense of process which was missing

in most of the hypotheses. The operational measures used in the inves-

tigations were quite varied, and yet researchers grouped them into two

general types--indicators of political processes and indicators of envi-

ronment. The complex dynamics by which various factors acquired import

for public policy were largely neglected. Specifically, in eliminating

the role of policy makers from their direct consideration both sides of

the debate had to content themselves with speculation about the manner

in which variables became important for public policy. Many of the most

important questions about the determination of public policy per force

had to remain unasked and unanswered.

For all the problems of the model developed by Dye, for all the cri-

ticisms of it, for all that, one impression remains. The literature of

political science has been fundamentally changed; its course redirected.

The preoccupation of political scientists with system variables noted by

Dye, while still evident, has given way to a generally shared concern


Pzll`Z* ~F~~~i~-l;lll~i~'~~UIC"~IC19C --`~EI~CE~bn'~




for public policy.16 Even traditional approaches to the study of poli-

tics are being re-evaluated in terms of their ability to shed light upon

considerations of policy.

In the last several years, students of public policy and of munici-

pal government have resorted to several approaches focusing on what Ver-

non Van Dyke has called salient-political-features.17 Lacking a well

defined or established approach to the study of public policy, some au-

thors have argued for the importance of elections in shaping municipal

policy. Local institutions, pressure groups and powerful influentials

have marked the investigations of still other researchers.18 Political

cleavages have also been taken to have meaning for a city's public poli-

cy, and researchers have examined the social and economic characteristics
19
of municipalities to produce surrogate measures of such divisions.1 In

this view, the political system is viewed "as a more or less competitive

arena in which groups compete for political advantage and in which indi-

viduals take cues for political behavior from the reference groups with

which they identify."20 Social and economic factors have been used to

buttress yet another conception of policy variation which appeals to the

underlying cultural differences of cities. In this view, the "ethos"

discernible in a city's social, economic, and political structures has

an effect upon the types of policies adopted by a city.21

All this is not to argue political influence is unimportant as a

concept or that various models of governance have no merit. To the con-

trary, what the growing body of policy studies in American urban politics

portends is a shift in perspective--one that extends greater importance

to policy and its consequences while encompassing the older concerns for

political influence and community stability. The recent literature





not only echoes the attention given the distribution of values in pre-

vious studies, but also questions the range and scope of policy alterna-

tives available. The much heralded policy focus involves a subtle shift

in our image of political man; he is now seen to be someone with a choice

to make. Politics has become an activity whereby the preferences of

some individuals are selected to be the choices of society.22

Choice, however, has been the point of departure for several concep-

tions of politics. For example, David Easton's attempt to deal with the

"authoritative allocation of values" naturally concerned itself with

choice and the manner in which norms and values in society act to both

limit and define the possible choices by governmental bodies in that so-

ciety. Similarly, the various schools of decision theorists have long

focused upon choice situations. Their manner of speculation on the ra-

mifications of choice for the entire polity is extensive and varied.23

To the extent a policy focus subsumes and borrows from these conceptions,

the approach will stand on the belief that choice is an important "focal

point of the social sciences, uniting normative, structural and behavioral

theory."24

Yet a policy focus entails more than decision making, more than a

mere concern for choice. "Policy making is the process of making choices,"

as Apter has suggested, but a research focus must be informed by more

than such a simple prescription.25 Perhaps best illustrating this point

is Eulau and Prewitt's insightful disquisition on "What Policy Is Not."26

Policy is not, they argue, a synonym for goals and objectives. It is not

an intent to do something. Nor is it a synomym for particular events in

the policy process, such as the actions of a governing body. And it is

most certainly not mere decisions or actions. Rather, "policy is a strict-


- _1~L_~BB~:







ly theoretical construct that is inferred from patterns of relevant

choice behavior by relevant political actors and the consequences of

choice behavior. . .If the behavioral patterns are consistent and re-

gular, then the existence of policy is inferred and identified."27

Choice is a key element in this formulation; it establishes the basis

for our inference about political man and the political system. But the

policy focus is more than"a theory of one segment."28 In order to be

meaningful, the focus itself must be illuminated by images of man and the

political system which supports its basic thrust. Its application must

handle the contextual relevance called for in Eulau's definition. In

particular, the images animating the focus must be geared for the re-

search in question.

Approaches to the Study of City Councils

A policy focus is fundamental to the problems to which this study

is addressed. However, such a focus cannot stand apart from the problems

themselves. To the contrary, the two complement each other. Moreover,

the tractibility of such problems is likely to vary considerably accor-

ding to the theoretical schemes employed to approach them. In social

science, where various images of man and society abound, some perspec-

tives appear suited for answering certain questions and ill-suited for

others. A brief review will show this to be true of the consideration

scholars have given city councils.

In the last ten years, several authors have attempted to deal with

city councils and their role in the formulation and promulgation of ur-

ban public policy. I will examine four studies. To varying degrees,

each exposition has been illuminated by different images of the politi-

cal system and of political man. Each has contributed valuable insights

into the evolving notion of municipal policy making.





John Crecine's Governmental Problem Solving is the most explicitly

modelled and narrowly conceived of the four. Testing a complicated simu-

lation model of big city budgetary processes against the actual expendi-

tures in Cleveland, Detroit, and Pittsburgh, Crecine postulates a model

consisting of three main elements or processes: a departmental or bureau

process, a moral process, and a city council process. Each interact to

produce a budget. For Crecine, though, "the role of the city council is

a limited one."29 In fact, he argues that the city council has only se-

ven feasible strategies for dealing with a mayor's proposed budget.30

As Crecine views the budgetary process, the serious constraints upon the

council in terms of time, expertness, and information make it little more

than a rubber stamp for the mayor's proposals. Thus it is not surprising

Crecine finds councils do tend to produce decisions on expenditures close-

ly paralleling those recommended by the mayor.31 Only in cases where

the council chooses to alter the tax rate does Crecine acknowledge dif-

ficulties with this conception.

The thrust of Crecine's conclusions are not dissimilar from those of
i
Wildavsky and others writing of federal-congressional budgeting. In a

sense, Crecine is saying that if the budget ceiling is fixed, no new re-

venues are raised (his model has a difficult time explaining that aspect

of council behavior), and no catastrophic events hit the community, then

the budget levels, examined by function, will not change very much. The

argument becomes that if very little changes within the community, then

budget levels change very little as well.

This is not to demeanCrecine's work, but rather to caution about its

validity under circumstances significantly dissimilar from those posited.

Certainly it is true that Crecine has captured an aspect of municipal bud-

geting observed in many cities throughout the United States. Namely,


iji





little change does occur from year to year in expenditures, and his as-

sumptions handle this particular phenomenon quite well. However, aside

from the simplified nature of the assumptions of the model, Crecine's

findings are limited by the process he studied--budget making--and the

context from which he gathered his data--large cities with strong mayor

forms of government, cities whose council's have been viewed as rela-

tively weak when compared to other community political actors.32 Clearly

for the problems to be considered in this study, Crecine's approach and

his assumptions seem inadequate.

John Dearlove has offered a different approach. Studying council

activities in Kensington and Chelsea, England, Dearlove starts from a

vastly different perspective than Crecine but nonetheless concludes by

buttressing Crecine's view. Dearlove suggests, after the manner of Cre-

cine, that there are strong tendencies among councils toward policy main-

tenance. However, this tendency flows not from the weakened position

of the council or the constraints upon its information gathering abili-

ties, but rather from the council's strengths. Dearlove concludes that

councils are entities able to control and shape the environment, but,

for several reasons, are not inclined toward change.33

Clearly, though, Dearlove is only considering one side of the coin.

By ignoring the conflictual nature of policy change, he has excluded an

exceedingly important aspect of the policy process. This is not an in-

advertant omission, however, but rather a product of the model used by

Dearlove. He relies extensively upon interaction theory, and more spe-

cifically, on role theory. Role theory, by positing a structured vision

of human behavior based upon regularized expectations, assumes a well

defined and persistent form of social organization.34 The veteran of

any council meeting might recognize the intuitive appeal of this position





for an analysis of council activities. But councils also share the

characteristics of small groups. Consensual aspects may predominate

but there exist strong tendencies toward shifts in behavior.

Additionally, councils are themselves not typical of many organized

small groups. In fact, they might aptly be characterized as organized
35
anarchies. That is, they are organizations with three general proper-

ties. The first is problematic preference; the council "discovers pre-

ferences through action more than it acts on the basis of preferences."36

Second, although the organization manages to survive and even operate

smoothly, its own processes are not understood well by its members; tri-

al and error procedures predominate. Third, "participants vary in the

amount of time and effort they devote to different domains; involvement

varies from one time to another."37 To the degree these statements ap-

pear to capture salient aspects of council life, the assumptions of role

theory become quite tenuous and of limited use in addressing the prob-

lems motivating this study.

The third of the major studies dealing with city council activities

and public policy was done by Bryan Downes. Quite distinctively, his

research sets out to investigate communities and councils marked by

changing values and divergent community interests. His dissertation,

along with several published articles, focuses upon 37 suburban munici-

palities in the St. Louis region which had experienced varying degrees

of population growth during the fifties. Downes begins his study with
-

a framework drawn from the work on community conflict conducted by James

Coleman and William Gamson.38

Downes arrives at several conclusions. He finds that councils do

disagree and often, that council decision processes have little effect on M

YIt
9:<

Hill 1 r~ilriff ~qf^BSR





aggregate measures of a community's policy commitments, that member

background seems to have some association with the policy choices of the

council, and that, in the main, "socioeconomic characteristics are pro-

bably more important determinants of municipal fiscal and land use poli-

cies" than political processes.39 He traces this importance to the man-

ner in which socioeconomic changes manifest themselves in the divergent

values and interests of the new community inhabitants associated with

population growth. This relationship is not an automatic one, however.

Rather different issues surface in different situations and spark vary-

ing degrees of dissension among the council members. For example,

Downes indicates "In very high growth rate communities, issue disagree-

ments appear to be largely a function of enduring factions, and these

cleavages affect the extent of conflict, on most issues. On the other

hand, in medium growth rate communities, issue disagreements did not ap-

pear to be the result of enduring council splits."40 While certain is-

sues such as zoning and land use planning seem to generate much contro-

versy in most cities, other issues are not clearly associated with growth

patterns or socioeconomic variables.

Downe's study is directed toward many of the same concerns motiva-

ting this research, and hence his approach is of particular interest.

He deals with council conflicts and unanimity. He stresses linkages be-

tween council member characteristics and municipal policy. Yet for se-

veral reasons, Downes' framework appears inadequate for the present pur-

poses. Methodologically the systems framework and the holistic concep-

tion of community are not well suited to considering council members as

the fundamental units of inquiry. When dealing with members, as he often

does, Downes resorts to ad hoc explanations; his theoretical framework







betrays the evidence he musters. Unlike Dearlove, who :--ork-s '-ithlin the

heritage of role theory, Downes' legacy flows from the i.-ellsprings of

conflict theory. When the unit of analysis is the comiiunit', and the

focus is on socioeconomic measures, all is well and Downes is convincing.

However, as Downes stoops to include council members and cieir relation

to policy, he falters. Having assumed council members to be reflections

of the broader community context, the framework fails to establish ade-

quately the link between members and policy, and ignores Dearlove's in-

sights regarding the autonomy of the council. Indeed nost .f D-o.ynes'

indicators of member perceptions only dimly illuminate the manner in

which members help determine-policy. The concern for community differ-

entiation, conflict and policy is salutary, but Downes' approach is un-

likely to produce answers to the puzzles considered in this stud',; it

assumes many of the links this study will investigate.

The City Council Research Project directed by Heinz Eulau and Ken-

neth Prewitt is perhaps the single most ambitious attempt to gauge the

relation between environmental challenges and policy responses in terms

of the purposive actions of decision makers.41 Begun primarily as a con-

tribution to the study of legislative institutions, the project emerged

as a highly complex and intricate view of the entire policy process, and

the most detailed and thorough study of council activities to date. Pro-

lific in number, varied in their focus, the several books and articles

derived from this investigation nonetheless seem to forge a common theme.

Namely, policy positions are emergent responses by decision makers to

the ecological situations and their own predispositions which have been

conditioned by specific policies adopted over time.42 MLuch of this work

argues that there is cognitive adjustment to environmental pressures and






to the previous patterns of decision making. The underlying image of

political man that issues from such conceptions emphasizes his adaptive

qualities, as one who engages in costly processes of social interaction

and who acts under conditions of uncertainty with notions of what has

been done before and what might work at any given moment.

For all its descriptive detail and calculation, however, the City

Council Research Project is supported by a meager theoretical framework.

In fact, little pretense is made at presenting a theoretical framework.

The major work of the project, Eulau and Prewitt's The Labyrinths of De-

mocracy, is content to portray its foundations in the metaphor supplied

in the title. As Eulau and Prewitt explain:

We find the metaphor of governance as a labyrinth prefer-
able to other metaphors used in politics. A labyrinth is
an enclosure with many entrances and exits. Its layout con-
sists of a maze of pathways, but the pathways can be marked
by signs that help one to avoid false moves and lead the
seeker where he wants to go. If the paths are not marked,
trial and error may yet lead to discovery and the long way
back. Democratic governance resembles the labyrinth. The
labyrinth has walls that serve as boundaries but are more
or less porous; it has major arteries and places of assem-
bly but also byways, detours, nooks and crannies. Passing
through the labyrinth may take more or less time, depend-
ing on the continuities and discontinuities in the journey
by what is known or unknown about the terrain.43

Understandably, Eulau and Prewitt wished to avoid portraying democratic

governance in oversimplified terms, to not "reduce the maze of democra-

tic governance to a simple formula."44 However, even this metaphor is

a sort of formula, albeit an intricate one, and to that degree draws the

politics of councils all too harshly. Moreover, this particular formula

seems endowed with few terms with which to model the political context

of councils or from which to draw hypotheses.

Of course, Eulau and Prewitt do animate their metaphor. Implicitly,

the perspectives they develop appear to be variants of interaction theory.






Social intercourse constitutes one of the main thrusts of their work.

In attempting to go beyond the assumptions of the Legislative System,

the authors draw a picture of city councils as adaptive organizations.45

The adaptive aspects of councils are painted in terms of internal pro-

cesses such as stratification, governing styles, communication, task

structure and decisional structures; external linkages such as consti-

tutive processes, petitioning processes and member orientations toward

these phenomena; and finally, policy as a response emerging from purpose-

ful, but adaptive interaction of council members.46

However, the study never goes much beyond these conceptions. Even

these concepts appear with little or no recourse to theoretical concerns

or the vast literature on empirical small group research. The interac-

tion perspective that guides much of their discussion deserts them in

forging the links between council processes and decision structures.

Only in the work of Eyestone and later Zisk does the project face the

task of relating conversion processes to policy outputs in an explicit

fashion.47 But these latter efforts, while themselves imbued with con-

siderable merit, lack the more comprehensive approach and detail that

sets Eulau and Prewitt's study apart. Much more explicitly Eulau and

Prewitt need to examine the kinds of connections which Banfield discussed

years ago; namely, how does the distribution of influence and demands

made upon decision makers combine with the decisional system so as to

explain the types of policy a city council adopts?48

An Exchange Perspective on Policy Making in City Councils

Each of the above studies has its merits. Crecine's successes stem

from his elucidation of portions of the budgetary process and the manner

in which he traces the limitations of the council in a severely constrained


IL - ~. L I~L-l ~ --1 II IC dBFir







political situation. Dearlove, on the other hand, is at his best when

the discussion turns to the extent of council latitude for action and

the means whereby such potential narrows under self-regulating adjust-

ments. The factional aspects of council politics and their importance

in many policy situations are handled deftly by Bryan Downes. His em-

phasis on change and conflict in the urban political scene is a welcome

addition to the literature on council politics and policies. Finally,

of course, Eulau and Prewitt's Labyrinths of Democracy provides the most

extensive and complex view of municipal governance to be found in the

literature. Replete with keen insights and encyclopedic in scope, it

portrays councils as both a product and producer of community policy.

Yet for all these merits, the various perspectives developed by these

authors are not well suited for the problems at hand. While one is well

fashioned to handle conflict and another unanimity or yet another to forge

the link between council activities and public policy, the common deno-

minators are few. Indeed most are not well suited to the notion that the

council itself is the focus of inquiry, the wellspring of a municipal

policy focus.

This study is predicated upon the notion that the council serves as

a breakwater for much of the public policy pursued within a community.

Regardless of its ultimate importance in the determination of a communi-

ty's public policy, the council nonetheless provides one of the few insti-

tutional arenas to which most everyone in the community can have recourse.

Minimally then, much of the potential for changes in community public

policy lies with the council and its politics. In order to capture the

possibilities inherent in the council circumstance, an adequate framework

must amplify the policy focus discussed earlier. If the policy focus




43


serves to direct our attention to the choices of the members themselves,

then it requires a conception of council politics descriptive of the

important links in the process of forming such policy. Although the

difficulties with the approaches adopted by Crecine, Dearlove, Downes

and Eulau have been described, this study hopes to benefit from their

insights indirectly by incorporating them into a different approach to

the problem. Exchange theoretical considerations will guide this inves-

tigation.49

Most contemporary versions of exchange theory revolve around and

differ with respect to three basic concepts: 1) the idea of utility or

costs and rewards; 2) rationality and its implications with regard to

information, alternatives, and time; 3) strategic interaction or the na-

ture of interpersonal interaction and anticipation among individuals.

As a whole, these concepts provide the link between council members

and public policy.

Most exchange theoretical expositions hold as their basic tenet an

assumption concerning the motivation of individual behavior. In essence,

such assumptions can be summarized as follows: "The representative or

the average individual, when confronted with real choice in exchange,

will choose 'more rather than less. 't50 In general, such propositions

are derived from considerations about an individual's psychological re-

action to various stimuli in terms of his inner needs or internal dispo-

sitions.51 When a person receives or has directed towards him an acti-

vity defined by that person as valuable, he is rewarded.52 Rewards, in

such conceptions, are what contribute to the desire for "more." On the

other hand, an activity that is punitive to the individual, or an alter-

native reward that is foregone in order to get other rewards, is a cost


r
i





















'

"-'
:r.
;'-
~ii-
~




i

-

;I:
'''







and makes up what has been called "less." In any case, the idea that

individuals have needs and seek values in their behavioral choices and

transactions is fundamental to exchange theory. Such needs and values

are of course quite diverse and the usefulness of exchange considerations

is to a large part to be found in how such factors are conceived with re-

ference to an empirical situation.

To say that men do have preferences and that in choice they seek

personal values seems reasonable only if such individuals are somehow

able to consistently realize their values in choice. While several ex-

planations can be had for the connection between internal dispositions

and external actions, exchange theory assumes that "people behave as if

they arranged their preferences in some logical pattern."53 That is,

somehow people are able to know which value they seek in a given situa-

tion. Regardless of the approach, the key to understanding the thrust

of exchange theorists on this point is to be found in the concept of ra-

tionality. "Rationality is the something we postulate in people that

makes them behave in a regular way. And the essence of that something

is that people relate their actions to their goals."54 In exchange theory,

man is a teleological entity and rationality is the manifestation of his

movement towards his goals, or desired rewards. Observed regularity is

the proof of the pudding.

While rational behavior may be purposeful, not all exchange theor-

ists view purposeful behavior as rational. This difference stems from

the various ways in which the concept of preference has been approached

by such theorists. Of course, preferences can never be observed directly.

Instead they must somehow be imputed. This inference can be made in ei-

ther of two ways: preference can be revealed or preference can be posited.






In the procedure of revealed preference, we assume ini-
tially that persons behave in accord with logical rules,
e.g., transitivity in ordering. Applying these rules to
choices, we then discover what goals must have existed
in order to lead logically to these choices. Goals are
therefore inferred from a)actual choices and b) assump-
tions about behavior.... In the procedure of posited pre-
ference, we assume initially that an actor has a given
goal (e.g., to win an election, to maximize profit) and
we infer that behaving in accordance with rules of logic
and this goal leads to particular choices. If such
choices actually occur, then we further infer both that
the actor actually does have the assumed goal and that
he behaves in accord with the rules of logic. If such
choices do not occur, then we are at a loss to discover
whether the fault lies in our attribution of logic or
our attribution of goals.55

In fact, there is switching back and forth between the two approaches.

But the use of posited preferences has suffered from a confusion over

their utility and their extensive association with the works of norma-

tive rational theorists.

In part, such confusion results from the failure of many research-

ers to apprehend the full meaning and usage of the term rational in the

context of the underlying notion of preference. If preference is taken

to be revealed in behavior, then the investigator has only to make the

connection between stimuli and behavior; few additional assumptions are

required. On the other hand, theorists using posited preferences re-

quire several additional assumptions. First, a set of preferences must

somehow be identified. Second, an assumption of transitivity is usually

required. Without this assumption to guarantee that persons can somehow

arrange their preferences in some order, the logical relationship between

values and goals would be impossible to establish. Finally, under posi-

ted preference connectivity is required. That is, in some manner the
T-
objects of value in an individual's mind must be comparable. Under con-

ditions of revealed preference, this is no problem. Instead, given a





........ ...... r t L 1 I I -I ....







perspective relying upon revealed preference, the primary difficulty

comes in knowing what irrationality is and in dealing with intensities

of preference.

Regardless of the approach taken, exchange models draw the relation

between internal dispositions and behavior. The manifestation of this

relation comes when an individual, such as a council member, attempts to

gain some benefit. Many conceptions of this process insist that the in-

dividual engages in utility maximization, others argue that people sa-

tisfice, while still others attempt to avoid the issue by concentrating

explicitly upon revealed preferences.56 In any case, the problem cen-

tral to such concerns deals with the way an individual or a council mem-

ber goes about generating different courses of action. That is, how are

alternative choices generated for individual consideration?

Clearly, an awareness of alternatives is critical to the possibility

of choice. In a rational process, an individual must choose his most

preferred alternative, yet the question arises of how the individual de-

termines the set of alternatives for evaluation and what the set may be.57

Under considerations of utility maximization, perfect information is re-

quired. However, this criterion appears untenable in any real council

situation, and thus a more reasonable standard might incorporate Herbert

Simon's famous view that men satisfice. That is, "in the complex envi-

ronment of decision making in the real world, choosers are not aware of

all possible alternatives, so that they choose not the best alternative

but a satisfactory one. They do not maximize, but satisfy."58

The main difficulty with such a notion, however, lies in the manner

in which the individual defines the set of alternatives from which he

ultimately makes his choice. If all possible alternatives are not evi-







dent, then an individual must search in order to establish the set.

How does an individual judge that his search has ended? Simon suggests

that an individual's satisficing level rises and falls with the ease by

which he gains a satisfactory outcome. Simon's description "allows for

a person's coping with the varying costs and benefits of searching for

alternatives across contexts, situations, and time. Thus, what satisfices

is related to the nature of the environment in which the chooser finds

himself, and the method of satisficing allows him to learn about the na-

ture of that environment."59 Under many of the most rigid exchange con-

ceptions, such a dynamic has no part; learning plays no role. Simon's

solution, however, places the choice situation in a context that incor-

porates conception of the interaction between an individual and his en-

vironment as well as his past experiences. Notice that this description

squares with the contextual aspect of policy as defined by Eulau.

The problems posed by assumptions of rationality are manifold. In

this study, only the weakest form of rationality is assumed--that coun-

cil members be able to select among their perceived set of alternatives

one which seems most likely to realize the most benefit in terms of their

goals. The calculus of benefits is based on probabilistic expectations

about the success of certain alternatives in obtaining given outcomes

within a given context.

This calculus involves more though than merely determining what

value is to be maximized. There is a crucial complementarity between

preference and the social dynamic imputed by the assumption of satisfi-

cing. Seldom are values pursued in isolation from the vagarities of

other people's actions. Moreover, even for a given council member, the

pursuit of values changes with time and circumstance. Implicit in most







exchange theoretical perspectives are the familiar notions derived from

economics, of marginal utility, strategic interaction, and externalities.

Pursuing one course of action, usually limits an individual's abili-

ty to actively pursue yet another course. Marginal utility connotes a

situation where additional units of a rewarding activity begin to gene-

rate costs that exceed the benefit received. Under conditions of ration-

ality, when these costs exceed the benefits accrued an individual ceases

his behavior. Of course, an individual will still desire "more rather

than less" but the concept of marginal utility raises the question of

how much more.

The pursuit of values is further constrained within the exchange

perspective by an additional assumption about behavior. Namely, the

complex set of internal calculations involved in the evaluation of per-

ceived alternatives will normally entail considerations about the "pro-

bable choices of others (given each of his own possible actions) when-

ever such choices by others would affect his evaluation of his own alter-

natives."60 Such behavior on the part of two or more interacting indi-

viduals is referred to as strategic interaction. This assumption amounts

to no more than the idea that each person's choice is contingent on the

choices of others. It means that despite the individualistic approach

indicated in exchange theory, there exists an implicit requirement for

the consideration of the collective context of individual decision ma-

king.61

The assumption of strategic interaction not only places bounds upon

an individual's attempts to realize his values, it also introduces an

indeterminacy into exchange models not present in many of the traditional

economic models of the market and economic man.62 This indeterminacy is




49 4


the result of the uncertainty concerning individual behavior introduced
-p
by the attention given the actions of others. In a collective body, in-

dividual rationality does not obtain. By creating considerable uncer-

tainty in the likelihood of certain alternatives to produce expected be-

nefits, and by diffusing the correspondence between choice and outcome

for an individual, the predictive aspects of the model are weakened. Ad-

ditionally, such difficulties demonstrate the necessity of tuning the ex-

change perspective more closely to the empirical referent than might other-

wise be the case. General theoretical statements need to be tempered by

additional assumptions about the characteristics of council members and

the council decision-making situation.

The notion of strategic interaction incorporates a familiar concept

into most exchange theoretical expositions--externalities. "An external-

ity exists if at least one person in addition to the subject can alter

the relation between outcome and alternatives with his choice."63 In the

calculations engendered by strategic interaction, it is the evaluation

and computation of perceived externalities which help determine an indi-

vidual's choice. However, in a collective choice situation it is useful

to differentiate between externalities over which a council member has

no control and those costs he incurs through his participation in a col-

lective decision-making process.

The individual's utility derived from any single human ac-
tivity is maximized when his share in the "net costs" of
organizing the activity is minimized. The possible benefits
that he secures from a particular method of operation are
included in this calculus as cost reductions, reductions
from that level which would be imposed on the individual if
the activity which were differently organized. There are
two separable and distinct elements in the expected costs
of any human activity which we want to isolate and to em-
phasize. First, there are costs that the individual ex-
pects to endure as a result of the actions of others over
which he has no direct control. To the individual these
'r






costs are external to his own behavior, and we shall call
them external costs, using the conventional and descrip-
tive terminology. Secondly, there are costs which the in-
dividual expects to incur as a result of his own partici-
pation in an organized activity. We shall call these de-
cision-making costs.64

Buchanan and Tullock argue that expected external costs can only be

eliminated if the decision-making rule by which choice is made requires

unanimous agreement. This is so since the rational individual will not

willingly allow others to impose external costs on him in most circum-

stances. On the other hand, under the rule of unanimity the expected de-

cision costs are quite high since every individual must be party to any

agreement. Thus the rational individual will attempt to reduce these

costs through some sort of balancing or trade-off between the two.65

In order to make meaningful theoretical statements about council

processes and public policy, the discussion above indicates the need for

some rough gauge of the benefits and costs accrued to council members.

The council member must make several judgements before he makes his choice,

weighing the several values that might be involved against the costs he

is able to predict associated with each action. In determining exactly

what the balance might be in a particular choice context, each member re-

lies, necessarily, on his previous experiences and a host of learned re-

sponses. In familiar choice situations, a member is likely to be more cer-

tain of his decision, and more likely to bear costs with an expectation

of benefit. In uncertain situations, a member's expected utility drops

and the importance of cost factors rises. Each member is conditioned by

a system of empirical beliefs, expressive symbols and values which help

define the situation in which he must choose.66 While each member may

pursue purposes of his own within the collective body, his ability to re-

alize these purposes is constrained by his own predispositions about the







manner in which values should be realized, and the costs incurred in

reaching and enforcing the means embodied in such actions. Analytically,

benefits and costs must be viewed in terms of posited preferences and

revealed predispositions.

These notions can be illustrated for city council members by appeal-

ing to the distinctions Richard Fenno has drawn about the motivations of

Congressmen. Fenno has argued that within the House of Representatives

members strive to attain various goals. In particular, he suggests the

importance of the following: re-election, influence within the House,

good public policy, a career beyond the house, and private gain.67 Simi-

lar purposes can be posited for council members. However, a member's e-

valuation of the benefits inherent in such goals cannot be viewed as in-

dependent of the predispositions toward implementing the goals. To use
.31
Rocheach's term,"primitive beliefs" act to structure man's understanding

of his environment, and help guide and inform rather specific goals.68

These primitive beliefs or predispositions are highly valued in them-

selves. Thus a "good policy" goal--for example, the elimination of po-

verty--will be associated with different action alternatives for differ-

ent individuals. Conservative or liberal societal predispositions will

significantly alter the course of action taken or alternative chosen by

an individual despite fundamentally similar goals.

Neither goals nor predispositions are immune from the context of

choice, however. It is the context of choice which determines the exter-

nalities and decision costs against which goals and predispositions are

to be weighed. The goals themselves are best achieved with reference to

different arenas or contexts. For example, "re-election" and "career be-

yond the council" goals appear most likely to involve individuals outside

b1




52


the council itself, or liable to be balanced against external costs such

as voting publics, special interest groups or powerful community indivi-

duals. "Good policy goals" and "influence within the council goals," on

the other hand, are most likely to be stymied by the structure of rela-

tionships within the council itself or the costs of reaching decisions.

The benefit captured by the "private gain" goal appears to have reference

to costs generated both internal and external to the council.

In a similar fashion, a council member's predispositions are mediated

by the context of the choice situation. However, predispositions, unlike

specific goals, are likely to be quite resilient to short-term circum-

stances, and hence much more important in the determination of public po-

licy. Predispositions which imbue ends with means appear more fundamen-

tally connected to the individual member's ability to absorb costs, ei-

ther external or internal to the council. A member's predispositions con-

stitute the core of his beliefs; actions and alternatives which evoke pre-

dispositions are likely to be perceived as conferring considerable bene-

fit. In choice situations where such predispositions are involved, it

might be expected that the costs a member would be willing to absorb

would be correspondingly high.

The costs a member might be reasonably expected to encounter have

already been considered to be either external to his own control or as-

sociated with the process of reaching a collective decision. The exter-

nal costs liable to an individual member originate from many sources.

First, local legislative politics are conditioned fundamentally by the

relationship of the council member and the electorate. At the local le-

vel, the elected official is joined in what Burke termed, "the strictest

union with his constituents'."69 This union connotes more than virtually
RF.








unlimited accessibility to the member by his constituents. It means the

member is not only an official of the community but also an official in

the community. Unlike officials at many other levels of government, the

typical member works, plays, and lives in the governmental unit whose

fate he helps determine. Wrapped as he is in the social nexus of the

community's activities, the council member is exposed to an unusual num-

ber and variety of potential sanctions. These factors have considerable

bearing on a member's ability to be re-elected and/or his future career

within the community.

Second, a council member seldom earns his livelihood from his salary

as a public official. A member typically pursues his council activities

in addition to or in determinant of his regular occupation. Despite the

benefits that may accrue him because of his position, council duties re-

main ancillary to other facets.of his life. Moreover, any motivation to

do otherwise is usually discouraged. Most municipalities re-enforce the

citizen legislator role by providing few services for the council member,

by purposively keeping salaries low, and by limiting the official office

space allotted for council members. In short, a member's ability to gain

from his tenure is heavily conditioned by his relations with the commu-

nity.

Severe external costs can also be generated for an individual coun-

cil member if he runs afoul of special interest groups within the commu-

nity or powerful individuals within the community. For a council member

wishing re-election, a career beyond the council, or private gain, choices

made by these individuals can result in great personal costs. In a simi-

lar fashion, the mayor of the municipality or portions of the local bu-

reaucracy can all act to derail a member's ambitions and hopes.







Of course, a council member will have goals within the council it-

self, and many of his actions within the council will be addressed to the

community audience he perceives as most relevant for the minimization of

external costs. Nonetheless the member must also work within the coun-

cil and with the other members. His ability to minimize external costs

will, in part, be a function of his ability to determine the costs in-

curred in decision making. Even in the event where external costs them-

selves are perceived as unimportant (as in the case of some retiring mem-

bers), a member's ability to pursue preferred alternatives still must be

balanced against decision making costs.

In perhaps the most perceptive article dealing with this aspect of

collective choice, Adrian and Press have suggested there are at least

eight decision costs involved in the process of forming a coalition ca-

pable of decision. The costs incorporate "a variety of economic and psy-

chological factors" which are related most intimately to a member's goals

for his tenure in the council, but also bear on his attempts to benefit

in arenas outside of the council.70 Adrian and Press argue decisions in-

clude: 1) informations costs; 2) responsibility costs; 3) intergame costs;

4) dissonance costs; 5) costs of the division of payoffs; 6) inertia

costs; 7) time costs; 8) persuasion costs.71

Analytically, these categories cover an incredible range of situa-

tions and evoke some measure of the complexity involved in attempting to

specify the meaning of costs. As the conceptual basis for an empirical

study, however, their complexity is overwhelming. Thus, in the pages to

follow, I will develop an argument which focuses on a behavioral indica-

tor capable of subsuming the intent of this categorization. The argument

begins by considering the fashion in which costs and choices are inter-

twined.







Exchange analysis features the act of choice. But the act of choice

is not simple. For example, different types of issues and different de-

cision-making arrangements should act to produce various contexts for

choice and trigger different member responses. If a decision is seen as

more than just some final choice--as a whole set of activities which con-

tribute to some final choice--then it can be argued that the process of

collective choice may be punctuated by antimonies of various sorts. In

collective bodies, the clashes between individuals are customarily hand-

led in accordance with procedures developed over time by the council bo-

dy. These procedures may be formal or informal. In any case, the pro-

cesses by which an organizaEion reacts to and resolves political rows do

establish themselves. Analytically, Thompson and Tuden identified four

general forms of response.72 Following Simon and March, I have labelled

them 1) problem solving, 2) persuasion, 3) bargaining, and 4) politics.73

Thompson and Tuden suggest that the ways in which decisions are usu-

ally handled evolve as the result of the interaction of two factors (see

Figure 2-1). The first factor is the degree of consensus or disagreement



Preferences About Possible Outcomes:
(Substantive Goals)
AGREEMENT NONAGREEMENT

Computation in Bargaining in
Bureaucratic Representative
AGREEMENT Structures Structures

Beliefs About [Problem Solving] [Bargaining]
Causation:
(Means) Majority Judgment Inspiration in
in Collegial "Anomic"
DISAGREEMENT Structures Structures
[Persuasion] [Politics]


Figure 2-1. A Typology of Decision Organization.
Source: adapted from James Thompson and Arthur Tuden, "Strategies, Struc-
tures, and Processes of Organizational Decision," in James Thomp-
son (ed.), Comparative Studies In Administration, (Pittsburgh:
University of Pittsburgh Press), 1959.




56


among the actors' general orientations towards the means used to reach

some alternative. The second factor is the preferences of the council

members for that alternative. These governing practices emerge over time

as the body acts, time and time again, to resolve the issues that come

before it.

While these are analytic categories and not concrete processing to

be associated with any one council, the assumption is that within a coun-

cil a governing practice or practices develop which minimize the decision

costs for that group of individuals. Such practices have an effect upon

the types of alternatives chosen by the group. For example, in a situa-

tion where there is consensus among the actors over the procedural norms

used to settle a conflict but disagreement over the goals espoused by dif-

ferent actors, then for collective action to occur, bargaining should be

initiated. Yet the fact that a group establishes bargaining as its de-

cision-handling technique has ramifications for the consideration of al-

ternatives which are different from those that would obtain if the group

handled these issues consensually.

This point can be amplified by considering the argument made by

Heinz Eulau which encompasses the interrelationship between individual pre-

dilections on issues and those detected within the entire council.75 He

argues that the organization of decision or "collective articulation" will

be impacted by the specificity or diffuseness of the preferences involved

at the individual level and at the council level. By diffuseness, he

means there is no hierarchy of preference which would give priority in

decision making to one demand over another; while specificity means there

is just such a hierarchy. The interaction of individual level and council

level articulation of preference, Eulau argues, is what determines the








manner in which decision situations are organized. This view is summa-

rized in Figure 2-2.



Articulation of Member:
SPECIFIC DIFFUSE

SPECIFIC Pluralism Monism
Articulation of Group:
DIFFUSE Dictatorship Anarchy


Figure 2-2. Decision Organization and Preference Articulation.
Source: adapted from Heinz Eulau, "Logics of Rationality in Unanimous
Decision Making," in Carl Friedrich (ed.), Rational Decisions,
Nomos VII, (New York: Atherton), 1967, 26-54.

Eulau maintains that collective articulation is likely to be plural

and democratic when the interests of every individual in the group, as

well as those of the group itself, are specifically articulated.76 An-

archic situations are the likely result of situations when the articula-

tion of both the members and the group is diffuse. On the other hand,

when there is an asymmetry between group and individual articulation then

unanimous decision making is a strong possibility. This is because where

the group's articulation is specific and the individual's preference is

diffuse, any individual will incur large costs in attempting to "buck"

the group's decisions. These costs will not be offset from the benefits

he might perceive in acting against the group because the issue does not

involve preferences he highly values. Similarly, if at least some indi-

viduals have specifically articulated preferences and the group as a

whole is diffuse on an issue, then group members having poorly ordered

preferences will incur costs and receive few benefits in opposing even

a minority of their fellow council members.

Eulau's argument demonstrates the real possibility for unanimity in

any situation of collective choice. Notice that highly contested situ-







nations are likely only when both the individual members and the group

preferences are quite specific. Given the external costs and decision-

making costs council members have been assumed to incur, unanimity on

councils would theoretically seem to be the most prevalent mode of arti-

culating preference. Specificity is just too costly.

This remains true even when the context of decision handling is

considered. Both the problem-solving and persuasion modes are readily

conducive to consensual forms of behavior. Under conditions indicating

bargaining, if there exists a specificity of group preferences but a dif-

fuseness of individual preference, then unanimity is quite likely. Since

no specific benefits are to be gained, external costs can be minimized

by a diffusion of responsibility. Moreover, decision making costs, usu-

ally quite- high in a consensual situation, will be lowered by the feli-

city of each member towards bargaining. In short, under such conditions

bargained unanimity will result. If, on the other hand, within the bar-

gaining mode, group interests are diffuse while individual interests are

specific, then unanimity is likely only if there is a failure in ration-

ality. Eulau has suggested failures are possible because unanimity under

such circumstances "serves as a kind of protective screen from responsi-

bility for failure to resolve an issue in those terms in which it was

originally defined by the individual with specific interest himself."77

Thus, in empirical situations, unanimity is the likely result even when

the assumption of individual rationality is made.

All this is not to argue unanimity will characterize every council

decision. Indeed the possibility of conflict is inherent in every situ-

ation where two members have goals or predispositions which are to some

degree incompatible. Costs, however, are extremely high for activists







engaged in conflictual situations. "Expenditures of time become greater,

stressful encounters multiply, information becomes harder to obtain yet

more vital as bluffing strategies are adopted. Even successful activists

find their average net utility reduced."78 Council conflicts are likely

then only when severe costs predicate no other course of action. Coun-

cil conflict can thus signal the presence of substantial costs. And con-

flict can be generated in many ways.

The disjunctive effects of member differences can flow from several

sources. Certainly the goals and predispositions of a member are a fun-

damental and important source. Basic life experiences may well hold the

seeds of conflict. A member's age, his education or occupation will help

determine his subjective evaluation of alternatives and the costs and be-

nefits they represent. In a sense, such experiences act as surrogate in-

dicators for a wealth of behavioral and psychological experiences. They

help establish the manner in which an individual will categorize specific

issues. It can also be assumed that such experiences are important for

establishing the cognitive processes of a member enabling him to formu-

late a way of acting upon alternatives once they have been cast.79 The

learned aspect of behavioral response in the context of such experiences

should produce cognitive processes that are relatively stable through

time. Such stability helps explain why members perceive costs in certain

situations, and why they feel cues taken from certain groups and indivi-

duals will orient them correctly in a given situation.

In the exchange perspective, the conflict or disagreement that char-

acterizes a council is conceived to be the result of a complex nexus that

finds the member at the center. His background (life experience), goals,

and predispositions are fundamental to his response. However, these fac-








tors can be mediated by the manner in which his perceptions subjectively

work to assign weights to various other factors in his environment. Im-

portant groups, habitual interactions with individuals in information ga-

thering forays, the manner in which the decisional structure of the coun-

cil evolves, the importance he attaches to his job; each affects the in-

dividual member's contribution to the collective choice.

While it is both interesting and helpful to consider the general

manner in which council members and their council as a whole interact, it

is specific, alternative policies which engender conflictual or consen-

sual interactions. This is in large measure what Lowi means by "policies

determine politics."80 The political reaction to a policy choice is mea-

sured by council members not only in terms of the institutional costs it

may generate but also in terms of the external costs expected from the

attending publics. Policy content triggers political calculations based

on each member's experientiallydetermined evaluations of the costs atten-

dant to such choices. These calculations, when made by an entire council,

determine the institutional future of a policy alternative.

Salisbury and Heinz have advanced this very argument. They maintain

different types of policy--distributive, redistributive, regulative, and

self-regulative--can be expected to emerge as outputs in relation to the

mix of external and decision-making costs attendant to policy choices

within an institutional setting. They construct a typology that consi-

ders 1) the costs resulting from the pattern of demand impinging upon the

decision makers and 2) the costs incurred by policy makers in forming the

coalition necessary for reaching a decision.81 Figure 2-3 displays their

hypothesized relationships. The thrust of this typology is towards spe-

cifying the conditions under which certain types of policy will be made.




61


It suggests a given type of policy will accrue, typically, benefits and

costs for various sectors of the community; each policy has characteris-

tics that generate costs for the policy makers.



Costs of Reaching a Decision:
LOW HIGH

INTEGRATED Redistributive Self-Regulative
Demand
Pattern: FRAGMENTED Distributive Regulative



Figure 2-3. The Conditions Under Which Different Policy Types Can Be Ex-
pected To Emerge From City Councils.
Source: Robert Salisbury and John Heinz, "A Theory of Policy Analysis
and Some Preliminary Applications," in Ira Sharkansky (ed.),
Policy Analysis in Political Science, (Chicago: Markham Press),
1970.

Salisbury and Heinz argue that a fundamental distinction can be

made between two types of policy. They suggest policies are either struc-

tural or allocative.

By allocative policies we mean decisions which confer di-
rect benefits, material or symbolic, upon individuals and
groups. Structural policies we take to mean policies which
establish authority structures or rules to guide future al-
locations. The latter policies are more abstractly formu-
lated and ambiguous in their effect than the former. Allo-
cative policies may vary along a distributive-redistributive
axis; structural policies may vary as between regulatory and
self-regulatory outcomes.82

The prime difference between these two types of policies is the decision

costs they generate. The presumption is "the more costly it is to orga-

nize the requisite coalition on an issue, the more likely it is that the

policy outcomes will be structural rather than allocative."83 The point

is that in many decisional systems rising information and negotiation

costs together with declining utilities will be countered by adopting

de facto structural rules which call for actual allocations to be deter-








mined elsewhere. Allocative policies are likely to be implemented only

if decision-making costs are low. On the other hand, allocative poli-

cies may well be associated with high costs when merely broug: into the

institutional arena for discussion and deliberation.

Salisbury and Heinz also argue that varying demand patterns impinge

upon the decision makers and affect the types of policies likely to re-

sult from collective choice. Integrated demand patterns, in general,

can confer great benefits or result in great costs. A member can ill-

afford to absorb the possibility of large external costs presented in in-

tegrated demand by acting counter to the wishes of the attentive public.

In the case of fragmented demand, external costs are likely to be much

lower since the member can generate benefits in some portions of the at-

tentive public only to lose in others. Of course, the perception of in-

tegrated or fragmented demand by a member is guided by his general orien-

tation to various segments of the public. In taking his cues from cer-

tain segments of the community, a member provides weights to his percep-

tion of demand. Significant efforts will be necessary for certain groups

to appear important enough and well integrated enough to pose serious ex-

ternal costs for a member. For other groups, a member will be sharply

attuned--high external costs are easily incurred. Thus in most communi-

ties where businessmen are among the most organized and best funded groups,

members are likely to pay inordinant attention to their wishes. The types

of costs and benefits other groups can confer are, in all probability,

much lower.

Distributive, redistributive, regulative and self-regulative poli-

cies result then from the balance of internal and external costs. Yet

this relationship can be determined only if one makes assumptions about







the manner in which the outcomes these policies represent square with po-

sited preferences and costs. The policies cannot be defined merely in

terms of the costs they generate; this would be tautological. Instead

the types must be defined in such a fashion as to relate to the costs

without explicitly involving behavioral manifestations.

One way out of this dilemma is to start with the distinction drawn

by Salisbury and Heinz between allocative and structural policies. Allo-

cative policies confer direct benefits; structural policies do not. Sup-

pose then that the conventional meaning of the terms are used. That is,

distributive policies are perceived to confer direct benefits upon one

or more groups according to previous rules or customary allocations. Re-

distributive policies connote a shift in the operant allocation of values.

In regulatory policy, a rule is established by which government can me-

diate disputes over values that can be harmful in and of themselves to

the public at large. Self-regulatory policy gives this control to bodies

other than governmental entities. Both regulative and redistributive po-

licies hold the possibility of immediate coercion, while, for self-regu-

lative and distributive policies, the possibility seems much more remote.

Intuitively this seems quite agreeable, but the reason is not imme-

diately apparent. When and why is coercion necessary? Lowi comes close

to providing an answer with a quite from David Hume:

Two neighbors may agree to drain a meadow which they possess
in common: because it is easy for them to know each other's
mind; and each must perceive, that the immediate consequence
of his failing in his part, is the abandoning of the whole
project. But it is very difficult, and indeed impossible,
that a thousand persons should agree in any such action; it
being difficult for them to concert so complicated a design,
and still more difficult for them to execute it; while each
seeks a pretext to free himself of the trouble and expense,
and would lay the whole burden on others. Political society
easily remedies both these inconveniences.






64

Coercion settles the problem of different expectation and utilities. Co-

ercion is used to establish a new set of expectations. This is the case

in the distinction drawn between distributive and redistributive policies.

Only a change in allocations characterizes them. In a sense, regulation

differs from self-regulation in the same fashion. Regulation is one of

the ways governments control societal and individual conduct. It repre-

sents a change from self-regulation wherein the individual controls con-

duct and helps establish societal rules and norms.

Thus the key to understanding the types of policy pursued and made

within city council is in positing the values that are supplanting them.

One such conception of local political values has been offered by Williams

and Adrian in their study of comparative policy making, Four Cities.

They argue that local governments act as though they pursue at least four

values: economic growth, providing or securing life's amenities, main-

taining traditional services, and arbitrating among conflicting interests.85

Traditionally the role of municipalities has been defined in terms of en-

suring economic growth and the maintenance of services.86 Increasingly,

though, the proper role of municipalities has come to encompass the newer

values of providing and securing amenities for its citizens and arbitra-

ting among conflicting interests. For municipalities, these changes re-

present attempts at redistribution and increasing regulations. The self-

regulation represented in older economic growth philosophies has been

challenged by new values. Traditional services are being supplemented

and augmented by new concerns for the well being of citizens.

These notions can be used to animate the original typology offered

by Salisbury and Heinz. Once incorporated into their scheme, new hypo-

theses may be formed as displayed in Figure 2-4.









Costs of Reaching a Decision:
LOW HIGH

INTEGRATED Amenities Promoting Growth
Demand
Pattern: FRAGMENTED Traditional Arbitrating A-
Services among Conflicting
interests



Figure 2-4. Modified Version of Salisbury and Heinz Typology.

The redistributive nature of many amenities will produce a situation

within councils such that policies favoring amenities will only be pro-

duced when decision-making costs are low relative to costs inherent in an

integrated demand pattern. In general, traditional services will pass

through councils with considerable ease since decision-making costs are

likely to be low and the demand pattern fragmented. On the other hand,

regulation seems to capture the essence of arbitrating between conflict-

ing interests. This type of value typically follows structural trends.

"Emphasis is placed upon the process rather than the substance of govern-

mental action."87 While not all arbitration involves regulation, most

regulation will consist of arbitration. In arbitration, the role of go-

vernment is to provide a "neutral" arena in which claims may be resolved.

Such an arena is needed when conflicting demands impinge upon the deci-

sion makers and the internal costs of making a decision become great.

Williams and Adrians' conception of the values promoting community growth

fits into the sense of self-regulation mainly because of the philosophy

which has stood behind such values. Namely, promoting growth has tradi-

tionally manifested itself as the set of municipal policies favoring busi-

ness. When municipal policy did intervene with the economic and popula-

tion growth of the city, it was to provide incentives. In most matters,

business interests were allowed to regulate their own growth. The assump-








tion was that the city might only prosper if its government allowed bu-

siness considerable latitude in determining the conditions in which it

would function.

Even after the relationship between community values and policy

types has been posited, there remain difficulties with the scheme pre-

sented by Salisbury and Heinz. Chief among these is the indeterminacy

associated with member costs. Although various costs and benefits have

been assumed to obtain, the salience of a specific issue for various

costs and benefits is likely to differ from member to member. Policy

content may be subjectively different for different members. In this

case, costs may be a function not only of the subjective unacceptability

of alternatives, but also of the uncertainty and incomparability associ-

ated with the alternatives.88 In other words, conflict among council

members may not be an accurate reflection of the clash of goals and pre-

dispositions.

Within the context of this study, this difficulty will be resolved

by assuming that members are similar in the uncertianty and incomparabil-

ity they assign to the alternatives that flow from the consideration of

an issue. Decision costs will be viewed in terms of the average amount

of opinion disagreement that prevails within the council. As developed

in Chapter eight, council disagreements will be taken as surrogates for

high subjective evaluations of costs. For all its convenience, this re-

presents nothing more than a restatement of the Axelrod hypothesis: the

greater the incomparability of the goals, the greater the conflictual be-

havior.89

By making disagreement the inverse of policy consensus, this concep-

tion fits squarely into the exchange perspective and encompasses the po-







licy focus. It is the case, however, that in the dynamic assumed to ex-

ist among normative, structural and behavioral factors, policy is a con-

venient, if arbitrary, focus. An issue gains salience and attracts some

patterns of support or opposition outside the council as well as within.

Issues or policies can be characterized by more than their substantive

content. It is important to recognize that policies can be defined in

different ways and their redefinition can affect the costs associated

with a member's stand on various alternatives.

One persuasive statement of the manner in which policies can be re-

defined has been made by Cobb and Elder. They maintain that issues can

be expanded to involve more individuals and groups within the community

to the degree that certain strategies are pursued. These strategies in-

clude attempts to define the matter unambiguousy, to mark its social sig-

nificance, to establish its immediate relevance, to avoid technical dis-

cussion, and to demonstrate its dissimilarity with previous policies.90

Such expansion increases the likelihood that costs will be associated

with positions taken on such issues. Clearly, some issues will become

more risky for council members than others despite a common content.

Conflicts that manage to erupt outside the council, that gain community

wide recognition should be more difficult to deal with.

An issue then has normative aspects which serve to coalesce struc-

tural and behavioral elements in certain fashions. For example, expan-

sion of an issue, means bringing it to different publics, thus activating

various cleavages within the political system. This in turn may even

change the types of issues that are considered within the council. This

facet of the consideration of a policy has serious ramifications for the

empirical investigation of the subject-. It means there might exist no







key by which to identify policies for examination. However, in the ex-

change perspective, even such changes must be considered in terms of the

patterns of habitual response taken by members. Their past behaviors

should give a cue to their future responses.91 Thus I will assume poli-

cies can be identified according to their substantive content. Their

classification captures generalized sorts of costs and benefits which

remain rather constant for the empirical situation.

Summary

To summarize, the exchange perspective offered here begins with

the conception that individual choice is fundamental to a policy focus.

Choice, whatever its form, is made within a calculus of costs and bene-

fits whose parameters include a council member's goals and predisposi-

tions as well as the choices made by other council members. Choice may

engender conflict or promote consensus. In either case, the type of po-

licy that emerges from a city council represents a reflection of the

costs involved in collective choice; it turns on the characteristics of

the individuals charged with that choice and the manner in which choice

situations are typically handled. In succeeding chapters, this study

will attempt to demonstrate the importance of such concerns and draw

more concretely many of the elements of policy making in city councils.















Notes

1 David Harvey, Social Justice and the City (Baltimore: Johns Hopkins,
1973), P. 22.

2 Roland Liebert, "Municipal Functions, Structures, and Expenditures:
A Reanalysis of Recent Research," Social Science Quarterly, Vol. LIV (March,
1974), pp. 765-783.

3 Ira Sharkansky, "The Political Scientist and Policy Analysis,"
policy Analysis in Political Science, ed. Ira Sharkansky (Chicago: Markham,
1970), P. i.

4 Charles 0. Jones, "State and Local Public Policy Analysis: A Review
of Progress," prepared for delivery at Conference on Political Science and
State and Local Government, American Political Science Association (Biloxi,
Miss.: October 26-29, 1972), p. 12.

5 V.O. Key, Southern Politics (New York: Knopf, 1949), p. 308.

6 Ibid., p. 664.

7 Ibid., p. 674.

8 Stuart Rakoff and Guenther Schaeffer, "Politics, Policy, and Politi-
cal Science: Theoretical Alternatives," Politics and Society, Vol. I (November,
1970), p. 53.

9 Richard Dawson and James Robinson, "Inter-Party Competition, Economic
Variables and Welfare Policies in the American States," The Journal of Politics,
Vol. XXV (November, 1963), p. 265.

10 Rakoff, op. cit., p. 55.

11 Thomas Dye, Politics, Economics and the Public (Chicago: Rand McNally,
1966), p. 299.

12 John Fenton and Donald Chamberlayne, "The Literature Dealing with
the Relationships Between Political Processes, Socioeconomic Conditions and
Public Policies in the American States: A Bibliographical Essay," Polity,
Vol. I (Spring, 1969), pp. 388-404.

13 Herbert Jacob and Michael Lipsky, "Outputs, Structure, and Power:
An Assessment of Changes in the Study of State and Local Politics," The Journal
of Politics, Vol. XXX (May, 1968), pp. 510-538.

14 Harold Sprout and Margaret Sprout, "Environmental Factors in the
Study of International Politics," The Shaping of Foreign Policy, eds. Harold









acobson and William Zimmerman (New York: Atherton, 1969), pp. 43-75. Harold
and Margaret Sprout have argued that the relationship between man and milieu
has been discussed by scholars in five, more or less, distinct ways: (1) en-
vironmental determinism, (2) free-will environmentalism, (3) environmental pos-
sibilism, (4) cognitive behaviorism, and (5) environmental probabilism. En-
vironmental determinism posits an "invariable correlation" between some set
of environmental "causes" and their environing "effects," (see p. 48). Free-
will environmentalism allows an individual more of a choice, but clearly has
environmental influences playing a direct role in his actions. Possibilism
avoids the notion of choice and postulates a set of constraining factors that
will affect the outcome of any course of action. In the possibilist world,
decisions and motivations are the givens of a situation not to be explained.
The cognitive behavior form of hypotheses emphasizes the conception that an
individual reacts to his environment as he apprehends it: it explicitly in-
corporates his past experiences into his future behavior. The distinction is
also drawn here between operational and psychological environments. Finally,
environmental probabilism names a relationship in which choices may be explained
with reference to a hypothetical norm. That is, through an examination of past
behavior and the norm established by it, inferences about future decisions
can be made.

15 One excellent review of this literature is found in Charles 0. Jones,
op. cit. See also Fenton and Chamberlayne, op. cit.; Edmund P. Fowler and
Robert L. Lineberry, "Comparative Policy Analysis and the Problem of Recipro-
cal Causation," Comparative Public Policy: Issues, Theories, and Methods, eds.
Craig Liske, William Loehr and John McCamant (New York: Halsted, 1975), pp.
243-259; and Terry Clark, "Community Structure, Decision-Making, Budget Ex-
penditures, and Urban Renewal in 51 American Communities," American Sociologi-
cal Review, Vol. XXXIII (August, 1968), pp. 576-592.

16 Dye, op. cit., p. 298.

17 Vernon Van Dyke, Political Science: A Philosophical Analysis (Stan-
ford, Calif.: Stanford University, 1960), pp. 131-157.

18 A discussion of these approaches can be found in John Dearlove,
The Politics of Policy in Local Government: The Making and Maintenance of
Public Policy in the Royal Boroughs of Kensington and Chelsea (London:
Cambridge University, 1973). Also, see John Kirlin and Steven Erie, "The
Study of City Governance and Public Policy Making: A Critical Appraisal,"
Warner Modular Publications (Andover, Mass.: Warner Modular Publications,
1973), pp. 1-12.

19 Edward Banfield and James Q. Wilson, City Politics (Cambridge, Mass.:
Harvard University and M.I.T., 1963).

20 Kirlin, op. cit., p. 4.

21 Ibid.

22 William Riker and Peter Ordeshook, An Introduction to Positive
Political Theory (Englewood Cliffs, N.J.: Prentice-Hall, 1973), p. 2.

23 See, for example, Raymond Bauer and Kenneth Gergen, The Study of
policy Formation (New York: Free Press, 1968).






71


24 David E. Apter, Choice and the Politics of Allocation: A Develop-
ntal Theory (New Haven, Conn.: Yale University, 1971), p. 9.

25 David E. Apter, The Politics of Modernization (Chicago: University
of Chicago, 1965), p. 39.

26 Heinz Eulau and Kenneth Prewitt, Labyrinths of Democracy: Adapta-
tions, Linkages, Representation, and Policies in Urban Politics (Indianapolis:
Bobbs-Merrill, 1973), p. 474.

27 Heinz Eulau, "Policy Making in American Cities: Comparisons in a
Quasi-Longitudinal, Quasi-Experimental Design," General Learning Press Reprint
(New York: General Learning, 1971), p. 2.

28 James March and Herbert Simon, Organizations (New York: Wiley, 1958),
p. 27.

29 John Crecine, Governmental Problem-Solving: A Computer Simulation
of Municipal Budgeting (Chicago: Rand McNally, 1969), p. 99.

30 Ibid., p. 101.

31 Ibid., p. 144.

32 Peter Rossi, Richard Berk and Bettye Edison, The Roots of Urban Dis-
content: Public Policy, Municipal Institutions, and the Ghetto (New York:
Wiley, 1974), p. 34.

33 Dearlove, op. cit., p. 231.

34 Jonathan Turner, The Structure of Sociological Theory (Homewood, Ill.:
Dorsey, 1974), p. 161.

35 Michael D. Cohen, James G. March and John Olsen, "A Garbage Can
Model of Organizational Choice," Administrative Science Quarterly, Vol. XVII
(March, 1972), pp. 1-25.

36 Ibid., p. 1.

37 Ibid., p. 2.

38 See James S. Coleman, Community Conflict (New York: Free Press, 1957),
and William A. Gamson, "Rancorous Conflict in Community Politics," American
Sociological Review, Vol. XXXI (February, 1966), pp. 71-80.

39 See Bryan T. Downes, "Issue Conflict, Factionalism, and Consensus in
Suburban City Councils," Urban Affairs Quarterly, Vol. IV (June, 1969), pp.
477-497. See also by the same author "Suburban Differentiation and Municipal
Policy Choices: A Comparative Analysis of Suburban Political Systems," Com-
munity Structure and Decision-Making Comparative Analyses, ed. Terry Clark
(San Francisco: Chandler, 1968), pp. 243-268.

40 Downes, "Issue Conflict, Factionalism, and Consensus in Suburban
City Councils," op. cit., p. 484.









41 The work coming out of the City Council Research Project is quite
varied. The most important pieces published thus far include the following:
Heinz Eulau and Kenneth Prewitt, op. cit.; Betty Zisk, Local Interest Politics:
A ne Way Street (Indianapolis: Bobbs-Merrill, 1968); Robert Eyestone, The
Treds of Public Policy: A Study in Policy Leadership (Indianapolis: Bobbs-
Merrill, 1971); Ronald Loveridge, City Managers in Legislative Politics (Indian-
apolis: Bobbs-Merrill, 1971); Heinz Eulau and Robert Eyestone, "Policy Maps
of City Councils and Policy Outcomes: A Developmental Analysis," APSR, Vol.
LXII (March, 1968), pp. 124-143; and Kenneth Prewitt, "Political Ambition,
Volunteerism and Electoral Accountability," APSR, Vol. LXIV (March, 1970), pp.
5-17.

42 Eulau and Prewitt, op. cit., p. 609.

43 Ibid., pp. 611-612.

44 Ibid., p. 612.

45 John C. Wahlke, Heinz Eulau, William Buchanan and Leroy Ferguson,
The Legislative System: Explanations in Legislative Behavior (New York:
Wiley, 1962).

46 Eulau and Prewitt, op. cit.

47 See Eyestone, op. cit., or Zisk, op. cit.

48 Edward Banfield, Political Influence (New York: Free Press, 1961).

49 Exchange theory is, of course, merely a broad rubric covering many
conceptual variants. Exchange theory flows from the traditional utilitarian
assumptions and concepts which viewed men as rationally seeking to maximize
their material benefits, or "utility," from transactions or exchanges with
others in a free and competitive marketplace. Gradually however, such ideas
have been supplemented and augmented to reflect an awareness of the constraints
of rationality, the plethora of individual values, and the social dynamic of
interpersonal interaction. While it is always risky to generalize about
sociological theory, it is probably safe to identify the following themes as
common to most contemporary exchange formulations:
(1) While men do not seek to maximize profits, they always
seek to make some profits in their social transactions with
others.
(2) While men are not perfectly rational, they engage in cal-
culations of costs and benefits in social transactions.
(3) While men do not have perfect information on all available
alternatives, they are usually aware of at least some alterna-
tives which form the basis for assessments of costs and benefits.
(4) While there are always some constraints on human activity,
men compete with each other in seeking to make a profit in their
transactions.
(5) While economic transactions in a clearly defined market-
place occur in all societies, they are only a special case of
more general exchange relations occurring among individuals
in virtually all social contexts
(6) While material goals typify exchanges in an economic








marketplace men also exchange other, nonmaterial commodities,
such as sentiments and services of various kinds. (See Jonathan
Turner, op. cit., pp. 212-213.

50 James Buchanan and Gordon Tullock, The Calculus of Consent: Logical
Foundations of Constitutional Democracy (Ann Arbor: University of Michigan, 1962)
p. 18.

51 Turner, op. cit., pp. 211-294.

52 Ibid., p. 264

53 Riker and Ordeshook, op. cit., p. 8.

54 Ibid., p. 12.

55 Ibid., p. 14.

56 Anthony Heath, "Review Article: Exchange Theory," British Journal of
Political Science, Vol. I (January, 1971), pp. 91-119.

57 Sidney R. Waldman, Foundations of Political Action: An Exchange Theory
of Politics (Boston: Little, Brown, 1972), p. 215.

58 Riker and Ordeshook, op. cit., p. 21.

59 Waldman, op. cit., p. 217.

60 Norman Frohlich, Joe Oppenheimer and Oran Young, Political Leadership
and Collective Goods (Princeton: Princeton University, 1971), p. 124.

61 Riker and Ordeshook, op. cit., p. 3.

62 Frohlich, Oppenheimer and Young, op. cit., p. 127.

63 Riker and Ordeshook, op. cit., p. 256.

64 Buchanan and Tullock, op. cit., p. 45.

65 Riker and Ordeshook, op. cit., pp. 272-330.

66 Robert Putnam, The Beliefs of Politicians: Ideology, Conflict, and
Democracy in Britain and Italy (New Haven, Conn.: Yale University, 1973), pp. 1-7.

67 Richard Fenno, Jr., Congressmen in Committees (Boston: Little, Brown,
1973), p. 1.

68 Putnam, op. cit., p. 5.

69 Quoted in Robert Morlan, "Life on the City Council: Realities of
Legislative Politics," Capitol, Courthouse and City Hall, ed. Robert Morlan
(4th ed.; Boston: Houghton-Mifflin, 1972), p. 212.

70 Charles R. Adrian and Charles Press, "Decision Costs in Coalition
Formation," American Political Science Review, Vol. LXII (June, 1968), p. 556.










71 Ibid., pp. 556-563.

72 James D. Thompson and Arthur R. Tuden, "Strategies, Structures, and
processes of Organizational Decision," Comparative Studies in Administration,
ed. James D. Thompson (Pittsburgh: University of Pittsburgh, 1959), pp. 195-
216.

73 March and Simon, op. cit.

74 Eulau and Prewitt, op. cit.

75 Heinz Eulau, "Logics of Rationality in Unanimous Decision-Making,"
Rational Decisions, Vol. VII, Nomos, ed. Carl Friedrich (New York: Atherton,
1967), pp. 26-54.

76 Ibid., pp. 26-30.

77 Ibid., p. 45.

78 Ian Budge, "Consensus Hypotheses and Conflict of Interest: An Attempt
at Theory Integration," British Journal of Political Science, Vol. III (January,
1973), p. 88.

79 Aage Clausen, How Congressmen Decide: A Policy Focus (New York:
St. Martin's, 1973), pp. 12-35.

80 Theodore Lowi, "Four Systems of Policy, Politics and Choice," pub-
lished for participating universities by tne Inter-University Case Program,
Syracuse, New York: November, 1971, p. 3.

81 Robert Salisbury and John Heinz, "A Theory of Policy Analysis and
Some Preliminary Applications," Policy Analysis in Political Science, ed.
Ira Sharkansky (Chicago: Markham, 1970), pp. 39-59.

82 Ibid., p. 3.

83 Ibid.

84 Theodore Lowi, "Decision Making Versus Policy Making: Toward an
Antidote for Technology," Public Administration Review, Vol. XXV (May-June,
1970), p. 31.

85 Oliver P. Williams and Charles R. Adrian, Four Cities: A Study in
Comparative Policy Making (Philadelphia: University of Pennsylvania, 1963),
p. 23.

86 Ibid., pp. 23-32.

87 Ibid., p. 28.

88 March and Simon, op. cit., p. 117.


89 Budge, op. cit., p. 73.





75


90 Roger Cobb and Charles Elder, Participation in American Politics:
The Dynamics of Agenda-Building (Boston: Allyn and Bacon, 1972).


91 Clausen, op. cit.
















CHAPTER THREE
THE DEMOGRAPHIC APPROACH TO FLORIDA MUNICIPAL POLICY

During the last decade, the demographic approach has dominated re-

search on public policy by political scientists.1 Most often concerned

with financial resources, this approach uses a simplified systems model

adapted from the early work of David Easton to explain variations in po-

licy priorities among different governmental units. Its components are

the now familiar concepts of Easton's political theory: system, input,

output and feedback. But within the demographic approach, these concepts

have assumed new forms. Inputs, which in Easton's work summarize a wealth

of activities having their origin in the performance of the system, are

typically conceived in more primitive fashion. They become the socio-

economic conditions which frame the context of a particular political

system. Outputs, on the other hand, are usually measured by the finan-

cial expenditures of the governmental unit under consideration. And the

political system itself is usually rendered in terms of its structural

characteristics, such as the degree of party government or the type of

government. As a rule, most authors writing within the demographic per-

spective have ignored feedback processes, and, rather, have assumed that

outputs are registered implicitly in a political system's socioeconomic

conditions. The interworkings of this approach, as depicted in the work

of Thomas Dye, are straight-forward. A set of forces in the socioecono-

mic environment interacts with the political system to produce public po-

licy outcomes. Accordingly, the determinants of public policy are to be

found in the nexus of environment and political system.

76








As a heuristic device, such a simple scheme has powerful appeal.

It seems applicable to a host of problems and to any number of situations.

For the researcher, however, this very generality is the source of many

difficulties. The encompassing conceptual nature of the environment

has lead many authors to comb census documents, extracting as many indi-

cators of socioeconomic conditions as appears reasonable. Scores of in-

dicators have been selected and collected on comparable units of analysis,

usually American cities or states. And for each of the units of analysis,

measures of policy output such as levels of municipal expenditure or per

capital expenditures have been compiled. In an attempt to capture some-

thing of the political system in each of the units of analysis, researchers

have gathered that information about political units which is most gener-

ally available. At the local level, this has included data on governmen-

tal type (whether city manager, mayor or commission form), electoral sys-

tem (whether partisan or non-partisan), election districts (whether ward

or at-large), and the intensity of competition (turnover or party compe-

tition). In short, the approach demands a substantial data gathering ef-

fort with attention to a broad range of variables. The connection be-

tween variables is specified loosely.

Following the lead of fiscal economists who have long attempted to

explain the differences in the level of spending among American cities

and states, most political scientists working within the demographic ap-

proach have resorted to statistical techniques for assistance in sorting

out the influences on expenditures.

The basic methodology is to apply statistical techniques
of correlation.and regression to these sets of variables.
The policy output measures, the dependent variables, are
"explained" to the extent that they are systematically re-
lated to political or environmental system variables. The







explanation is a statistical one--it measures the percen-
tage of inter-case variation in the dependent variable that
is predicted by variation in the independent variable. 0-
ther statistical manipulations can combine or control for
the effect of other variables.2

Early work using this approach, pioneered in political science by

Thomas Dye, appeared to indicate economic and social factors in the en-

vironment were more influential in shaping governmental policies than

were any of the political variables. Indeed, Dye argued most of the sta-

tistical relationships found between political variables and policy out-

comes were a result of the pervasive influence of economic development on

both the political system and the policies chosen in that system.3 His

research indicated that when political factors were controlled for, eco-

nomic development continued to have a significant impact upon public po-

licy. But when the effects of economic development were statistically

controlled, political factors diminished in their importance for public

policy.

While there have been extensive criticisms of Dye's work and those

who followed his lead, the contribution of the demographic approach should

not be blurred by admissions of its difficulties.4 A number of generali-

zations about the interrelation of environment, politics and policy can be

drawn from the work founded on this conceptual scheme. In almost every

instance, research has demonstrated large and dense populations correlate

well with general city expenditures and with spending for specific muni-

cipal services. Large minority populations are associated with similar

expenditure patterns. But it is population size which has the most sys-

tematic, independent effect on expenditures. As the size of a city in-

creases, so does its expenditures. And this holds not only for total ex-

penditures, but also for expenditures per capital 5








Authors such as Brett Hawkins have been persuaded such findings can

be interpreted to bear upon the consideration of policy. He suggests

that large, dense and heterogeneous socioeconomic environments generate

demands for services and that city governments often respond favorably to

such demands.6 While this explanation would appear to neglect the manner

in which socioeconomic factors are translated into action demands by a

population, it does frame the larger concern standing behind much of the

demographic analysis. Namely, which is more important in the determina-

tion of public policy, socioeconomic or political factors? Hawkins ar-

gues "that the impact of socioeconomic heterogeneity [on policy] is not

materially affected by any intervening variable."7 In other words, the

political system, its structure and processes have little impact on the

pattern or level of a municipality's expenditures.

Not even the most ardent socioeconomic determinist denies the impor-

tance of intervening political system variables altogether. For example,

Hawkins writes:

a high level of socioeconomic heterogeneity encourages the
retention of politicized, responsive, group-accomodating in-
stitutions and influences other system properties (e.g., a
greater sensitivity to problems on the part of official po-
licy makers that in turn promotes a higher level of expen-
ditures and affects the distribution of policy benefits).8

Moreover, not all studies fashioned from the demographic perspective

have rejected the importance of political factors on policy. Political

factors have been found important in the adaptation of floridation pro-

posals, in the responsiveness by cities to the demands of citizens, and

even, on occasion, in spending priorities.9 In short, the demographic

perspective has not single-mindedly supported the impact of socioeconomic

factors.








It remains to be demonstrated whether Florida cities spend their

revenues in a way which highlights their political institutions and prac-

tices or whether socioeconomic factors will prove more important. In

either event, a systematic investigation of the perspective will allow

for a determination of the significant factors bearing on public policies

in Florida cities. It will facilitate an exposition of the worth of the

demographic approach. Moreover, the results will condition further exa-

mination of policy as it is promulgated by municipal councils. These re-

sults can be incorporated into a research design in which council influ-

ence on policy can be studied. In the next few pages, I will specify the

units variables and linkages portrayed in Figure 3-1.




[ENVIRONMENT]
SES Factors



[POLITICAL SYSTEM] [POLICY]
Structural Aspects of Expenditures
City Government





Figure 3-1. The Demographic Perspective: Macro-phase Indicators


The Units of Analysis

The demographic model has had wide currency in the study of municipal

expenditures, but its use has engendered an extended debate. Much criti-

cism surrounding its application has turned on methodological difficul-

ties. For example, it has been argued the reliance on correlations rather
JI
than regression coefficients, the severe problems of multicollinearity,

and the cross-sectional nature of most studies, all act to weaken its






findings.10 However, many of the problems most frequently cited can be

addressed by a careful choice of the units of analysis.

In choosing only Florida cities as the site of my investigation,

many of the confounding factors present in any comparative research pro-

ject are minimized. The municipalities are subject to a common body of

state law,.and state requirements are relatively similar for all cities

throughout the state. Moreover, to the degree urban environments vary

greatly from state to state as the literature on political culture and

state politics indicates, inter-state variations will be controlled.11

There are additional advantages to conducting the research within

a single state. In the demographic model, policy outputs are measured

in terms of fiscal indicators. Research involving expenditures in the

municipalities of a single state is unlikely to suffer from the same

weaknesses as comparative studies involving cities in different states or

regions. Costs should be similar across the sample of cities. Compar-

able services will require similar expenditures because the costs of li-

ving will be relatively uniform. And there is yet another advantage.

Within Florida, municipal expenditures are standardized. In inter-state

comparisons, expenditures are derived from aggregate figures compiled by

different agencies using different procedures. Bookkeeping decisions

vary; transfers are handled differently. This is not true in Florida

where the Commission on Local Government, created by the Florida legis-

lature in 1972, developed a questionnaire which formed the basis of Flo-

rida's Fiscal Reporting, and Uniform Accounting and Budgeting System.12

All the fiscal measures used in this study have been gathered and re-

ported by municipalities on the basis of a common questionnaire.


qIIPYIP 1 111 _1 L Y~~r$ll




Full Text

PAGE 1

CITY COUNCILS AND PUBLIC POLICY By Robert Bernard Bradley A Dissertation Presented to the Graduate Council of The University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy UNIVERSITY OF FLORIDA 1977

PAGE 2

To the Memory of My Parents

PAGE 3

ACKNOWLEDGEMENTS I began this project almost four years ago. During that time I have benefited immensely from the insight and counsel of Dr. Alfred B. Clubok. He is an extraordinary teacher and a valuable friend. He has been patient and sensitive, critical and constructive. And in my association with him I have learned more than I might easily say. For what he lias given, for what he is, I would like to extend my sincere appreciation. I would also like to express my appreciation to Dr. Bert E. Swanson who has patiently read through several versions of this dissertation and whose comments improved the entire effort markedly. This dissertation also benefited from the advice and commentary of Dr. Keith Legg, Dr. Eugene Wittkopf and Dr. Robert Ziller. Dr. Legg, in particular, contributed more than he might know since it was he along with Dr. James F. Morrison who first convinced me that political science rather than physics might provide fascinating avenues of intellectual development. In a work such as this numerous other individuals provided comfort and assistance along the way. In particular I would like to thank my good friends Dr. Clem Bezold and Dr. Paul Peretz with whom I discussed aspects of my argument and without whose counsel this work would be much the worse. Throughout the entire time I was preoccupied with this exercise, I have had the recourse to the advice of Dr. Carolyn Herrington. She has consistently urged me to clarity though I be muddled, pushed me to precision when I was vague, and counseled me to simplicity when I was roundabout. That I have often failed to this good counsel is a blame I regretfully accept. 1X1

PAGE 4

TABLE OF CONTENTS ACKNOWLEDGEMENTS iii LIST OF TABLES vii LIST OF FIGURES xii ABSTRACT ™ CHAPTER ONE INTRODUCTION 1 The Problem and Its Importance 1 The Scope and Limits of the Research The Argument of the Dissertation 10 Notes 24 TWO THE POLICY FOCUS, CITY COUNCILS AND EXCHANGE THEORY . 26 Policy Models and the Policy Outcomes 27 Approaches to the Study of City Councils .... 34 An Exchange Perspective on Policy Making in City Councils , ^ 1 Summary Notes 69 THREE THE DEMOGRAPHIC APPROACH TO FLORIDA MUNICIPAL POLICY 76 The Units of Analysis 8° The Dependent Variable 83 The Independent Variables 87 The Intervening Variables 90 Specifying the Linkages Methods and Analysis "' A Commentary on the Findings 103 Notes 107 FOUR CLUSTERING FLORIDA CITIES .... 112 Considering the Operational Environment 112 The Establishment of Environment 116 The Reliability of the Factor Solution 119 City Classification: Socioeconomic Environment . . 133 City Classification: Expenditure Types 142 Summary ± -' u Notes 157 IV

PAGE 5

CHAPTER FIVE SIX SEVEN EIGHT NINE Page FLORIDA COUNCIL MEMBERS AND THEIR COUNCILS .... 161 The Typical Florida Council Member 162 Florida Council Members: Characteristics and Perceptions 167 The Characteristics of Florida City Councils . . . 179 The City Councils of SES Group F 195 Summary nnr Notes 209 PERSONAL ATTRIBUTES, CONTEXTUAL FACTORS AND COUNCIL RELATIONS 211 Members and the Form of Their Cognitive Structures . 212 Filling in the Members' Cognitive Structure .... 223 Cognitive Structures and Council Relations .... 231 The Council Context 240 Council Relations in SES Group F 249 Summary 261 Notes • • • ...'.'.'.'.'. 264 COUNCIL CONFLICT AND DECISION-MAKING MODES .... 266 Council Membership and the Prospect of Conflict . . 267 Exchange Theory and Council Conflict 275 Council Conflicts 279 Decision-Making Modes 288 The Correlates of Decision-Making Modes 296 Decision-Making Modes in Group F 301 Decision-Making Modes and Conflict 309 Summary o-,o Notes '. '.'.'.'.'.'.'. 320 DECISION-MAKING MODES AND POLICY 323 The Importance of Issues 325 Conflict, Issues, and Decision Modes 335 Issue Conflict Clusters and Decision-Making Modes . . 347 Summary o/-o Notes " ' '. '. '. '. '.'..'.'.'. 364 POLICY ISSUE CONFLICT AND POLICY TYPES 366 Conflict Over Issues 355 Revealed Interest Versus Posited Interests .... 378 An Examination of Council Policy 390 An Examination of Policy Types 405 Summary ,-, r Notes 419

PAGE 6

CHAPTER Page TEN POLITICS AND POLICY 422 Policy Types and Expenditure Groups in the High SES ^ • ^ ... 424 Cities Political Features of the High SES Cities and Their v. i • ... 435 Policies Summary x T ^ .... 444 Notes ELEVEN SALVAGING THE POLITICAL SYSTEM 445 The City Council in Policy Making 445 The Implications for Policy Explanation .... 455 Policy Research and the Policy Problem 458 ., . 465 Notes APPENDIX A THE RESEARCH INSTRUMENT AND SCALE VARIABLES ... 467 Representativeness of the Sample 476 The Scale Variables ^ The Conservatism-Liberalism Scales ^°^ Group Influence Scales Policy Disagreement Scales Policy Agreement Scales Scales of Various Council Activities 4yz Scales of Council Evaluation of Council Performance 493 Information Cue Scales .... Council Relations Scales J: Scales of Problem Perceptions ^8 Notes B ANALYSIS OF FINANCIAL DATA FOR FY 1972 FOR CITIES . 502 . . • • 508 Notes '•' C DEMOGRAPHIC VARIABLES USED IN STUDY 509 513 Notes D VARIMAX ROTATION OF SEVENTY-THREE SOCIOECONOMIC VARIABLES 5 528 Notes E CLUSTERING ISSUE DISAGREEMENTS . . 529 . . • • 535 Notes F STATISTICS Box's M Wilks Lambda ... 5 Jo Notes 539 LIST OF REFERENCES BIOGRAPHICAL SKETCH VI

PAGE 7

LIST OF TABLES Table Page 3-1 Alphabetical List of Cities in Study 84 3-2 Variables Contained in the Fiscal Reporting and Uniform Accounting and Budgeting System for the State of Florida 86 3-3 Classification of Variables Employed in Factorial Ecology 89 3-4 Variables Relating to City Political Structure ... 9 2 4-1 Factor Analysis of Expenditure Variables: Varimax Rotated Factor Matrix 121 4-2 Factor Analysis of Percent of Total Expenditures by Functional Area Variables: Varimax Rotated Factor Matrix 123 4-3 Beta Weights of Regression of SES (Factor Scores) on Expenditures and Percentage of Expenditures by Functional Area (Factor Scores) with All SES Variables Forced into the Equation 129 4-4 Relative Depression of Multiple Correlation Coefficients 132 4-5 Socioeconomic Groupings Used Throughout Study: Groupings of Florida Cities Over 10,000 Population Based on Discriminant Analysis of 73 SES Variables Reduced by Varimax Factor Rotation into 12 Scores Per City (Primarily 1970 Data) 135 4-6 Contiguity Analysis of Cities in the Socioeconomic Groups 137 4-7 Summary Listing of Final Assignments by Discriminant Analyses of Cities on Total and Relative Proportions of Expenditures in Each Functional Category .... 144 4-8 Grouping of Florida Cities Based on Discriminant Analysis of Expenditure Variables 147 VII

PAGE 8

Table 4-9 4-10 5-1 5-2 5-3 6-1 6-2 6-3 6-4 6-5 6-6 6-7 6-8 6-9 6-10 6-11 Page Grouping of Florida Cities Based on Discriminant Inalysis of Percent of Total Expenditure Varxables^ ^ by Functional Area Discriminant Analysis of Percentage of Total Expenditures by Functional Category Within Selected ^ SES Groups Frequency Distributions of Member Response on ^ Scale Variables Analysis of Variance On Scale Variables For Coun^ cil Means of SES Groups Group Council on Council Relations Scales for SES Groups: Normalized to Range 1-10 Kendall's Tau-c Between Conservatism Scales and ^ Member Life Experiences The Relationship Between Measures of Conserva^ tism and Occupation Kendall's Tau-c Between Measures of Personal ^ Goals and Life Experiences . • Kendall's Tau-c Between Measures of Conservatism ^ and Members' Goals Council Relations Scales Used to Tap Each Element of Eulau and Prewitt's Concept of "Triple Bonding . Kendall's Tau-c Between Selected Measures and Council Characteristics for All Individual Council ^ ^ Members Spearman's Rho Between Measures of Council Charac-^ ^ teristics Kendall's Tau-c Between Selected Measures and Council ^ Characteristics for All Councxls Kendall's Tau-c Between Selected Measures and Council Characteristics For Councils of Group F . . . • Kendall's Tau-c Between Measures of Council Characteristics for High SES Cities of Group F .... Kendall's Tau-c Between Measures of /"dispositions and Differences in Predispositions for Councxls of High SES^ Cities in Group F viii

PAGE 9

Table Page 7-1 Typology of Legislative Forms 270 7-2 Kendall's Tau-c Between Measure of Council Performance and Council Characteristics (Relations) 278 7-3 Kendall's Tau-c Between Selected Measures of Council Life and Total Council Conflict for All Councils . . 282 7-4 Number of Councils Above and Below the Sample Mean on Policy Disagreement (MODES) Scales 291 7-5 Number of Councils Above Group F Mean on Only One of the Policy Disagreement (MODES) Scales 291 7-6 Council Means on Decision-Making Mode Scales for SES Groups: Normalized in Range from 1 to 10 ... 7-7 Kendall's Tau-c Between Selected Measures of Council Life and Decision-Making Modes for All Councils . . 7-8 Z-Scores of Council Means on Decision-Making Modes for SES Group F Cities 7-9 Kendall's Tau-c Between Measures of Decision-Making Modes and Conflict 7-10 Generalized Disagreement and Bargaining in SES Group F 312 7-11 Amicable Agreement and Generalized Disagreement in SES Group F 7-12 Number of Decision-Making Modes Referenced Highly by Each City Council in SES Group F 8-1 Number of Council Members and Councils Citing Policy Issues as Important 8-2 Policy or Funding Area Ranked by Proportion of Members Indicating Council Disagreement 8-3 A Comparison of Policy Issues According to Their Number of Mentions and Their Propensity to Generate Council tn.341 Disagreement 8-4 Average Number of Issue Disagreements Per Member for Each Council in SES Groups O A O 8-5 Clusters of Council Issue Disagreements IX

PAGE 10

Table Page 8-9 8-6 Average Disagreement in Each Cluster Per Member Per Issue for All Members of SES Group F 8-7 Average Proportion of Disagreement Per Issue Cluster Per Member Per Issue for Group F Councils 8-8 Kendall's Tau-c Between Decision-Making Modes and Council Divisions on Clusters for Group F Cxties .... Kendall's Tau-c Between Council Division Clusters and Total Amount of Council Conflict on Issues tor Coun ^ cils of SES Group F 8-10 Kendall's Tau-c Between Member Problem Perception and Council Division Clusters for Councils of SES Croup F . 9-1 Proportion of Council Members Indicating Disagreement, Citizen Contacts and Group Actions by General Policy ^ Area .... 9-2 Phi Associations Between Council Disagreement and Types of Contacting by General Policy Area . . • • 9-3 Classification of Policy Issues into Four Types . . . 9-4 Z-Scores of Proportion of Council Members Indicating Disagreement, Citizen Contacts and Group Actions by ^ General Policy Area Difference Between Z-Scores for Proportions of Members Indicating General Policy Area and Council Divided, ^ or Attracted Group and Citizen Activity 9-5 9-6 Z-Scores of Proportion of Council Members Indicating Disagreement, Citizen Contact, and Group Action by General Policy Area and Classified Policy Types . . 9-7 Proportion of Council Members Indicating Agreement With Intent of Statement About Policies of Importance . . 10-1 Average Proportion of Members Reporting Council Disa greement, Citizen Contacts, and Group Actions by Policy .ype in Each Expenditure Group of High SES Cities (Group F) 426 10-2 SES Group F Council Members' Characterization of the Political Aspects of Traditional Issues 10-3 Vector Analysis of Cross-Tabulation of Important Issues Involving Council Disagreement and Group Conflict m ^ Cities of Group F x

PAGE 11

Table 10-4 Page 10-5 A-l A-2 A3 A-4 A-5 A6 A7 B-l D-l E-l Average Proportion of Members Reporting Council Disagreement, Citizen Contacts, and Group Actions by Policy Type for All Council Members and Those in SES Group F Average Scores of Selected Indices in Each Expenditure Group of High SES Cities (Group F) Interview Schedule Response Rate by SES Group Councils Represented in Each SES Group by Number of Council Members Responding Breakdown of All Members' Responses to Question: What was your impression of this questionnaire ? A Comparison of Sample and Population Characteristics by Sex Scales and Their Mnemonics Reliability Coefficients for Scale Variables General Information on Cities by Population Class: 19 72 Varimax Rotated Factor Matrix of Demographic Variables Varimax Rotation of Issue Disagreement Areas Evaluated by Members 437 438 468 475 476 476 477 481 485 50 3 515 534 XI

PAGE 12

Table Page 8-6 Average Disagreement in Each Cluster Per Member Per Issue for All Members of SES Group F 8-7 Average Proportion of Disagreement Per Issue Cluster Per Member Per Issue for Group F Councils 8-8 Kendall's Tau-c Between Decision-Making Modes and Council Divisions on Clusters for Group F Cxties .... Kendall's Tau-c Between Council Division Clusters and Total Amount of Council Conflict on Issues ror Coun^ cils of SES Group F 8-9 8-10 Kendall's Tau-c Between Member Problem Perception and Council Division Clusters for Councils of SES Croup F . 361 9-1 Proportion of Council Members Indicating Disagreement, Citizen Contacts and Group Actions by General Policy ^ Area . 9-2 Phi Associations Between Council Disagreement and Types of Contacting by General Policy Area . . • 9-3 Classification of Policy Issues into Four Types . . . 9-4 Z-Scores of Proportion of Council Members Indicating Disagreement, Citizen Contacts and Group Actions by ^ General Policy Area 9-5 Difference Between Z-Scores for Proportions of Members Indicating General Policy Area and Council Divided, ^ or Attracted Group and Citizen Activity 9-6 Z-Scores of Proportion of Council Members Indicating Disagreement, Citizen Contact, and Group Action by General Policy Area and Classified Policy Types . . . 9-7 Proportion of Council Members Indicating Agreement With Intent of Statement About Policies of Importance . . 10-1 Average Proportion of Members Reporting °™^ J*"*^ merit, Citizen Contacts, and Group Actions by Policy .ype in Each Expenditure Group of High SES Cities (Group F) 426 10-2 SES Group F Council Members' Characterization of the Political Aspects of Traditional Issues 10-3 Vector Analysis of Cross-Tabulation of Important Issues Involving Council Disagreement and Group Conflict m ^ Cities of Group F x

PAGE 13

Table 10-4 Page 10-5 A-l A-2 A3 A-4 A-5 A-6 A7 B-l D-l E-l Average Proportion of Members Reporting Council Disagreement, Citizen Contacts, and Group Actions by Policy Type for All Council Members and Those in SES Group F Average Scores of Selected Indices in Each Expenditure Group of High SES Cities (Group F) Interview Schedule Response Rate by SES Group Councils Represented in Each SES Group by Number of Council Members Responding Breakdown of All Members' Responses to Question: What was your impression of this questionnaire ? A Comparison of Sample and Population Characteristics by Sex Scales and Their Mnemonics Reliability Coefficients for Scale Variables General Information on Cities by Population Class: 1972 Varimax Rotated Factor Matrix of Demographic Variables Varimax Rotation of Issue Disagreement Areas Evaluated by Members 437 438 468 475 476 476 477 481 485 50 3 515 5 34 XI

PAGE 14

Figure 1-1 1-2 1-3 2-1 2-2 2-3 2-4 3-1 4-1 5-1 5-2 5-3 5-4 5-5 5-6 5-7 LIST OF FIGURES Page The Systems Approach to Policy Outcomes ...,-. 14 Overview of the Major Conceptual Categories and Relationships Investigated in the Micro-phase Overview of the Major Conceptual Categories, Analytical Variables, and Relationships Investigated in the Microphase of the Study A Typology of Decision Organization Decision Organization and Preference Articulation ... 57 The Conditions Under Which Different Policy Types Can Be Expected To Emerge From City Councils Modified Version of Salisbury and Heinz Typology ... The Demographic Perspective: Macro-phase Indicators . 80 Plot of First and Second Canonical Variables For Each City in Group: Based Upon Discriminant Function of SES Variables Mean Scores For All Council Members on Evaluation of Community Influence of Groups and Organizations 165 Mean Scores For All Council Members on Evaluation of the Seriousness of Community Problems 166 Normalized Means on Conservation Scales For Councils of SES Groups 186 Normalized Means on Group Influence Scales for Councils of SES Groups 188 Normalized Means on Problem Perception Scales for Councils of SES Groups 191 Normalized Council Means on Conservatism Scales for Councils of SES Group F 197 Normalized Council Means on Group Influence Scales f or ^ ^ SES Group F xii

PAGE 15

Figure p age 5-8 Normalized Council Means on Problem Perception Scales for SES Group F 202 5-9 Normalized Council Means on Council RElations Scales for Councils of SES Group F 204 6-1 Major Categories and Variables Treated in Chapter Six . 213 6-2 Closed Networks of Bivariate Relationships Centering on Measures of Problem Perceptions and Political Predispositions for All Members 225 6-3 Closed Networks of Bivariate Relstionships Centering on Measures of Council Relations for All Members .... 238 6-4 Closed Networks of Bivariate Relationships Centering on Measures of Council Relations for All Councils (N = 61) . 246 6-5 Closed Networks of Bivariate Relationships Centering on Measures of Council Relations for the Councils of High SES Cities (N = 12) 257 7-1 Major Categories and Variables Treated in Chapter Seven . 268 7-2 Conflict in the SES Groups 280 7-3 Closed Network of Bivariate Relationships Centering on Measures of Council Relations and Total Council Conflict. 284 7-4 z-Scores of Council Means in SES Group F on CONFLICT . . 287 7-5 Closed Networks of Bivariate Relationships Centering on Decision-Making Modes for Councils of the High SES Cities in Group F 305 7-6 Correlations of Decision-Making Modes and Conflict . . 318 8-1 Major Categories and Variables Treated in Chapter Eight . 324 8-2 Correlation of Decision-Making Modes With Average Disagreement Over Issues in Group F (Kendall's Tau-C) .... 346 9-1 Major Categories and Variables Treated in Chapter Nine . 36 7 10-1 Systems Model of Policy Process: Micro-Phase Indicators. 42 j E-l Plot of Stress Against Dimensionality in Multidimensional Scaling Analysis of Issue Conflicts (MDSCAL Program) . . 531 E-2 Plot of the Minimum Average Correlation Necessary for Inclusion in Cluster Against the Number of Clusters Produced at That Level of Correlation (CLUSTER Program) .... 531 xiii

PAGE 16

Figure E-3 Two-Dimensional Solution Scaling Issue Disagreements (Stress = .270) Page 532 1 xiv

PAGE 17

Abstract of Dissertation Presented to the Graduate Council in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy at the University of Florida CITY COUNCILS AND PUBLIC POLICY by Robert Bernard Bradley December 19 77 Chairmen: Alfred B. Clubok Major Department: Political Science City councils are breakwaters for public policy. There is a strong link between the political features of councils and their communities and the types of policies which they pursue. So argues this dissertation as it examines city councils and their handling of public policy in cities throughout Florida. In the last decade, research undertaken within the premises of the demographic model of public policy has shown socioeconomic factors to systematically account for municipal expenditures better than the structural features of municipal government. However, when 78 Florida cities with population greater than 10,000 were grouped into six clusters, each sharing relatively similar socioeconomic conditions, expenditure patterns within several clusters were found to be different. Cities displayed different expenditure profiles despite the similarity of their environments. This suggested the possibility that political forces might act in a crucial way to determine public policy. City councils, it was thought, might be crucibles where these political forces are given play. To gauge the differences in the political situation of the many councils throughout Florida, this study relied on responses of council xv '4 '4 L 'i-

PAGE 18

members in 61 cities to a mail questionnaire and personal interviews. These responses were investigated for the manner in which the personal attributes of the members, their perceptions of significant contextual factors and the characteristics of the councils to which they belong act to alter each council's decision organization. City councils, it was found, develop a repertoire of decision-making modes by which they handle different policy choice situations. Seven different modes, falling into two broad types, were investigated: 1) disagreement modes which include generalized, particularized and idiosyncratic splits, and 2) agreement modes which include bargaining, amicable agreement, consensual and informal agreement. Typically, the personal attributes of the members and especially their political and religious predispositions along with council relations will help shape the mix of modes a council adopts. But the specific modes a council adopts will evolve out of the kinds of issues considered and the historical circumstance of the council. Decision-making modes capture conceptually the orientation members evidence toward matters which come before the council. The contextual component of such orientations is great. Not every matter in a general policy area will be resolved with the same decision mode. Nor are particular modes likely to be used for any given issue across a range of councils. Nonetheless, some decision modes are more permeable to certain issues than others. And in any event, the referencing of modes appears to be a highly complex and discriminating process. In part, this complexity results from the impact of issues on each council and their meaning within that council. A simple classification scheme, such as that suggested by Lowi, appears inadequate to capture the manner in which politics and policy relate across the range of issues 1 l xvi

PAGE 19

which councils consider. But as the salience of the policies increases, as they become more important in the perceptions of the council members, policies can be expected to result in a predictable fashion from the pattern of council decision making and demands placed on the council by the community. In fact, the systematic policy differences that are observed between cities appear related to systematic differences in the political characteristics of councils and the communities which they serve. — . — — i^-^ — "X Chairman v^k XV11

PAGE 20

CHAPTER ONE INTRODUCTION In principle, municipal government stands at the very cere of the American political system. So conceived as to serve the smallest constituency in the nationwide areal division of powers, its place in the federal system has traditionally been viewed in purely functional terms. While Madison and others might have intended such an areal division as part of the cure for the mischiefs of faction, legal doctrine and the vast body of American political theory came to regard cities as entities with limited jurisdiction, and what is worse, limited capacity to solve their own problems. As such, the ability of municipal governments to exercise powers in service of their constituencies was severely circumscribed. As municipal corporations, cities often were not extended the inherent right of municipal self-governance by the courts. Municipalities existed at the pleasure of the states and frequently languished in their neglect. Gradually, over the last 20 years or so, this position has given way to what James Sundquist calls a "policy of deference." 2 Increasingly, the conscious policy of the federal government, and state governments as well, has been to defer to local judgement. Local political institutions have acquired a new importance, if not always a new potency. The Problem and Its Importance Among the structural pillars of local government stands the city council or commission. As the basic policy-making apparatus of the city, the council is nominally charged with matters of taxation, appropriation, 1

PAGE 21

and the supervision of city administration. In fact, however, there is little concerning the social, political, and physical life of a city into which its council does not or cannot inject itself. Everything from sponsoring the local beauty queen to approving the course of genetic research comes within its purview. For many of the most important issues facing any municipality, councils serve as public crucibles. In theory at least, they are used to amalgamate the demands of the citizenry, while charging the community with commitments. This description belies the place councils have found in the discussion of urban problems and politics. For most scholars, the machinations of local councils have been of little importance. The real politics of the community and the real determinants of local public policy, it has been argued, are seldom found in institutional bodies such as councils or in the results of their formal sessions. Persuasively, and at length, scholars and experienced observers have warned against the dangers of too fortnalistic a view of local politics. They look instead to the manner in which elite preferences are reflected in community choices, to the manner in which community groups and organizations help shape local policy, and to the manner in which the socioeconomic context of the city conditions its policy alternatives. However, the literature on urban processes has not always reflected this view. During the first part of this century the importance of formal structures was presupposed. Since the early 1950 's a different perspective on local government has held sway. During the early years of the century, the hopes of city government reformers turned on the notion that institutional change would result in better government; the problems with government could be resolved in tinkering with the machinery. After the Second World War, this optimistic view gave way to a period of reflection. Re-

PAGE 22

forms appeared to have little effect. To determine why this was the case, researchers re-examined their assumptions, and began anew to describe the processes of city governance. For the most part, the workings of government dominated the discussions. The concern for "good" government was left to administrators and amateur philosophers . J Still, the tradition of reform has not been entirely abandoned. In the aftermath of the events of the late sixties, a new concern for the ability of local governments to deal effectively with their problems has arisen. Whether seen in the harsh pronouncements of the Unheavenly City or in the criticism made of Banfield's iconoclasm, a "new reform' literature is upon us. 4 And the thrust of the new reform movement concerns the capacity of local government to satisfy the claims of its citizenry, especially minority and disadvantaged groups. This emphasis carries a renewed concern for the institutions of local government. This interest is tempered by the research and concerns of the last decade, however. There is the explicit recognition, amplified in the last twenty years of research, that city hall does not itself have the capacity to govern within the game as it is now played, nor is it an institution which can be treated independently from the community of which it is a part. In fact, policy initiatives may be more a function of the objective resource capabilities of the community than the desires of local decision makers. Politics itself, it has been argued, "may have litte impact on local policy. 5 But even where politics does have a role, public policy, it is realized, is made in many sites throughout the community, oftentimes outside of government. Bankers and brokers, planners and bureaucrats, public and private — all interact to allocate the values in a municipality. Municipal public policy is a complex result of many forces. There are few studies, though, which can comprehend such complexity. JHBMHMHMMMHRWHBMMHRMBMHHiMnMHHIAMMt

PAGE 23

Indeed, this study will not attempt anything so ambitious. The perspective and research advanced here are aimed at capturing that part of public policy which percolates through a city's council or commission. Clearly not all of local policy will be covered. Instead, I have elected to consider only that portion handled by local councils. Some scholars might suggest this emphasis represents the triumph of form over substance, of structure over behavior. This view, I think, is mistaken. Holistic conceptions of political systems can illuminate stubborn problems, but are difficult to document and research. In deciding to examine city councils and their role in municipal policy formulation, I have chosen to view councils as breakwaters for the policy disputes of a community; that is, as points at which the political ramifications of choice situations may become evident and highly visible. This constitutes not so much a rejection of theses on the significance of community forces in making policy, but rather an appreciation of the ability of governmental units to selectively take sustenance from their surroundings. Collective bodies have the ability to defy their surroundings or to ratify the situation as it obtains. For many segments of society, it is only through or with reference to collective decision-making bodies that recourse from systemic properties can be had. City councils as possible initiators of change or as umbrellas of the status quo play an important role in policy formation for any municipality. If politics can be important for the policies" adopted by a community, councils should be involved in the politics and part of the policy determination. Research on the role of councils in the formulation of municipal policy should illuminate the fashion in which politics can impact on policy. Part of the reason why city councils have attracted so few researchers probably lies in their failure to appear dramatic. In fact, observing a

PAGE 24

„8 council is often an exercise in tedium; the flow of action is difficult to capture, while the minutia of city governance appears all too evident. "The casual or occasional visitor to the city council meetings is often distressed at what he sees; major pieces of legislation passed by unanimous vote with little or no discussion, while an hour is spent over the implications of an apparently minor request for zoning reclassification." As with most collective deliberative bodies, debate tends to drag on, and decisive actions are infrequent. Decisions, when they occur, tend to terminate with unanimous or near unanimous voting. In short, councils may appear to bear little resemblance to democratic bodies "designed to institutionalize conflicts and facilitate the clarification, crystallization, and resolution of political differences.' From this perspective, their policy-making role becomes problematical. In fact, we might well question whether environmental factors or demographic characteristics do not really determine much of local public policy as a good portion of the literature dealing with the subject appears to suggest. Perhaps the political decisions made in council chambers have little effect on local policy when compared to the impact of the characteristics of the city itself. But if councils are breakwaters for the consideration of local policy, then council politics should make a difference, and in a way which depends on the issues which a council considers and the politics that is played out among the members. I have taken such expectations and distilled them into three broad concerns which I will address in this study. First, at the aggregate level and for the councils under consideration, this study addresses problems of municipal expenditures and their relationship to the socioeconomic and political, structural characteristics of cities. My intent is to <»-,

PAGE 25

examine the relationship between these factors in Florida cities and to isolate major indicators of municipal dynamics which impact upon Florida council members. This analysis is designed with a view to controlling, in the subsequent analyses, those aspects of the council members' socioeconomic environment which appear to have the most importance for the council's allocation of funds. The relevant caveat to be applied in such analysis is, although one would expect environmental conditions to determine largely at what level expenditures will be set, politics is likely to be pivotal in establishing the distribution of rewards and benefits at that level. The pivotal role of politics is likely to be discovered in similar socioeconomic situations. Council politics should create different expenditure priorities despite similar socioeconomic contexts. The second major concern focuses on the relationship between individual council members, their characteristics, the nature of the decisional structures in which they operate, and the environmental factors that impinge upon them. The inquiry will center on the manner in which the political and social milieu of members help determine the prevalence of conflict or unanimity on the council, and the manner in which council relations and member characteristics shape the council's orientation toward choice situations and the consideration of different policy issues. Finally, the investigation deals with the manner in which councils and members are affected by the types of policy issues which they consider. Do certain types of issues act as triggers which initiate conflict or promote accord? How do the issues which councils handle affect the political methods by which councils settle such issues? Stated briefly then, this study will examine various characteristics of individual council members and salient aspects of their sociopolitical environment in order to determine how the structure of decision within

PAGE 26

the council is organized and to explain the presence of conflict or consensus among council members. Additionally, different types of policy issues will be investigated in order to determine if they are associated with certain patterns of conflict or consensus among the members or with other aspects of a council member's circumstance. Finally, various aspects of local councils will be treated to determine if the political characteristics of a council make a difference for the types of policy a council pursues. Do councils establish similar profiles of political activity in similar socioeconomic settings? Or are there political differences among councils with similar contexts which one might reasonably associate with the peculiar policies of the councils? The Scope and Limits of the Research In an attempt to answer such questions, a questionnaire was sent in the fall and winter of 1974-1975 to 441 council members throughout the state of Florida asking their opinions on a series of items relating to council activities. This included their perceptions and attitudes on various aspects of municipal politics and social issues (see Appendix A) . Every council member elected to office in a city with 10,000 or more residents was sent a form. These cities represent almost half of Florida's total population (49.9%), over 60 percent of the total urban population of the state, and almost 90 percent of all residents living in incorporated areas. ° Two waves of questionnaires were mailed out; the first in November, the second in December. Of those asked to participate, 173 responded or 39.6 percent of the 437 members occupying council seats in this time period. This rate, while itself low, compares favorably with the return rates reported in other questionnaire-based, mail surveys of council members. In addition to the mail-out questionnaire, however, face-

PAGE 27

8 to-face interviews were conducted with council members in twelve cities selected for their socioeconomic similarities. Among those cities, 81.6 percent of the council members were interviewed in sessions that lasted from 45 minutes to an hour and one-half. In all, 61 cities were represented in the sample. In 51 of the cities, two or more members responded, and in 25 cities, a majority of the council members are represented in the results. These results naturally limit the general applicability of the findings, especially for those parts of the study which treat the entire sample of members. On the other hand, within the selected group of cities where nearly all members responded, the results appear to bear considerable validity. The high response rate within this one group allows for some check on the validity of the entire sample. While such inferences must be treated x^ith caution, a check of certain design variables in both portions of the sample reveals the entire group of council members does not differ significantly from the members who have been sampled more intensively (see Appendix A). Of course, it is difficult to impute total similarity on the basis of demographic patterns alone. Nonetheless, the responses do represent a considerable amount of information on the city politics and council life of communities throughout the state of Florida. Statistical problems abound, but the end result promises to be worth the effort. This study is limited by more than its data base. Conceptions of urban politics that draw their focus narrowly are likely to run afoul of the open and dynamic nature of municipalities. My concern for councils issues from the expectation that they stand as the center of a swirling dynamic interwoven with a city's policymaking. My own theoretical

PAGE 28

and practical concerns would have led me to encompass additional elements of the whole, but practical considerations required that 1 limit the problem. I am keenly aware of the caution issued by pioneers in the study of legislative behavior: The notion of the legislature as a decisional system tends to limit the focus of inquiry more than seems appropriate for the purpose of specifying just how legislature and political system are interdependent. . . . The decisional approach ignores the fact that a good deal of legislative behavior does not involve decisionmaking at all.1 -5 Yet I believe the approach I have taken to city councils avoids many of the pitfalls that characterized the decision-making approach per se. Another limitation of this study should be noted; it does not come to grips with one of the most significant aspects of institutional policy making, namely, the issue of non-decision. It has long been realized "that the most important part of the legislative decision process was the decision about which decisions to consider."1 Every community has a systematic agenda consisting of all "issues that are commonly perceived by members of the political community as meriting public attention and as involving matters within the legitimate jurisdiction of existing governmental authority."-^ Yet the systematic agenda will never be considered as a whole for inclusion on the institutional agenda. Only certain elements merit such consideration as a result of the patterns of support and opposition within the community. But the institutional agenda is important not only for those individuals and groups able to control its content, but also for those segments of the community whose participation it is able to influence. This is a vital issue in the consideration of all public policies, however; it has been excluded from explicit consideration in this study for practical reasons. Non-decisions require that an elaborate set of assumptions be posited about a given situation, WMMbq

PAGE 29

10 and that all stages of the decision-making process be monitored. This, in turn, requires an enormous research effort far beyond the scope of this project . There is one further, major limitation to this study; that is, the context in which it is cast. Surveys elicit responses from which a picture of a social situation can be painted. But the questions used to generate the responses, and the responses themselves stand as stony datum. There is little way of assessing the remarks. Only tangentially, through a member's own colleagues, can the situation be evaluated. Of course, in those twelve instances where indepth field work was conducted, the overall perspective can be augmented. Even here, however, no interviewing of a systematic nature was done in the community. Thus specific interpretations may neglect important keys to a more substantial understanding. My only recourse is to fall back upon the questionnaire itself which was designed so that important questions were re-enforced with internal checks upon member consistency. Beyond this I must assume that the perceptions of the members, their responses, are indicative of the situation as it does obtain at one point in time. The Argument of the Dissertation In the next few pages, I will preview the shape of the argument developed in this study. For each chapter, I will present a brief sketch of the approach used and the major findings. The study of public policy has increasingly come to turn on the question asked years ago by Harold Lasswell, "Who Gets What, When, and How?" 16 In particular, recent research on public policy has addressed itself to the "How" of public policy making and its link to "What" is decided upon. Focusing upon the distinctions made in the systems approach, numerous researchers have examined the relative importance of environmental variables

PAGE 30

as opposed to political variables for policy when measured at the aggreage level. 17 Others have suggested the solution lies in approaches at1 8 tentive to group dynamics, mass-elite interactions or community power. This study argues that the concern should be directed toward the conceptual issue of choice and its social manifestations. In fact, it argues the conceptual underpinnings of a policy focus involves a subtle shift in our image of political man; he is someone with a choice to make. Municipal policy is formed by many different actors and in several arenas. In this study, I will consider only the fashion in which the local council helps shape policy through the choices it makes. Unlike many studies of policy-making bodies and city councils in particular, I will not rely heavily upon the conceptual apparatus associated with interaction theory. In Chapter two, I argue that council actions can be considered in terms of exchange theoretical propositions. Borrowing from the legacy of classical economics and primitive utilitarianism, I operate from a series of assumptions: 1. While men do not seek to maximize profits, they always seek to make some profit in their social transactions with others. 2. While men are not perfectly rational, they engage in calculations of costs and benefits in social transactions. 3. While some men do not have perfect information of all available alternatives,, they usually are aware of at least some alternatives which form the basis for their assessments of costs and benefits. 4. While there are always constraints on human activity, men compete with each other in seeking to make profit in their transactions . 5. While economic transactions are typified by material goals in the economic marketplace, men also exchange other, nonmaterial commodities, such as sentiments and services of various kinds. 19 In addition, my conception of social organization also includes the notion of strategic interaction. That is, the behavior of an invididual in a group is conditioned and predicated upon expectations regarding the behaviors of others in the group. ?.^? sgaaam a i

PAGE 31

12 Of course, the major difficulty posed by this framework is the concept of "costs" and its measurement. While there is no general agreement over the manner in which costs can be operationalized in empirical research, the various branches of exchange theory appear to agree costs refer ultimately to the internal calculations of value performed by the individual actor. Costs are the results of engaging in exchange processes where an individual's preferences or goals are thwarted in some fashion. Costs can be dealt with in terms of goals or preferences, but it is the threat to goal fulfillment which is important. Thus. I have assumed costs can be measured in the clash of members over issues, over personalities, over styles. Costs are incurred in the manifestation of conflict. The point of adopting the exchange framework is to present a means of drawing the relationship between politics and public policy as determined in local councils. Specifically, I argue different types of policy can be expected to emerge from the interaction of two kinds of costs incumbent to council members: 1) those associated with the pattern of demands impinging upon the members from sources external to the council, and 2) those incurred by policy makers in the formation of decision coalitions among the members. For example, the more conflictual the policy area, the more likely the policies involved are to be redistributive. But this framework can also be used to investigate the ways in which council members reach decisions. In particular, this approach can be used to explain the considerable strains toward unanimity that exist in local councils . The exchange framework provides a means of animating the policy process that is typically neglected in the systems view. This neglect is

PAGE 32

illuminated in Chapter three. At the conceptual level, the systems view links environmental factors and the political system with policy outcomes as shown in Figure 1-la . Researchers such as Dye and Hawkins , arguing within this framework, have established the connection between socioeconomic factors and expenditures using the aggregate characteristics of cities portrayed in census statistics. 21 Yet their specification of the systems model has been quite ambiguous. Without an explicit consideration of the role played by policy-making institutions such as city councils, any number of hypotheses can link socioeconomic factors and various expenditure measures. Nonetheless, their findings stand. Socioeconomic factors appear to condition the pursuit of certain public policies. The objective social and economic realities of a community can act to impose constraints on a council member's choice. These factors can confound the investigation of political behaviors and their link to policy formulation. Thus I have conducted my research with an eye toward capturing the link between politics and policy by engaging in a dual strategy. In the macro-phase as shown in Figure 1-lb , I examine the link between aggregate measures of socioeconomic factors, certain structural aspects of city government and a range of expenditure measures for the 78 cities in Florida having over 10,000 population. In Chapter three, I pursue this analysis to test whether Florida municipalities differ in these linkages from cities around the country and to display in explicit fashion the strengths and weaknesses of this approach for our understanding of public policy. The lessons of the macro-phase were incorporated into the research design investigating the manner in which the political features of city councils and their contexts affect municipal policy and expenditures.

PAGE 34

15 Chapter four initiates the micro-phase of the research. As shown in Figure 1-lc , socioeconomic factors were controlled in order to highlight the relationship between political features and policy. This was done by grouping the 78 Florida cities into clusters with similar socioeconomic characteristics. Six different group were generated. And while council members were surveyed in each of the SES clusters, extensive interviewing was conducted in the cities ranked highest according to the socioeconomic status of their citizens. These cities, it will be shown, exhibit different expenditure patterns, despite their relatively similar socioeconomic milieu. Extensive interviewing, using a questionnaire incorporating exchange theoretical notions, helps establish why this might be. Figure 1-2 represents an overview of the major conceptual categories and relationships investigated in Chapters five through nine. These features are used to develop the micro-phase inquiry pursued in this study. In particular, it shows the manner in which policy choices are conditioned by on-going political machinations in the council and the community and act as mediators of such political events. Figure 1-3 provides a glimpse of the analytical variables subsumed within each of the major conceptual categories together with the relationships to be examined. These variables and the labelled relationships will be referenced in the discussion which follows. Chapter five begins a detailed investigation of council members and councils throughout Florida aimed at forming an empirical foundation for the exchange assumptions made in Chapter two. This chapter provides a description of the characteristics and opinions of the individual council members and their councils. It presents their responses on the many

PAGE 35

16 PERSONAL ATTRIBUTES a. •4 *CONTEXTUAL FACTORS POLICY ISSUES POLITICAL FEATURES OF COUNCIL AND COMMUNITY Figure 1-2. Overview of the Major Conceptual Categories and Relationships Investigated in the Micro-phase

PAGE 36

17 [PERSONAL ATTRIBUTES] [CONTEXTUAL FACTORS] DECISION ORGANIZATION d. Figure 1-3. Overview of the Major Conceptual Categories, Analytical Variables, and Relationships Investigated in the Microphase of the Study

PAGE 37

18 c. [DECISION ORGANIZATION] AGREEMENT MODES DISAGREEMENT MODES POLICY ISSUE CONFLICT PARTICIPATION BY GROUPS AND INDIVIDUAL CITIZENS ON POLICY ISSUES: EXTERNAL COSTS POLITICAL BEHAVIOR OF COUNCIL MEMBERS ON POLICY ISSUES: INTERNAL COSTS POLICY TYPES: 1. Traditional Services 2. Amenities 3. Promoting Economic Growth 4. Arbitrating Between Conflicting Interests EXPENDITURE TYPES: 1. Service Expenditures 2. Service/Newer Public Works Expenditures 3. Newer Public Works Expenditures Figure 1-3. Continued

PAGE 38

19 scales and measures tapping members' personal attributes, perceptions of contextual factors and council characteristics. It also considers the manner in which councils with similar socioeconomic environments differ according to such features. In the wake of this analysis, council members emerge as conservatives, politically and religiously, whose perceptions of local problems are dominated by growth-related issues. As with their colleagues across the United States, Florida council members are predominantly white, male, businessmen in their forties who tend to be better educated than the rest of the population. However, their predispositions, their perceptions of community problems, their reliance on information sources (cues), their evaluation of influential groups, and the types of relations which obtain among members on the same council do not appear related, in any systematic fashion, to the socioeconomic milieu of the councils themselves. The ability of environmental conditions to constrain the attitudes or behaviors of city councils appears limited when examined in this fashion. The univariate analysis presented in Chapter five underwrites the data search that marks the next stage of the analysis. Among the high SES cities of Florida, for example, the Boca Raton council evidences consistent strains of polarization together with rather liberal political predispositions. Contrariwise, Cocoa Beach is both conservative and relatively cohesive. The coincidence of several such measures in the univariate analysis prompted an examination of the cognitive structures of the council members and the manner in which these structures predicate council characteristics. In Chapter six, links a and b of Figure 1^3 are examined through a systematic investigation of relationships between the many analytical

PAGE 39

20 variables. For example, liberal members are shown to be sensitive to human resource problems in their community, to perceive people-oriented groups as important in community decision making, and to be receptive to information from citizens. In this case, personal attributes are seen to be associated with certain contextual factors. Further, the associational measures used in Chapter six help establish the manner in which certain council member attributes and predispositions favor the evolution of forms of interaction among members. Predispositions appear to condition the manner in which a council's members take sustenance from groups within the community. Conservatives, for example, are much more likely to attend to the influence of business groups than are political liberals. Religious conservatives, on the other hand, evidence a completely different set of tendencies. The thrust of the argument initiated in Chapter six is that given sets of empirically defined predispositions and established patterns of council relations may condition the appearance of conflict among the council members. And if the exchange theoretical notions developed in Chapter two are correct, they should also condition the way in which costs are apportioned to various types of policies. In fact, the predispositions of council members and the relations obtaining among council members do appear to prejudice the manifestation of conflict in the council. This association, together with the other relationships specified in link £ of Figure 1-3 , is investigated in Chapter seven. Not surprisingly, conflictual relations among the members such as polarization, stratification and the perception of community pressure are associated with high levels of council conflict. However, the most decisive political predisposition biasing conflictual situations

PAGE 40

appears to be the degree to which the members evidence religious conservatism. The greater the religious conservatism of the council, the less conflictual relations will be and the less conflict will surface over matters of policy. Among high SES cities, issue conflict is most likely where the differences among the members on religious conservatism are large. The amount of conflict which surfaces in a council over matters of policy or between the council and the city manager may affect the manner in which the council chooses to decide on matters of policy. No council faces every issue which comes before it in exactly the same fashion. In Chapter seven, I argue councils are likely to develop and use a repertoire of modes to resolve policy choice situations. If a council has established a mix of disagreement and agreement modes by which it organizes itself to resolve most of the issues before it, then issues which generate considerable costs should evoke disagreement modes. This is the argument pursued in Chapter eight; it examines the link between decision organization and policy issue conflict shown in Figure 1-3 , relationship d. And it does appear to capture a bit of the empirical situation. However, councils are frequently able to handle controversial issues within the context of agreement modes which rely upon bargaining schemes or decision tacts which call upon all members to work toward amicable agreement. On the other hand, controversial policies appear to fracture decision schemes based on interpersonal relations as the foundation for consent. The analysis presented in Chapter eight demonstrates that ongoing political procedures developed to handle issue choices can survive even the most controversial situations. But it also suggests that certain !T*w^; ^^ ' w q m j w^ra^

PAGE 41

22 political arrangements will be unable to withstand the costs of certain issue decisions. For example, certain policies regardless of the conflict they generate, do not engage generalized patterns of member disagreement. Members appear to act as though consensus is necessary in certain policy areas. Similarly, bargaining is typically excluded from issues which generate large amounts of conflict and which attract the attention of significant community groups. Exchange theoretical conceptions suggest members will have highly differentiated views of a given policy; certain policies may have components which act to exclude their consideration by a given decision mode. Policies not only evoke decision modes differentially, they also generate different amounts of decision-making costs and act to galvanize community groups or citizen actions in different ways. The exposition developed in Chapter nine examines these relationships, labelled e in Figure 1-3. It demonstrates that amenities such as public welfare and community development projects constitute a set of policies quite likely to provoke council disagreement and community actions. Traditional municipal policies such as the provision of police and fire protection are unlikely to attract such attention. The analysis developed in chapter nine makes it clear that not every council is capable of sustaining the political costs necessary for the pursuit of amenities. Cities which share relatively similar socioeconomic characteristics may pursue quite different policy priorities. This is exactly the relationship depicted in Figure 1-lc and examined in Chapter ten. The findings buttress the conclusions of Chapter four. Not only do cities with similar environments spend their monies differently, they also vary systematically in the substantive policies

PAGE 42

they pursue and in the most important issues which surface in their communities. The politics of the cities differ and have different effects. Not surprisingly, liberal, people-oriented councils where the religious differences among the members are large are more likely to pursue policies which provide for amenities than conservative, business-oriented councils. By focusing upon the empirically defined preferences and interests of council members, it is possible to illuminate the fashion in which councils help determine the policies of local government. The political features of the local council are related to the policy priorities of the community. These political features remain important even when the socioeconomic context of the council remains relatively constant. But just as the policy choices of a community are conditioned by the ongoing political features of the council, so, too, these features may be mediated by the policies considered by the council. As certain policy choices make their way to the council chamber, their consideration can alter the the way in which the council does its business. In the end, the formulation and consideration of policy by local policy-making bodies remains an immensely complex process.

PAGE 43

Notes 1 Arthur Maas , "Division of Powers: An Areal Analysis," A Theory of .1 Government , ed. Arthur Maas (Glencoe, 111.: Free Press, 1959), pp. 2 James Sundquist, Making Federalism Work: A Study of Program Coordinai at the Community Level (Washington, D.C.: Brookings Institution, 1969), Wf. 3 Matthew Holden, "The Politics of Urbanization," People and Politics i Jrban Society , Vol. VI: Urban Affairs Annual Review , ed. Harlan Hahn perly Hills: Sage, 1972), pp. 557-602. 4 Ibid . 5 See Thomas Dye, Understanding Public Policy (Englewood Cliffs, N.J.: ntice-Hall, 1972) for a discussion of a number of models of policy analysis. Chapter eleven, Dye concentrates on what I term the demographic model. 6 Robert Morlan, "Life on the City Council: Realities of Legislative itics," Capitol, Courthouse and City Hall , ed. Robert Morlan (4th ed; [ton: Houghton-Mifflin, 1972), p. 213. 7 Heinz Eulau, "Logic of Rationality in Unanimous Decision-Making," ional Decisions , Vol. VII, Nomos , ed. Carl Friedrich (New York: Atherton, ;7), pp. 26-54. 8 Ibid ., p. 29. 9 For a complete survey of this literature, see Brett Hawkins, Politics Urban Policies (Indianapolis: Bobbs-Merrill, 1971). 10 Calculated from statistics provided in U.S. Department of Commerce, i'eau of Census, General Social and Economic Characteristics: Florida , 1970 sus of Population, PC(1)-C11. 11 Responses from four cities indicated four different council seats fe vacant either through death or resignation. The list of council members j obtained using Florida League of Cities, Officials of Florida Municipalities : [4 (Tallahassee, Florida: Florida League of Cities, 1974). 12 The response rates from mail questionnaires are often quite low. taever, the problem is especially acute when the respondents are not provisionals of some sort. See Delbert Miller, Handbook of Research Design Social Measurement (New York: McKay, 1964), p. 77. For example, Benedict der had a response rate of 40 percent among council members in cities of r 10,000. See Benedict Kweder, "Beliefs, Attitudes, and Policy Preferences:

PAGE 44

25 Views of Some North Carolina Mayors, Councilmen, and City Managers," ;. (unpublished), University of North Carolina at Chapel Hill, 1972, 87. Lyman Kellstedt had a response rate of 34 percent among Philadelphia Lcil members. See Lyman Kellstedt, "Precinct Committeemen in the Philaihia Metropolitan Area: An Analysis of Roles,: Diss, (unpublished), Uni;ity of Illinois, 1965. Ronald Matheny's response was just 29.3 percent. Ronald Matheny, "Role Conflict: A Quantitative Analysis of the Urban ^ Lslator," prepared for delivery at American Political Science Association :ing (Chicago: August 29-September 2, 1974). Only when the researcher is i to get the help of governmental officials does the response rate exceed >ercent. See Ross Robson and James Sheffield, Jr., "Effects of Population ; on Problem Perceptions and Policy Preferences of Local Government Of fiLs," prepared for delivery at Annual Convention of the Southern Political mce Association (New Orleans, Louisiana: November 7-9, 1974). Additional Drmation can be found in Eli Cox, III, Thomas Anderson and David Fulcher, appraising Mail Survey Response Rates," Journal of Marketing Research , . II (June, 1974), pp. 413-417, and Herbert Blumberg, Carolyn Fuller and L Hare, "Response Rates in Postal Surveys," Public Opinion Quarterly , Vol. 7III (March, 1974), pp. 113-123. 13 John C. Wahlke, Heinz Eulau, William Buchanan and Leroy Ferguson, Legislative System: Explorations in Legislative Behavior (New York: ey, 1962), p. 378. 14 Raymond A. Bauer, Ithiel de Sola Pool and Lewis Dexter, American iness and Public Policy: Politics of Foreign Trade (New York: Atherton, 3), p. 404. 15 Roger Cobb and Charles Elder, Participation in American Politics : Dynamics of Agenda-Building (Boston: Allyn and Bacon, 1972), p. 58. 16 Harold Lasswell, Politics: Who Gets What, When ^Jjow. (New York: raw-Hill, 1936). 17 Hawkins, op. cit . 18 Dye, op. cit . 19 Jonathan Turner, The Structure of Sociological Theory (Homewood, : Dorsey, 1974), p. 212. 20 This approach is adopted from that presented in James Buchanan and don Tullock, The Calculus of Consent: Logical Foundations of Constitutional ocracy (Ann Arbor: University of Michigan, 1962). 21 Terry Clark, Community Power and Policy Outputs: A Review of Urban earch (Beverly Hills: Sage, 1973).

PAGE 45

CHAPTER TWO THE POLICY FOCUS, CITY COUNCILS AND EXCHANGE THEORY "The city is manifestly a complicated thing." No less is local government. It stands as an intricate artifice serving several intents. At base though, local government is charged with performing several policy functions. However, municipal public policy issues from many sources within local government and is conditioned by a welter of community forces. In this chapter, I am concerned with sketching a framework whereby the policy machinations of local councils might be considered. My focus is narrow. I deal only with city councils and the way in which they shape policy. But the policies with which councils deal range across a broad spectrum of concerns. The scope of these matters is not easily apprehended by many of the perspectives commonly used to analyze public policy. Moreover, many of these policy perspectives do not facilitate investigation of the policy process, and it is in this process that the council fits. The policy focus I present here turns on the notion of choice and flows from exchange theoretical considerations. This view seeks to incorporate a regard for the manner in which facets of the council decision-making process interact with the pressures placed on the council by forces in the community and thereby shape different policies. It also recognizes that policies themselves take on meanings within a political context and act to condition the politics of their consideration. 26

PAGE 46

Policy Models and the Policy Focus While it is true that "policy is not a new concern of political science," it is nonetheless the case that the last decade has witnessed a renewed concern with and interest in matters of public policy. 3 Certainly the explosive social events of the sixties and the great number of federal initiatives in social legislation contributed to this re-awakening. But the phenomenon can also be traced to ongoing discussions between scholars and a changing view of politics and public policy within the literature of political science. "Once upon a time," Charles 0. Jones has written, "V.O. Key reflected on some of the possible consequences of one-party factionalism in the South. He (Key) was particularly interested in speculating about the effect of 'disorganized politics' or 'loose factionalism' on the match between the 'haves' and the 'have-nots.'"^ At the state level, Key argued, the match usually came down to matters of taxation and expenditure. Viewing politics as "the conflict between those who have and those who have less," Key speculated that for the South of the late forties an understanding of governmental action was to be found in the varying shades of political organization within the states. Key searched the factional arrangements of the southern states for the basis of both their differences and their similarities .5 However, Key's interest in such arrangements stemmed from a larger concern with the way in which southern policy preferences over the issue of race helped perpetuate one-party systems. The need to portray a single front in national political forums such as Congress helped establish a reliance on a single party for national articulation of policy. Key felt that "southern political regionalism derives basically from the

PAGE 47

28 influence of the Negro," and he concerned himself with demonstrating the manner in which this influence affected politics in general and party politics in particular. 6 Key assumed two-party politics might signal a change in policy, but he recognized changes would not come automatically with a move to two-party politics. Key was intent upon sketching the broad outlines of policy movement. As a result, his portrayal of the determinants of public policy avoids a rigidly deterministic conception of the consequences of organization for policy. Instead, Key's approach to the consequences of politics for public policy involved a very complex and subtle view of the interplay of forces working within the southern states. It is not surprising, of course, that political scientists such as Key would undertake studies involving the relationship between politics and policy, and ultimately conclude politics did influence policy. Indeed, such findings almost seemed axiomatic; the study of politics was predicated upon its importance for policy. However, not all researchers have documented this importance. The level and nature of public expenditure, or the form of public policy, had always been a relevant concern for economists. And interestingly enough, when such economists attempted to account for the differences in the levels of spending of the various states, political variables were seldom included in their models. Economists such as Solomon Fabricant and Harvey Brazer found certain aspects of the social and economic environments of the states were the most important determinants of spending levels. Factors such as urbanization, income level, and industrialization were viewed as important to the exclusion of political variables. "In brief, the political scientists, considering only political

PAGE 48

variables, found these to be determinants of policy, while the economists, Q using only economic variables, found economics to be crucial.'" In an attempt to reconcile the difference between the two perspectives, Dawson and Robinson set out to "discover the relationships among the extent of inter-party competition, the presence of certain economic factors, and the extent of nine public welfare policies, using the American States as the units for investigation." 9 Dawson and Robinson relied on the well-known model of the political system framed in the work of David Easton. At heart, this conception of politics argues that certain environmental conditions, primarily socioeconomic in nature, combine with certain characteristics of the political system to produce public policy. . . . Policy, in this view, is the outcome of activity or interaction among external conditions, political system and political process. 1U As employed by many scholars, this model was designed to bridge a gap in previous studies. It was intended to include both political variables and environmental variables, and to be generally inclusive of a wide range of political and social situations. Moreover, this model allowed for a systematic and precise examination of the effects of each of the variables on measures of public policy through multivariate techniques. Indeed the synthesis embodied in the model did seem to focus squarely upon the determinants of public policy. The Dawson-Robinson study, like those of Thomas Dye which followed, was able to conclude the role of political processes in the determination of public policy had been overdrawn. Dye, for example, suggested that "political science has been guilty of viewing political life as a closed system." 11 The thrust of

PAGE 49

30 Dye's findings pointed toward the relative importance of certain external conditions in the formulations of public policy. Such conditions, he argued, were more influential in shaping policy than had been suggested in the writings of many political scientists. However, many political scientists demurred in the wake of such findings. New tacts were taken to demonstrate the pre-eminence of politics. New variables were introduced into the model; expenditures were criticized as poor indicators of policy; additional linkages were specified. 12 At best, the results of these efforts were ambiguous. They never seemed to catch the importance of politics for policy which appeared all too obvious to political scientists. The poignant observations of scholars such as Key were by-passed in the refinements of the new approach. While the faults of the analysis offered by Dye and others have, by this time, been repeatedly dealt with in the literature of political science, one point often goes unnoticed. 1 ^ Language both informs thought and betrays it. The synthesis constructed by Dye and accepted by many others as the basis for discussion about matters of policy purported to deal with the same questions raised in previous considerations of the topic. Heuristically, the simplified systems model promised to clarify the discussion of public policy and to highlight prominent relationships. In fact, this seemingly benign simplification significantly altered the political and economic images embedded in the conceptions of public policy which had preceded it. The terminology persisted, but acted to serve different ends. The ramifications of the "systems" model used by Dye and many others entailed a synthesis that was mainly illusion, based upon a common vocabulary but offering quite different hypotheses about the nature of the influences shaping public policy.

PAGE 50

In most of the discussions that animated the debate ever the synthesis suggested by Dye, the concern focused upon the appropriateness and importance of either political or environmental variables. Hypotheses were offered that talked of the determinants of public policy, but the form of the hypotheses was typically ignored. The particular form of the systems model assumed during the debate cast explanation of policy in terms of two alternatives, environmental determinism as opposed to free-will environmentalism. 1 ^ Authors such as Fenton, Lineberry, Fowler, and Clarke who contested the evidence pointing to the importance of environmental considerations in the determination of outputs responded with metatheoretical statements about the mediating role of political processes. 1 But, in fact, it was a sense of process which was missing in most of the hypotheses. The operational measures used in the investigations were quite varied, and yet researchers grouped them into two general types — indicators of political processes and indicators of environment. The complex dynamics by which various factors acquired import for public policy were largely neglected. Specifically, in eliminating the role of policy makers from their direct consideration both sides of the debate had to content themselves with speculation about the manner in which variables became important for public policy. Many of the most important questions about the determination of public policy per force had to remain unasked and unanswered. For all the problems of the model developed by Dye, for all the criticisms of it, for all that, one impression remains. The literature of political science has been fundamentally changed; its course redirected. The preoccupation of political scientists with system variables noted by Dye, while still evident, has given way to a generally shared concern

PAGE 51

for public policy. 16 Even traditional approaches to the study of politics are being re-evaluated in terms of their ability to shed light upon considerations of policy. In the last several years, students of public policy and of municipal government have resorted to several approaches focusing on what Vernon Van Dyke has called salient-political-features. *' Lacking a well defined or established approach to the study of public policy, some authors have argued for the importance of elections in shaping municipal policy. Local institutions, pressure groups and powerful influentials have marked the investigations of still other researchers. 18 Political cleavages have also been taken to have meaning for a city's public policy, and researchers have examined the social and economic characteristics 19 of municipalities to produce surrogate measures of such divisions. In this view, the political system is viewed "as a more or less competitive arena in which groups compete for political advantage and in which individuals take cues for political behavior from the reference groups with which they identify. "< ) Social and economic factors have been used to buttress yet another conception of policy variation which appeals to the underlying cultural differences of cities. In this view, the "ethos" discernible in a city's social, economic, and political structures has an effect upon the types of policies adopted by a city.^1 All this is not to argue political influence is unimportant as a concept or that various models of governance have no merit. To the contrary, what the growing body of policy studies in American urban politics portends is a shift in perspective — one that extends greater importance to policy and its consequences while encompassing the older concerns for political influence and community stability. The recent literature

PAGE 52

not only echoes the attention given the distribution of values in previous studies, but also questions the range and scope of policy alternatives available. The much heralded policy focus involves a subtle shift in our image of political man; he is now seen to be someone with a choice to make. Politics has become an activity whereby the preferences of some individuals are selected to be the choices of society, Choice, however, has been the point of departure for several conceptions of politics. For example, David Easton's attempt to deal with the "authoritative allocation of values" naturally concerned itself with choice and the manner in which norms and values in society act to both limit and define the possible choices by governmental bodies in that society. Similarly, the various schools of decision theorists have long focused upon choice situations. Their manner of speculation on the ramifications of choice for the entire polity is extensive and varied." To the extent a policy focus subsumes and borrows from these conceptions, the approach will stand on the belief that choice is an important "focal point of the social sciences, uniting normative, structural and behavioral theory." 24 Yet a policy focus entails more than decision making, more than a mere concern for choice. "Policy making is the process of making choices," as Apter has suggested, but a research focus must be informed by more than such a simple prescription. 25 Perhaps best illustrating this point • is Eulau and Prewitt's insightful disquisition on "What Policy Is Not." 26 Policy is not, they argue, a synonym for goals and objectives. It is not an intent to do something. Nor is it a synomym for particular events in the policy process, such as the actions of a governing body. And it is m ost certainly not mere decisions or actions. Rather, "policy is a strictrt^>^ ^^affrf fts^'i^Wiif<^^^^* > ^^'^^

PAGE 53

34 ly theoretical construct that is inferred from patterns of relevant choice behavior by relevant political actors and the consequences of choice behavior. . . .If the behavioral patterns are consistent and regular, then the existence of policy is inferred and identified." Choice is a key element in this formulation; it establishes the basis for our inference about political man and the political system. But the policy focus is more than "a theory of one segment." 28 In order to be meaningful, the focus itself must be illuminated by images of man and the political system which supports its basic thrust. Its application must handle the contextual relevance called for in Eulau's definition. In particular, the images animating the focus must be geared for the research in question. Approaches to the Study of City Councils A policy focus is fundamental to the problems to which this study is addressed. However, such a focus cannot stand apart from the problems themselves. To the contrary, the two complement each other. Moreover, the tractibility of such problems is likely to vary considerably according to the theoretical schemes employed to approach them. In social science, where various images of man and society abound, some perspectives appear suited for answering certain questions and ill-suited for others. A brief review will show this to be true of the consideration scholars have given city councils. In the last ten years, several authors have attempted to deal with city councils and their role in the formulation and promulgation of urban public policy. I will examine four studies. To varying degrees, each exposition has been illuminated by different images of the political system and of political man. Each has contributed valuable insights into the evolving notion of municipal policy making.

PAGE 54

John Crecine's Governmental Problem Solving is the most explicitly modelled and narrowly conceived of the four. Testing a complicated simulation model of big city budgetary processes against the actual expenditures in Cleveland, Detroit, and Pittsburgh, Crecine postulates a model consisting of three main elements or processes: a departmental or bureau process, a moral process, and a city council process. Each interact to produce a budget. For Crecine, though, "the role of the city council is a limited one." 29 In fact, he argues that the city council has only seven feasible strategies for dealing with a mayor's proposed budget. DU As Crecine views the budgetary process, the serious constraints upon the council in terms of time, expertness, and information make it little more than a rubber stamp for the mayor's proposals. Thus it is not surprising Crecine finds councils do tend to produce decisions on expenditures closely paralleling those recommended by the mayor. 31 Only in cases where the council chooses to alter the tax rate does Crecine acknowledge difficulties with this conception. The thrust of Crecine's conclusions are not dissimilar from those of Wildavsky and others writing of federal-congressional budgeting. In a sense, Crecine is saying that if the budget ceiling is fixed, no new revenues are raised (his model has a difficult time explaining that aspect of council behavior), and no catastrophic events hit the community, then the budget levels, examined by function, will not change very much. The argument becomes that if very little changes within the community, then budget levels change very little as well. This is not to demean Crecine's work, but rather to caution about its validity under circumstances significantly dissimilar from those posited. Certainly it is true that Crecine has captured an aspect of municipal budgeting observed in many cities throughout the United States. Namely, -v-r yt&tfpsBrtre&cr''*

PAGE 55

little change does occur from year to year in expenditures, and his assumptions handle this particular phenomenon quite well. However, aside from the simplified nature of the assumptions of the model, Crecine's findings are limited by the process he studied — budget making — and the context from which he gathered his data — large cities with strong mayor forms of government, cities whose council's have been viewed as relatively weak when compared to other community political actors. 2 Clearly for the problems to be considered in this study, Crecine's approach and his assumptions seem inadequate. John Dearlove has offered a different approach. Studying council activities in Kensington and Chelsea, England, Dearlove starts from a vastly different perspective than Crecine but nonetheless concludes by buttressing Crecine's view. Dearlove suggests, after the manner of Crecine, that there are strong tendencies among councils toward policy maintenance. However, this tendency flows not from the weakened position of the council or the constraints upon its information gathering abilities, but rather from the council's strengths. Dearlove concludes that councils are entities able to control and shape the environment, but, for several reasons, are not inclined toward change. 33 Clearly, though, Dearlove is only considering one side of the coin. By ignoring the conflictual nature of policy change, he has excluded an exceedingly important aspect of the policy process. This is not an inadvertant omission, however, but rather a product of the model used by Dearlove. He relies extensively upon interaction theory, and more specifically, on role theory. Role theory, by positing a structured vision of human behavior based upon regularized expectations, assumes a well defined and persistent form of social organization. 34 The veteran of any council meeting might recognize the intuitive appeal of this position ae**Ais.TiL r .-t*i: HMPKfHHSM

PAGE 56

for an analysis of council activities. But councils also share the characteristics of small groups. Consensual aspects may predominate but there exist strong tendencies toward shifts in behavior. Additionally, councils are themselves not typical of many organized small groups. In fact, they might aptly be characterized as organized anarchies. 35 That is, they are organizations with three general properties. The first is problematic preference; the council "discovers preferences through action more than it acts on the basis of preferences." Second, although the organization manages to survive and even operate smoothly, its own processes are not understood well by its members; trial and error procedures predominate. Third, "participants vary in the amount of time and effort they devote to different domains; involvement varies from one time to another." 37 To the degree these statements appear to capture salient aspects of council life, the assumptions of role theory become quite tenuous and of limited use in addressing the problems motivating this study. The third of the major studies dealing with city council activities and public policy was done by Bryan Downes. Quite distinctively, his research sets out to investigate communities and councils marked by changing values and divergent community interests. His dissertation, along with several published articles, focuses upon 37 suburban municipalities in the St. Louis region which had experienced varying degrees of population growth during the fifties. Downes begins his study with a framework drawn from the work on community conflict conducted by James Coleman and William Gamson. JO U Downes arrives at several conclusions. He finds that councils do 1 disagree and often, that council decision processes have little effect on wwae-flEf***"***"

PAGE 57

aggregate measures of a community's policy commitments, that member background seems to have some association with the policy choices of the council, and that, in the main, "socioeconomic characteristics are probably more important determinants of municipal fiscal and land use policies than political processes. He traces this importance to the manner in which socioeconomic changes manifest themselves in the divergent values and interests of the new community inhabitants associated with population growth. This relationship is not an automatic one, however. Rather different issues surface in different situations and spark varying degrees of dissension among the council members. For example, Downes indicates "In very high growth rate communities, issue disagreements appear to be largely a function of enduring factions, and these cleavages affect the extent of conflict, on most issues. On the other hand, in medium growth rate communities, issue disagreements did not appear to be the result of enduring council splits." While certain issues such as zoning and land use planning seem to generate much controversy in most cities, other issues are not clearly associated with growth patterns or socioeconomic variables. Downe's study is directed toward many of the same concerns motivating this research, and hence his approach is of particular interest. He deals with council conflicts and unanimity. He stresses linkages between council member characteristics and municipal policy. Yet for several reasons, Downes' framework appears inadequate for the present purposes. Methodologically the systems framework and the holistic conception of community are not well suited to considering council members as the fundamental units of inquiry. When dealing with members, as he often does, Downes resorts to ad hoc explanations; his theoretical framework •**3W!?l»?Ni»v

PAGE 58

39 betrays the evidence he musters. Unlike Dearlove, who works within the heritage of role theory, Downes' legacy flows from the wellsprings of conflict theory. When the unit of analysis is the community and the focus is on socioeconomic measures, all is well and Downes is convincing. However, as Downes stoops to include council members and their relation to policy, he falters. Having assumed council members to be reflections of the broader community context, the framework fails to establish adequately the link between members and policy, and ignores Dearlove' s insights regarding the autonomy of the council. Indeed most cf Downes' indicators of member perceptions only dimly illuminate the manner in which members help determine policy. The concern for community differentiation, conflict and policy is salutary, but Downes' approach is unlikely to produce answers to the puzzles considered in this study; it assumes many of the links this study will investigate. The City Council Research Project directed by Heinz Eulau and Kenneth Prewitt is perhaps the single most ambitious attempt to gauge the relation between environmental challenges and policy responses in terms of the purposive actions of decision makers. 41 Begun primarily as a contribution to the study of legislative institutions, the project emerged as a highly complex and intricate view of the entire policy process, and the most detailed and thorough study of council activities to date. Prolific in number, varied in their focus, the several books and articles derived from this investigation nonetheless seem to forge a common theme. Namely, policy positions are emergent responses by decision makers to the ecological situations and their own predispositions which have been conditioned by specific policies adopted over time. 42 Much of this work argues that there is cognitive adjustment to environmental pressures and K?£!

PAGE 59

40 to the previous patterns of decision making. The underlying image of political man that issues from such conceptions emphasizes his adaptive qualities, as one who engages in costly processes of social interaction and who acts under conditions of uncertainty with notions of what has been don e before and what might work at any given moment. For all its descriptive detail and calculation, however, the City Council Research Project is supported by a meager theoretical framework. In fact, little pretense is made at presenting a theoretical framework. The major work of the project, Eulau and Prewitt's The Labyrinths of Democracy, is content to portray its foundations in the metaphor supplied in the title. As Eulau and Prewitt explain: We find the metaphor of governance as a labyrinth preferable to other metaphors used in politics. A labyrinth is an enclosure with many entrances and exits. Its layout consists of a maze of pathways, but the pathways can be marked by signs that help one to avoid false moves and lead the seeker where he wants to go. If the paths are not marked, trial and error may yet lead to discovery and the long way back. Democratic governance resembles the labyrinth. The labyrinth has walls that serve as boundaries but are more or less porous; it has major arteries and places of assembly but also byways, detours, nooks and crannies. Passing through the labyrinth may take more or less time, depending on the continuities and discontinuities in the journey by what is known or unknown about the terrain. 4 -* Understandably, Eulau and Prewitt wished to avoid portraying democratic governance in oversimplified terms, to not "reduce the maze of democratic governance to a simple formula." 44 However, even this metaphor is a sort of formula, albeit an intricate one, and to that degree draws the politics of councils all too harshly. Moreover, this particular formula seems endowed with few terms with which to model the political context of councils or from which to draw hypotheses. Of course, Eulau and Prewitt do animate their metaphor. Implicitly, the perspectives they develop appear to be variants of interaction theory.

PAGE 60

Social intercourse constitutes one of the main thrusts of their work. In attempting to go beyond the assumptions of the Legislative System, the authors draw a picture of city councils as adaptxve organizations. The adaptive aspects of councils are painted in terms of internal processes such as stratification, governing styles, communication, task structure and decisional structures; external linkages such as constitutive processes, petitioning processes and member orientations toward these phenomena; and finally, policy as a reponse emerging from purposeful, but adaptive interaction of council members. However, the study never goes much beyond these conceptions. Even these concepts appear with little or no recourse to theoretical concerns or the vast literature on empirical small group research. The interaction perspective that guides much of their discussion deserts them in forging the links between council processes and decision structures. Only in the work of Eyestone and later Zisk does the project face the task of relating conversion processes to policy outputs in an explicit fashion. 47 But these latter efforts, while themselves imbued with considerable merit, lack the more comprehensive approach and detail that sets Eulau and Prewitt's study apart. Much more explicitly Eulau and Prewitt need to examine the kinds of connections which Banfield discussed years ago; namely, how does the distribution of influence and demands made upon decision makers combine with the decisional system so as to j ->48 explain the types of policy a city council adopts.' An Exchange Perspective on Policy Making in Cit y Councils Each of the above studies has its merits. Crecine's successes stem from his elucidation of portions of the budgetary process and the manner in which he traces the limitations of the council in a severely constrained

PAGE 61

42 political situation. Dearlove, on the other hand, is at his best when the discussion turns to the extent of council latitude for action and the means whereby such potential narrows under self-regulating adjustments. The factional aspects of council politics and their importance in many policy situations are handled deftly by Bryan Downes. His emphasis on change and conflict in the urban political scene is a welcome addition to the literature on council politics and policies. Finally, of course, Eulau and Prewitt's Labyrinths of Democracy provides the most extensive and complex view of municipal governance to be found in the literature. Replete with keen insights and encyclopedic in scope, it portrays councils as both a product and producer of community policy. Yet for all these merits, the various perspectives developed by these authors are not well suited for the problems at hand. While one is well fashioned to handle conflict and another unanimity or yet another to forge the link between council activities and public policy, the common denominators are few. Indeed most are not well suited to the notion that the council itself is the focus of inquiry, the wellspring of a municipal policy focus. This study is predicated upon the notion that the council serves as a breakwater for much of the public policy pursued within a community. Regardless of its ultimate importance in the determination of a community's public policy, the council nonetheless provides one of the few institutional arenas to which most everyone in the community can have recourse. Minimally then, much of the potential for changes in community public policy lies with the council and its politics. In order to capture the possibilities inherent in the council circumstance, an adequate framework must amplify the policy focus discussed earlier. If the policy focus

PAGE 62

43 serves to direct our attention to the choices of the members themselves, then it requires a conception of council politics descriptive of the important links in the process of forming such policy. Although the difficulties with the approaches adopted by Crecine, Dearlove, Downes and Eulau have been described, this study hopes to benefit from their insights indirectly by incorporating them into a different approach to the problem. Exchange theoretical considerations will guide this inves49 tigation. Most contemporary versions of exchange theory revolve around and differ with respect to three basic concepts: 1) the idea of utility or costs and rewards; 2) rationality and its implications with regard to information, alternatives, and time; 3) strategic interaction or the nature of interpersonal interaction and anticipation among individuals. As a whole, these concepts provide the link between council members and public policy. Most exchange theoretical expositions hold as their basic tenet an assumption concerning the motivation of individual behavior. In essence, such assumptions can be summarized as follows: "The representative or the average individual, when confronted with real choice in exchange, will choose 'more rather than less. ""'50 In general, such propositions are derived from considerations about an individual's psychological reaction to various stimuli in terms of his inner needs or internal dispositions. 51 When a person receives or has directed towards him an activity defined by that person as valuable, he is rewarded. 52 Rewards, in such conceptions, are what contribute to the desire for "more." On the other hand, an activity that is punitive to the individual, or an alternative reward that is foregone in order to get other rewards, is a cost

PAGE 63

44 and makes up what has been called "less." In any case, the idea that individuals have needs and seek values in their behavioral choices and transactions is fundamental to exchange theory. Such needs and values are of course quite diverse and the usefulness of exchange considerations is to a large part to be found in how such factors are conceived with reference to an empirical situation. To say that men do have preferences and that in choice they seek personal values seems reasonable only if such individuals are somehow able to consistently realize their values in choice. While several explanations can be had for the connection between internal dispositions and external actions, exchange theory assumes that "people behave as if they arranged their preferences in some logical pattern. "" That is, somehow people are able to know which value they seek in a given situation. Regardless of the approach, the key to understanding the thrust of exchange theorists on this point is to be found in the concept of rationality. "Rationality is the something we postulate in people that makes them behave in a regular way. And the essence of that something is that people relate their actions to their goals." In exchange theory, man is a teleological entity and rationality is the manifestation of his movement towards his goals, or desired rewards. Observed regularity is the proof of the pudding. While rational behavior may be purposeful, not all exchange theorists view purposeful behavior as rational. This difference stems from the various ways in which the concept of preference has been approached by such theorists. Of course, preferences can never be observed directly. Instead they must somehow be imputed. This inference can be made in either of two ways: preference can be revealed or preference can be posited.

PAGE 64

In the procedure of revealed preference, we assume initially that persons behave in accord with logical rules, e.g., transitivity in ordering. Applying these rules to choices, we then discover what goals must have existed in order to lead logically to these choices. Goals are therefore inferred from a) actual choices and b) assumptions about behavior In the procedure of posited preference, we assume initially that an actor has a given goal (e.g., to win an election, to maximize profit) and we infer that behaving in accordance with rules of logic and this goal leads to particular choices. If such choices actually occur, then we further infer both that the actor actually does have the assumed goal and that he behaves in accord with the rules of logic. If such choices do not occur, then we are at a loss to discover whether the fault lies in our attribution of logic or our attribution of goals. In fact, there is switching back and forth between the two approaches. But the use of posited preferences has suffered from a confusion over their utility and their extensive association with the works of normative rational theorists. In part, such confusion results from the failure of many researchers to apprehend the full meaning and usage of the term rational in the context of the underlying notion of preference. If preference is taken to be revealed in behavior, then the investigator has only to make the connection between stimuli and behavior; few additional assumptions are required. On the other hand, theorists using posited preferences require several additional assumptions. First, a set of preferences must somehow be identified. Second, an assumption of transitivity is usually required. Without this assumption to guarantee that persons can somehow arrange their preferences in some order, the logical relationship between values and goals would be impossible to establish. Finally, under posited preference connectivity is required. That is, in some manner the objects of value in an individual's mind must be comparable. Under conditions of revealed preference, this is no problem. Instead, given a -

PAGE 65

46 perspective relying upon revealed preference, the primary difficulty comes in knowing what irrationality is and in dealing with intensities of preference. Regardless of the approach taken, exchange models draw the relation between internal dispositions and behavior. The manifestation of this relation comes when an individual, such as a council member, attempts to gain some benefit. Many conceptions of this process insist that the individual engages in utility maximization, others argue that people satisfice, while still others attempt to avoid the issue by concentrating explicitly upon revealed preferences. In any case, the problem central to such concerns deals with the way an individual or a council member goes about generating different courses of action. That is, how are alternative choices generated for individual consideration? Clearly, an awareness of alternatives is critical to the possibility of choice. In a rational process, an individual must choose his most preferred alternative, yet the question arises of how the individual determines the set of alternatives for evaluation and what the set may be. Under considerations of utility maximization, perfect information is required. However, this criterion appears untenable in any real council situation, and thus a more reasonable standard might incorporate Herbert Simon's famous view that men satisfice. That is, "in the complex environment of decision making in the real world, choosers are not aware of all possible alternatives, so that they choose not the best alternative but a satisfactory one. They do not maximize, but satisfy." 5 ^ The main difficulty with such a notion, however, lies in the manner in which the individual defines the set of alternatives from which he ultimately makes his choice. If all possible alternatives are not evi-

PAGE 66

47 dent, then an individual must search in order to establish the set. How does an individual judge that his search has ended? Simon suggests that an individual's satisf icing level rises and falls with the ease by which he gains a satisfactory outcome. Simon's description "allows for a person's coping with the varying costs and benefits of searching for alternatives across contexts, situations, and time. Thus, what satisf ices is related to the nature of the environment in which the chooser finds himself, and the method of satisf icing allows him to learn about the nature of that environment." 59 Under many of the most rigid exchange conceptions, such a dynamic has no part; learning plays no role. Simon's solution, however, places the choice situation in a context that incorporates conception of the interaction between an individual and his environment as well as his past experiences. Notice that this description squares with the contextual aspect of policy as defined by Eulau. The problems posed by assumptions of rationality are manifold. In this study, only the weakest form of rationality is assumed— that council members be able to select among their perceived set of alternatives one which seems most likely to realize the most benefit in terms of their goals. The calculus of benefits is based on probabilistic expectations about the success of certain alternatives in obtaining given outcomes within a given context. This calculus involves more though than merely determining what value is to be maximized. There is a crucial complementarity between preference and the social dynamic imputed by the assumption of satisficing. Seldom are values pursued in isolation from the vagarities of other people's actions. Moreover, even for a given council member, the pursuit of values changes with time and circumstance. Implicit in most -

PAGE 67

48 exchange theoretical perspectives are the familiar notions, derived from economics, of marginal utility, strategic interaction, and externalities. Pursuing one course of action, usually limits an individual's ability to actively pursue yet another course. Marginal utility connotes a situation where additional units of a rewarding activity begin to generate costs that exceed the benefit received. Under conditions of rationality, when these costs exceed the benefits accrued an individual ceases his behavior. Of course, an individual will still desire "more rather than less" but the concept of marginal utility raises the question of how much more. The pursuit of values is further constrained within the exchange perspective by an additional assumption about behavior. Namely, the complex set of internal calculations involved in the evaluation of perceived alternatives will normally entail considerations about the "probable choices of others (given each of his own possible actions) whenever such choices by others would affect his evaluation of his own alternatives." Such behavior on the part of two or more interacting individuals is referred to as strategic interaction. This assumption amounts to no more than the idea that each person's choice is contingent on the choices of others. It means that despite the individualistic approach indicated in exchange theory, there exists an implicit requirement for the consideration of the collective context of individual decision making. 61 The assumption of strategic interaction not only places bounds upon an individual's attempts to realize his values, it also introduces an indeterminacy into exchange models not present in many of the traditional economic models of the market and economic man. This indeterminacy is

PAGE 68

49 the result of the uncertainty concerning individual behavior introduced by the attention given the actions of others. In a collective body, individual rationality does not obtain. By creating considerable uncertainty in the likelihood of certain alternatives to produce expected benefits, and by diffusing the correspondence between choice and outcome for an individual, the predictive aspects of the model are weakened. Additionally, such difficulties demonstrate the necessity of tuning the exchange perspective more closely to the empirical referent than might otherwise be the case. General theoretical statements need to be tempered by additional assumptions about the characteristics of council members and the council decision-making situation. The notion of strategic interaction incorporates a familiar concept into most exchange theoretical expositions— externalities. "An externality exists if at least one person in addition to the subject can alter the relation between outcome and alternatives with his choice.'" In the calculations engendered by strategic interaction, it is the evaluation and computation of perceived externalities which help determine an individual's choice. However, in a collective choice situation it is useful to differentiate between externalities over which a council member has no control and those costs he incurs through his participation in a collective decision-making process. The individual's utility derived from any single human activity is maximized when his share in the "net costs" of organizing the activity is minimized. The possible benefits that he secures from a particular method of operation are included in this calculus as cost reductions, reductions from that level which would be imposed on the individual if the activity which were differently organized. There are two separable and distinct elements in the expected costs of any human activity which we want to isolate and to emphasize. First, there are costs that the individual expects to endure as a result of the actions of others over which he has no direct control. To the individual these

PAGE 69

50 costs are external to his own behavior, and we shall call them external costs , using the conventional and descriptive terminology. Secondly, there are costs which the individual expects to incur as a result of his own participation in an organized activity. We shall call these decision-making costs . 64 Buchanan and Tullock argue that expected external costs can only be eliminated if the decision-making rule by which choice is made requires unanimous agreement. This is so since the rational individual will not willingly allow others to impose external costs on him in most circumstances. On the other hand, under the rule of unanimity the expected decision costs are quite high since every individual must be party to any agreement. Thus the rational individual will attempt to reduce these costs through some sort of balancing or trade-off between the two. ^ In order to make meaningful theoretical statements about council processes and public policy, the discussion above indicates the need for some rough gauge of the benefits and costs accrued to council members. The council member must make several judgements before he makes his choice, weighing the several values that might be involved against the costs he is able to predict associated with each action. In determining exactly what the balance might be in a particular choice context, each member relies, necessarily, on his previous experiences and a host of learned responses. In familiar choice situations, a member is likely to be more certain of his decision, and more likely to bear costs with an expectation of benefit. In uncertain situations, a member's expected utility drops and the importance of cost factors rises. Each member is conditioned by a system of empirical beliefs, expressive symbols and values which help define the situation in which he must choose."" While each member may pursue purposes of his own within the collective body, his ability to realize these purposes is constrained by his own predispositions about the

PAGE 70

51 manner in which values should be realized, and the costs incurred in reaching and enforcing the means embodied in such actions. Analytically, benefits and costs must be viewed in terms of posited preferences and revealed predispositions. These notions can be illustrated for city council members by appealing to the distinctions Richard Fenno has drawn about the motivations of Congressmen. Fenno has argued that within the House of Representatives members strive to attain various goals. In particular, he suggests the importance of the following: re-election, influence within the House, good public policy, a career beyond the house, and private gain. 67 Similar purposes can be posited for council members. However, a member's evaluation of the benefits inherent in such goals cannot be viewed as independent of the predispositions toward implementing the goals. To use Rocheach's term, "primitive beliefs" act to structure man's understanding of his environment, and help guide and inform rather specific goals. These primitive beliefs or predispositions are highly valued in themselves. Thus a "good policy" goal—for example, the elimination of poverty—will be associated with different action alternatives for different individuals. Conservative or liberal societal predispositions will significantly alter the course of action taken or alternative chosen by an individual despite fundamentally similar goals. Neither goals nor predispositions are immune from the context of choice, however. It is the context of choice which determines the externalities and decision costs against which goals and predispositions are to be weighed. The goals themselves are best achieved with reference to different arenas or contexts. For example, "re-election" and "career beyond the council" goals appear most likely to involve individuals outside

PAGE 71

52 the council itself, or liable to be balanced against external costs such as voting publics, special interest groups or powerful community individuals. "Good policy goals" and "influence within the council goals," on the other hand, are most likely to be stymied by the structure of relationships within the council itself or the costs of reaching decisions. The benefit captured by the "private gain" goal appears to have reference to costs generated both internal and external to the council. In a similar fashion, a council member's predispositions are mediated by the context of the choice situation. However, predispositions, unlike specific goals, are likely to be quite resilient to short-term circumstances, and hence much more important in the determination of public policy. Predispositions which imbue ends with means appear more fundamentally connected to the individual member's ability to absorb costs, either external or internal to the council. A member's predispositions constitute the core of his beliefs; actions and alternatives which evoke predispositions are likely to be perceived as conferring considerable benefit. In choice situations where such predispositions are involved, it might be expected that the costs a member would be willing to absorb would be correspondingly high. The costs a member might be reasonably expected to encounter have already been considered to be either external to his own control or associated with the process of reaching a collective decision. The external costs liable to an individual member originate from many sources. First, local legislative politics are conditioned fundamentally by the relationship of the council member and the electorate. At the local level, the elected official is joined in what Burke termed, "the strictest union with his constituents." 69 This union connotes more than virtually

PAGE 72

5 3 unlimited accessibility to the member by his constituents. It means the member is not only an official of the community but also an official in the community. Unlike officials at many other levels of government, the typical member works, plays, and lives in the governmental unit whose fate he helps determine. Wrapped as he is in the social nexus of the community's activities, the council member is exposed to an unusual number and variety of potential sanctions. These factors have considerable bearing on a member's ability to be re-elected and/or his future career within the community. Second, a council member seldom earns his livelihood from his salary as a public official. A member typically pursues his council activities in addition to or in determinant of his regular occupation. Despite the benefits that may accrue him because of his position, council duties remain ancillary to other facets. of his life. Moreover, any motivation to do otherwise is usually discouraged. Most municipalities re-enforce the citizen legislator role by providing few services for the council member, by purposively keeping salaries low, and by limiting the official office space allotted for council members. In short, a member's ability to gain from his tenure is heavily conditioned by his relations with the community. Severe external costs can also be generated for an individual council member if he runs afoul of special interest groups within the community or powerful individuals within the community. For a council member wishing re-election, a career beyond the council, or private gain, choices made by these individuals can result in great personal costs. In a similar fashion, the mayor of the municipality or portions of the local bureaucracy can all act to derail a member's ambitions and hopes.

PAGE 73

54 Of course, a council member will have goals within the council itself, and many of his actions within the council will be addressed to the community audience he perceives as most relevant for the minimization of external costs. Nonetheless the member must also work within the council and with the other members. His ability to minimize external costs will, in part, be a function of his ability to determine the costs incurred in decision making. Even in the event where external costs themselves are perceived as unimportant (as in the case of some retiring members) , a member's ability to pursue preferred alternatives still must be balanced against decision making costs. In perhaps the most perceptive article dealing with this aspect of collective choice, Adrian and Press have suggested there are at least eight decision costs involved in the process of forming a coalition capable of decision. The costs incorporate "a variety of economic and psychological factors" which are related most intimately to a member's goals for his tenure in the council, but also bear on his attempts to benefit in arenas outside of the council. Adrian and Press argue decisions include: 1) informations costs; 2) responsibility costs; 3) intergame costs; 4) dissonance costs; 5) costs of the division of payoffs; 6) inertia costs; 7) time costs; 8) persuasion costs. Analytically, these categories cover an incredible range of situations and evoke some measure of the complexity involved in attempting to specify the meaning of costs. As the conceptual basis for an empirical study, however, their complexity is overwhelming. Thus, in the pages to follow, I will develop an argument which focuses on a behavioral indicator capable of subsuming the intent of this categorization. The argument begins by considering the fashion in which costs and choices are intertwined.

PAGE 74

55 Exchange analysis features the act of choice. Sut the act of choice is not simple. For example, different types of issues and different decision-making arrangements should act to produce various contexts for choice and trigger different member responses. If a decision is seen as more than just some final choice — as a whole set of activities which contribute to some final choice — then it can be argued that the process of collective choice may be punctuated by antimonies of various sorts. In collective bodies, the clashes between individuals are customarily handled in accordance with procedures developed over time by the council body. These procedures may be formal or informal. In any case, the processes by which an organization reacts to and resolves political rows do establish themselves. Analytically, Thompson and Tuden identified four general forms of response. 72 Following Simon and March, I have labelled them 1) problem solving, 2) persuasion, 3) bargaining, and 4) politics Thompson and Tuden suggest that the ways in which decisions are usually handled evolve as the result of the interaction of two factors (see Figure 2-1) . The first factor is the degree of consensus or disagreement 73 AGREEMENT Beliefs About Causation: (Means) DISAGREEMENT Preferences About Possible Outcomes (Substantive Goals) AGREEMENT NONAGREEMENT Computation in Bureaucratic Structures [Problem Solving] Majority Judgment in Collegial Structures [Persuasion] Bargaining in Representative Structures [Bargaining] Inspiration in "Anomic" Structures [Politics] Figure 2-1. A Typology of Decision Organization. Source : adapted from James Thompson and Arthur Tuden, "Strategies, Structures, and Processes of Organizational Decision," in James Thompson (ed.), Comparative Studies In Administration , (Pittsburgh: University of Pittsburgh Press), 1959.

PAGE 75

56 among the actors' general orientations towards the means used to reach some alternative. The second factor is the preferences of the council members for that alternative. These governing practices emerge over time as the body acts, time and time again, to resolve the issues that come before xt. While these are analytic categories and not concrete processing to be associated with any one council, the assumption is that within a council a governing practice or practices develop which minimize the decision costs for that group of individuals. Such practices have an effect upon the types of alternatives chosen by the group. For example, in a situation where there is consensus among the actors over the procedural norms used to settle a conflict but disagreement over the goals espoused by different actors, then for collective action to occur, bargaining should be initiated. Yet the fact that a group establishes bargaining as its decision-handling technique has ramifications for the consideration of alternatives which are different from those that would obtain if the group handled these issues consensually. This point can be amplified by considering the argument made by Heinz Eulau which encompasses the interrelationship between individual predilections on issues and those detected within the entire council. * He argues that the organization of decision or "collective articulation" will be impacted by the specificity or diffuseness of the preferences involved at the individual level and at the council level. By diffuseness, he means there is no hierarchy of preference which would give priority in decision making to one demand over another; while specificity means there is just such a hierarchy. The interaction of individual level and council level articulation of preference, Eulau argues, is what determines the

PAGE 76

57 manner in which decision situations are organized. This view is summarized in Figure 2-2 . Articulation of Member: SPECIFIC DIFFUSE SPECIFIC Pluralism Monism Articulation of Group: DIFFUSE Dictatorship AnarchyFigure 2-2. Decision Organization and Preference Articulation. Source : adapted from Heinz Eulau, "Logics of Rationality in Unanimous Decision Making," in Carl Friedrich (ed.), Rational Decisions , Nomos VII, (New York: Atherton) , 1967, 26-54. Eulau maintains that collective articulation is likely to be plural and democratic when the interests of every individual in the group, as well as those of the group itself, are specifically articulated. Anarchic situations are the likely result of situations when the articulation of both the members and the group is diffuse. On the other hand, when there is an asymmetry between group and individual articulation then unanimous decision making is a strong possibility. This is because where the group's articulation is specific and the individual's preference is diffuse, any individual will incur large costs in attempting to "buck" the group's decisions. These costs will not be offset from the benefits he might perceive in acting against the group because the issue does not involve preferences he highly values. Similarly, if at least some individuals have specifically articulated preferences and the group as a whole is diffuse on an issue, then group members having poorly ordered preferences will incur costs and receive few benefits in opposing even a minority of their fellow council members. Eulau' s argument demonstrates the real possibility for unanimity in any situation of collective choice. Notice that highly contested situ-

PAGE 77

I ations are likely only when both the individual members and the group preferences are quite specific. Given the external costs and decisionmaking costs council members have been assumed to incur, unanimity on councils would theoretically seem to be the most prevalent mode of articulating preference. Specificity is just too costly. This remains true even when the context of decision handling is considered. Both the problem-solving and persuasion modes are readily conducive to consensual forms of behavior. Under conditions indicating bargaining, if there exists a specificity of group preferences but a diffuseness of individual preference, then unanimity is quite likely. Since no specific benefits are to be gained, external costs can be minimized by a diffusion of responsibility. Moreover, decision making costs, usually quite high in a consensual situation, will be lowered by the felicity of each member towards bargaining. In short, under such conditions bargained unanimity will result. If, on the other hand, within the bargaining mode, group interests are diffuse while individual interests are specific, then unanimity is likely only if there is a failure in rationality. Eulau has suggested failures are possible because unanimity under such circumstances "serves as a kind of protective screen from responsibility for failure to resolve an issue in those terms in which it was originally defined by the individual with specific interest himself. "^7 Thus, in empirical situations, unanimity is the likely result even when the assumption of individual rationality is made. All this is not to argue unanimity will characterize every council decision. Indeed the possibility of conflict is inherent in every situation where two members have goals or predispositions which are to some degree incompatible. Costs, however, are extremely high for activists

PAGE 78

59 engaged in conflictual situations. "Expenditures of time become greater, stressful encounters multiply, information becomes harder to obtain yet more vital as bluffing strategies are adopted. Even successful activists find their average net utility reduced." 78 Council conflicts are likely then only when severe costs predicate no other course of action. Council conflict can thus signal the presence of substantial costs. And conflict can be generated in many ways. The disjunctive effects of member differences can flow from several sources. Certainly the goals and predispositions of a member are a fundamental and important source. Basic life experiences may well hold the seeds of conflict. A member's age, his education or occupation will help determine his subjective evaluation of alternatives and the costs and benefits they represent. In a sense, such experiences act as surrogate indicators for a wealth of behavioral and psychological experiences. They help establish the manner in which an individual will categorize specific issues. It can also be assumed that such experiences are important for establishing the cognitive processes of a member enabling him to formulate a way of acting upon alternatives once they have been cast. 79 The learned aspect of behavioral response in the context of such experiences should produce cognitive processes that are relatively stable through time. Such stability helps explain why members perceive costs in certain situations, and why they feel cues taken from certain groups and individuals will orient them correctly in a given situation. In the exchange perspective, the conflict or disagreement that characterizes a council is conceived to be the result of a complex nexus that finds the member at the center. His background (life experience), goals, and predispositions are fundamental to his response. However, these fac-

PAGE 79

60 tors can be mediated by the manner in which his perceptions subjectively work to assign weights to various other factors in his environment. Important groups, habitual interactions with individuals in information gathering forays, the manner in which the decisional structure of the council evolves, the importance he attaches to his job; each affects the individual member's contribution to the collective choice. While it is both interesting and helpful to consider the general manner in which council members and their council as a whole interact, it is specific, alternative policies which engender cor.flictual or consensual interactions. This is in large measure what Lowi means by "policies determine politics. ^ The political reaction to a policy choice is measured by council members not only in terms of the institutional costs it may generate but also in terms of the external costs expected from the attending publics. Policy content triggers political calculations based on each member's experientially determined evaluations of the costs attendant to such choices. These calculations, when made by an entire council, determine the institutional future of a policy alternative. Salisbury and Heinz have advanced this very argument. They maintain different types of policy — distributive, redistributive, regulative, and self-regulative — can be expected to emerge as outputs in relation to the mix of external and decision-making costs attendant to policy choices within an institutional setting. They construct a typology that considers 1) the costs resulting from the pattern of demand impinging upon the decision makers and 2) the costs incurred by policy makers in forming the coalition necessary for reaching a decision. 81 Figure 2-3 displays their hypothesized relationships. The thrust of this typology is towards specifying the conditions under which certain types of policy will be made.

PAGE 80

61 It suggests a given type of policy will accrue, typically, benefits and costs for various sectors of the community; each policy has characteristics that generate costs for the policy makers. Demand Pattern: INTEGRATED FRAGMENTED Costs of Reaching a Decision: LOW HIGH Redistributive Self-Regulative Distributive Regulative Figure 2-3. The Conditions Under Which Different Policy Types Can Be Expected To Emerge From City Councils. Source : Robert Salisbury and John Heinz, "A Theory of Policy Analysis and Some Preliminary Applications," in Ira Sharkansky (ed.), Policy Analysis in Political Science , (Chicago: Markham Press), 1970. Salisbury and Heinz argue that a fundamental distinction can be made between two types of policy. They suggest policies are either structural or allocative. By allocative policies we mean decisions which confer direct benefits, material or symbolic, upon individuals and groups. Structural policies we take to mean policies which establish authority structures or rules to guide future allocations. The latter policies are more abstractly formulated and ambiguous in their effect than the former. Allocative policies may vary along a distributive-redistributive axis; structural policies may vary as between regulatory and self-regulatory outcomes. 8 The prime difference between these two types of policies is the decision costs they generate. The presumption is "the more costly it is to organize the requisite coalition on an issue, the more likely it is that the on policy outcomes will be structural rather than allocative."' The point is that in many decisional systems rising information and negotiation costs together with declining utilities will be countered by adopting de facto structural rules which call for actual allocations to be deter-

PAGE 81

62 mined elsewhere. Allocative policies are likely to be implemented only if decision-making costs are low. On the other hand, allocative policies may well be associated with high costs when merely brought into the institutional arena for discussion and deliberation . Salisbury and Heinz also argue that varying demand patterns impinge upon the decision makers and affect the types of policies likely to result from collective choice. Integrated demand patterns, in general, can confer great benefits or result in great costs. A member can illafford to absorb the possibility of large external costs presented in integrated demand by acting counter to the wishes of the attentive public. In the case of fragmented demand, external costs are likely to be much lower since the member can generate benefits in some portions of the attentive public only to lose in others. Of course, the perception of integrated or fragmented demand by a member is guided by his general orientation to various segments of the public. In taking his cues from certain segments of the community, a member provides weights to his perception of demand. Significant efforts will be necessary for certain groups to appear important enough and well integrated enough to pose serious external costs for a member. For other groups, a member will be sharply attuned — high external costs are easily incurred. Thus in most communities where businessmen are among the most organized and best funded groups, members are likely to pay inordinant attention to their wishes. The types of costs and benefits other groups can confer are, in all probability, much lower. Distributive, redistributive, regulative and self-regulative policies result then from the balance of internal and external costs. Yet this relationship can be determined only if one makes assumptions about

PAGE 82

6 3 the manner in which the outcomes these policies represent square with posited preferences and costs. The policies cannot be defined merely in terms of the costs they generate; this would be tautological. Instead the types must be defined in such a fashion as to relate to the costs without explicitly involving behavioral manifestations. One way out of this dilemma is to start with the distinction drawn by Salisbury and Heinz between allocative and structural policies. Allocative policies confer direct benefits; structural policies do not. Suppose then that the conventional meaning of the terms are used. That is, distributive policies are perceived to confer direct benefits upon one or more groups according to previous rules or customary allocations. Redistributive policies connote a shift in the operant allocation of values. In regulatory policy, a rule is established by which government can mediate disputes over values that can be harmful in and of themselves to the public at large. Self-regulatory policy gives this control to bodies other than governmental entities. Both regulative and redistributive policies hold the possibility of immediate coercion, while, for self-regulative and distributive policies, the possibility seems much more remote. Intuitively this seems quite agreeable, but the reason is not immediately apparent. When and why is coercion necessary? Lowi comes close to providing an answer with a quite from David Hume: Two neighbors may agree to drain a meadow which they possess in common: because it is easy for them to know each other's mind; and each must perceive, that the immediate consequence of his failing in his part, is the abandoning of the whole project. But it is very difficult, and indeed impossible, that a thousand persons should agree in any such action; it being difficult for them to concert so complicated a design, and still more difficult for them to execute it; while each seeks a pretext to free himself of the trouble and expense, and would lay the whole burden on others. Political society easily remedies both these inconveniences.

PAGE 83

64 Coercion settles the problem of different expectation and utilities. Coercion is used to establish a new set of expectations. This is the case in the distinction drawn between distributive and redistributive policies. Only a change in allocations characterizes them. In a sense, regulation differs from self-regulation in the same fashion. Regulation is one of the ways governments control societal and individual conduct. It represents a change from selfregulation wherein the individual controls conduct and helps establish societal rules and norms. Thus the key to understanding the types of policy pursued and made within city council Is in positing the values that are supplanting them. One such conception of local political values has been offered by Williams and Adrian in their study of comparative policy making, Four Cities . They argue that local governments act as though they pursue at least four values: economic growth, providing or securing life's amenities, maintaining traditional services, and arbitrating among conflicting interests. °-> Traditionally the role of municipalities has been defined in terms of ensuring economic growth and the maintenance of services. Increasingly, Jhough, the proper role of municipalities has come to encompass the newer values of providing and securing amenities for its citizens and arbitrating among conflicting interests. For municipalities, these changes represent attempts at redistribution and increasing regulations. The selfregulation represented in older economic growth philosophies has been challenged by new values. Traditional services are being supplemented and augmented by new concerns for the well being of citizens. These notions can be used to animate the original typology offered by Salisbury and Heinz. Once incorporated into their scheme, new hypotheses may be formed as displayed in Figure 2-4.

PAGE 84

Demand Pattern: 65 Costs of Reaching a Decision: LOW HIGH INTEGRATED Amenities Promoting Growth FRAGMENTED Traditional Arbitrating AServices mong Conflicting Interests Figure 2-4. Modified Version of Salisbury and Heinz Typology. The redistributive nature of many amenities will produce a situation within councils such that policies favoring amenities will only be produced when decision-making costs are low relative to costs inherent in an integrated demand pattern. In general, traditional services will pass through councils with considerable ease since decision-making costs are likely to be low and the demand pattern fragmented. On the other hand, regulation seems to capture the essence of arbitrating between conflicting interests. This type of value typically follows structural trends. "Emphasis is placed upon the process rather than the substance of governmental action." 87 While not all arbitration involves regulation, most regulation will consist of arbitration. In arbitration, the role of government is to provide a "neutral" arena in which claims may be resolved. Such an arena is needed when conflicting demands impinge upon the decision makers and the internal costs of making a decision become great. Williams and Adrians' conception of the values promoting community growth fits into the sense of self-regulation mainly because of the philosophy which has stood behind such values. Namely, promoting growth has traditionally manifested itself as the set of municipal policies favoring business. When municipal policy did intervene with the economic and population growth of the city, it was to provide incentives. In most matters, business interests were allowed to regulate their own growth. The assump-

PAGE 85

! 66 tion was that the city might only prosper if its government allowed business considerable latitude in determining the conditions in which it would function. Even after the relationship between community values and policy i types has been posited, there remain difficulties with the scheme presented by Salisbury and Heinz. Chief among these is the indeterminacy associated with member costs. Although various costs and benefits have been assumed to obtain, the salience of a specific issue for various costs and benefits is likely to differ from member to member. Policy content may be subjectively different for different members. In this Case, costs may be a function not only of the subjective unacceptability of alternatives, but also of the uncertainty and incomparability associated with the alternatives."" In other words, conflict among council members may not be an accurate reflection of the clash of goals and predispositions. Within the context of this study, this difficulty will be resolved by assuming that members are similar in the uncertianty and incomparability they assign to the alternatives that flow from the consideration of an issue. Decision costs will be viewed in terms of the average amount of opinion disagreement that prevails within the council. As developed in Chapter eight, council disagreements will be taken as surrogates for high subjective evaluations of costs. For all its convenience, this represents nothing more than a restatement of the Axelrod hypothesis: the greater the incomparability of the goals, the greater the conflictual behavior. °9 By making disagreement the inverse of policy consensus, this conception fits squarely into the exchange perspective and encompasses the po-

PAGE 86

67 licy focus. It is the case, however, that in the dynamic assumed to exist among normative, structural and behavioral factors, policy is a convenient, if arbitrary, focus. An issue gains salience and attracts some patterns of support or opposition outside the council as well as within. Issues or policies can be characterized by more than their substantivecontent. It is important to recognize that policies can be defined in different ways and their redefinition can affect the costs associated with a member's stand on various alternatives. One persuasive statement of the manner in which policies can be redefined has been made by Cobb and Elder. They maintain that issues can be expanded to involve more individuals and groups within the community to the degree that certain strategies are pursued. These strategies include attempts to define the matter unambiguously, to mark its social significance, to establish its immediate relevance, to avoid technical dis90 cussion, and to demonstrate its dissimilarity with previous policies. Such expansion increases the likelihood that costs will be associated with positions taken on such issues. Clearly, some issues will become more risky for council members than others despite a common content. Conflicts that manage to erupt outside the council, that gain community wide recognition should be more difficult to deal with. An issue then has normative aspects which serve to coalesce structural and behavioral elements in certain fashions. For example, expansion of an issue, means bringing it to different publics, thus activating various cleavages within the political system. This in turn may even change the types of issues that are considered within the council. This facet of the consideration of a policy has serious ramifications for the empirical investigation of the sub j ec't-. It means there might exist no

PAGE 87

68 key by which to identify policies for examination. However, in the exchange perspective, even such changes must be considered in terms of the patterns of habitual response taken by members. Their past behaviors should give a cue to their future responses. 91 Thus I will assume policies can be identified according to their substantive content. Their classification captures generalized sorts of costs and benefits which remain rather constant for the empirical situation. Summary To summarize, the exchange perspective offered here begins with the conception that individual choice is fundamental to a policy focus. Choice, whatever its form, is made within a calculus of costs and benefits whose parameters include a council member's goals and predispositions as well as the choices made by other council members. Choice may engender conflict or promote consensus. In either case, the type of policy that emerges from a city council represents a reflection of the costs involved in collective choice; it turns on the characteristics of the individuals charged with that choice and the manner in which choice situations are typically handled. In succeeding chapters, this study will attempt to demonstrate the importance of such concerns and draw more concretely many of the elements of policy making in city councils.

PAGE 88

69 Notes 1 David Harvey, Social Justice and the City (Baltimore: Johns Hopkins, 1973), p. 22. 2 Roland Liebert, "Municipal Functions, Structures, and Expenditures: A Reanalysis of Recent Research," Social Science Quarterly , Vol. LIV (March, 1974), pp. 765-783. 3 Ira Sharkansky, "The Political Scientist and Policy Analysis," Policy Analysis in Political Science , ed. Ira Sharkansky (Chicago: Markham, 1970), p. 1. 4 Charles 0. Jones, "State and Local Public Policy Analysis: A Review of Progress," prepared for delivery at Conference on Political Science and State and Local Government, American Political Science Association (Biloxi, Miss.: October 26-29, 1972), p. 12. 5 V.O. Key, Southern Politics (New York: Knopf, 1949), p. 308. 6 Ibid ., p. 664. 7 Ibid . , p. 674. 8 Stuart Rakoff and Guenther Schaeffer, "Politics, Policy, and Political Science: Theoretical Alternatives," Politics and Society , Vol. I (November, 1970), p. 53. 9 Richard Dawson and James Robinson, "Inter-Party Competition, Economic Variables and Welfare Policies in the American States," The Journal of Politics , Vol. XXV (November, 1963), p. 265. 10 Rakoff, op. cit ., p. 55. 11 Thomas Dye, Politics, Economics and the Public (Chicago: Rand McNally, 1966), p. 299. 12 John Fenton and Donald Chamberlayne, "The Literature Dealing with the Relationships Between Political Processes, Socioeconomic Conditions and Public Policies in the American States: A Bibliographical Essay," Polity , Vol. I (Spring, 1969), pp. 388-404. 13 Herbert Jacob and Michael Lipsky, "Outputs, Structure, and Power: An Assessment of Changes in the Study of State and Local Politics," The Journal of Politics , Vol. XXX (May, 1968), pp. 510-538. 14 Harold Sprout and Margaret Sprout, "Environmental Factors in the Study of International Politics," The Shaping of Foreign Policy , eds. Harold

PAGE 89

70 bs0 n and William Zimmerman (New York: Atherton, 1969), pp. 43-75. Harold Ja d°Margaret Sprout have argued that the relationship between man and milieu 311 been discussed by scholars in five, more or less, distinct ways: (1) enhaS nmental determinism, (2) free-will environmentalism, (3) environmental pos^hilism (4j cognitive behaviorism, and (5) environmental probabilism. Environmental determinism posits an "invariable correlation" between some set V f environmental "causes" and their environing "effects," (see p. 48). Free•11 environmentalism allows an individual more of a choice, but clearly has environmental influences playing a direct role in his actions. Possibilism avoids the notion of choice and postulates a set of constraining factors that ill affect the outcome of any course of action. In the possibilist world, decisions and motivations are the givens of a situation not to be explained. The cognitive behavior form of hypotheses emphasizes the conception that an individual reacts to his environment as he apprehends it: it explicitly incorporates his past experiences into his future behavior. The distinction is also drawn here between operational and psychological environments. Finally, environmental probabilism names a relationship in which choices may be explained with reference to a hypothetical norm. That is, through an examination of past behavior and the norm established by it, inferences about future decisions can be made. 15 One excellent review of this literature is found in Charles 0. Jones, op. cit. See also Fenton and Chamberlayne, op. cit .; Edmund P. Fowler and Robert L. Lineberry, "Comparative Policy Analysis and the Problem of Reciprocal Causation," Comparative Public Policy: Issues, Theories, and Methods , eds . Craig Liske, William Loehr and John McCamant (New York: Halsted, 1975), pp. 243-259; and Terry Clark, "Community Structure, Decision-Making, Budget Expenditures, and Urban Renewal in 51 American Communities," American Sociological Review , Vol. XXXIII (August, 1968), pp. 576-592. 16 Dye, op. cit ., p. 298. 17 Vernon Van Dyke, Political Science: A Philosophical Analysis (Stanford, Calif.: Stanford University, 1960), pp. 131-157. 18 A discussion of these approaches can be found in John Dearlove, The Politics of Policy in Local Government: The Making and Maintenance of Public Policy in the Royal Boroughs of Kensington and Chelsea (London: Cambridge University, 1973). Also, see John Kirlin and Steven Erie, "The Study of City Governance and Public Policy Making: A Critical Appraisal," Warner Modular Publications (Andover, Mass.: Warner Modular Publications, 1973), pp. 1-12. 19 Edward Banfield and James Q. Wilson, City Politics (Cambridge, Mass.: Harvard University and M. I.T., 1963). 20 Kirlin, op. cit . , p. 4. 21 Ibid . 22 William Riker and Peter Ordeshook, An Introduction to Positive Poli tical Theory (Englewood Cliffs, N.J.: Prentice-Hall, 1973), p. 2. 23 See, for example, Raymond Bauer and Kenneth Gergen, The Study of £°licy_ Formation (New York: Free Press, 1968).

PAGE 90

71 24 David E. Apter, Choice and the Politics of Allocation: A Developmental Theory (New Haven, Conn.: Yale University, 1971), p. 9. 25 David E. Apter, The Politics of Modernization (Chicago: University f Chicago, 1965), p. 39. 26 Heinz Eulau and Kenneth Prewitt, Labyrinths of Democracy: Adaptation s. Linkages, Representation, and Policies in Urban Politics (Indianapolis: B^bbs^Merrill, 1973), p. 474. 27 Heinz Eulau, "Policy Making in American Cities: Comparisons in a Quasi-Longitudinal, Quasi-Experimental Design," General Learning Press Reprint (New York: General Learning, 1971), p. 2. 28 James March and Herbert Simon, Organizations (New York: Wiley, 1958), p. 27. 29 John Crecine, Governmental Problem-Solving: A Computer Simulation of Municipal Budgeting (Chicago: Rand McNally, 1969), p. 99. 30 Ibid . , p. 101. 31 Ibid . , p. 144. 32 Peter Rossi, Richard Berk and Bettye Edison, The Roots of Urban Discontent: Public Policy, Municipal Institutions, and the Ghetto (New York: Wiley, 1974), p. 34. 33 Dearlove, op. cit . , p. 231. 34 Jonathan Turner, The Structure of Sociological Theory (Homewood, 111.: Dorsey, 1974), p. 161. 35 Michael D. Cohen, James G. March and John Olsen, "A Garbage Can Model of Organizational Choice," Administrative Science Quarterly , Vol. XVII (March, 1972), pp. 1-25. 36 Ibid . , p. 1. 37 Ibid . , p. 2. 38 See James S. Coleman, Community Conflict (New York: Free Press, 1957), and William A. Gamson, "Rancorous Conflict in Community Politics," American So ciological Review , Vol. XXXI (February, 1966), pp. 71-80. 39 See Bryan T. Downes, "Issue Conflict, Factionalism, and Consensus in Suburban City Councils," Urban Affairs Quarterly , Vol. IV (June, 1969), pp. 477-497. See also by the same author "Suburban Differentiation and Municipal Policy Choices: A Comparative Analysis of Suburban Political Systems," ComBun ity Structure and Decision-Making Comparative Analyses , ed. Terry Clark (San Francisco: Chandler, 1968), pp. 243-268. 40 Downes, "Issue Conflict, Factionalism, and Consensus in Suburban Ci ty Councils," op. cit ., p. 484.

PAGE 91

72 41 The work coming out of the City Council Research Project is quite 1 The most important pieces published thus far include the following: V8 'i! Eulau and Kenneth Prewitt, op. clt . ; Betty Zisk, Local Interest Poises: Be £p Fay Street (Indianapolis: Bobbs-Merrill, 1968); Robert Eyestone, The ^^T^^ir/P nl.vv: A Study in Policy Leadership (Indianapolis: Bobbsp££ffpf^TTT ^ a lH LvP.r1d B e. Cl tv Managers in-^sla t^LJgli^ (IndianM S-sBobbs-Merrill, 1971); Heinz Eulau and Robert Eyestone, Policy aps Tritv Councils and Policy Outcomes: A Developmental Analysis, APSR, Vol. It torch 1968) pp. 124-143; and Kenneth Prewitt, "Political Ambition, SunSrism and Electoral Accountability," APSR, Vol. LXIV (March, 1970), pp. 5-17. 42 Eulau and Prewitt, op. cit . , p. 609. 43 Ibid . , pp. 611-612. f 44 Ibid . , p. 612. 45 John C. Wahlke, Heinz Eulau, William Buchanan and Leroy Ferguson, Thp J.P-gislative System: Explanations in Legislative Behav ior (New York: Wiliy,"l962). 46 Eulau and Prewitt, op. cit . 47 See Eyestone, op. cit ., or Zisk, op. cit . 48 Edward Banfield, Political Influence (New York: Free Press, 1961). 49 Exchange theory is, of course, merely a broad rubric covering many conceptual variants. Exchange theory flows from the traditional utilitarian assumptions and concepts which viewed men as rationally seeking to maximize their material benefits, or "utility," from transactions or exchanges with others in a free and competitive marketplace. Gradually however, such ideas have been supplemented and augmented to reflect an awareness of the constraints of rationality, the plethora of individual values, and the social dynamic of interpersonal interaction. While it is always risky to generalize about sociological theory, it is probably safe to identify the following themes as common to most contemporary exchange formulations: (1) While men do not seek to maximize profits, they always seek to make some profits in their social transactions with others. . 1 (2) While men are not perfectly rational, they engage m calculations of costs and benefits in social transactions. (3) While men do not have perfect information on all available alternatives, they are usually aware of at least some alternatives which form the basis for assessments of costs and benefits. (4) While there are always some constraints on human activity, men compete with each other in seeking to make a profit in their transactions. (5) While economic transactions in a clearly defined marketplace occur in all societies, they are only a special case of more general exchange relations occurring among individuals in virtually all social contexts (6) While material goals typify exchanges in an economic

PAGE 92

73 marketplace men also exchange other, nonmaterial commodities, such as sentiments and services of various kinds. (See Jonathan Turner, op. cit ., pp. 212-213. 50 James Buchanan and Gordon Tullock, The Calculus of Consent: Logical Fou ndations of Constitutional Democracy (Ann Arbor: University of Michigan, 1962) 51 Turner, op. cit ., pp. 211-294. 52 Ibid . , p. 264 53 Riker and Ordeshook, op. cit . , p. 8. 54 Ibid ., p. 12. 55 Ibid ., p. 14. 56 Anthony Heath, "Review Article: Exchange Theory," British Journal of Political Science , Vol. I (January, 1971), pp. 91-119. 57 Sidney R. Waldman, Foundations of Political Action: An Exchange Theory of Politics (Boston: Little, Brown, 1972), p. 215. 58 Riker and Ordeshook, op. cit . , p. 21. 59 Waldman, op. cit . , p. 217. 60 Norman Frohlich, Joe Oppenheimer and Oran Young, Political Leadership and Collective Goods (Princeton: Princeton University, 1971), p. 124. 61 Riker and Ordeshook, op. cit . , p. 3. 62 Frohlich, Oppenheimer and Young, op. cit . , p. 127. 63 Riker and Ordeshook, op. cit . , p. 256. 64 Buchanan and Tullock, op. cit . , p. 45. 65 Riker and Ordeshook, op. cit . , pp. 272-330. 66 Robert Putnam, The Beliefs of Politicians: Ideology, Conflict, and Democracy in Britain and Italy (New Haven, Conn.: Yale University, 1973), pp. 1-7. 67 Richard Fenno, Jr., Congressmen in Committees (Boston: Little, Brown, 1973), p. 1. 68 Putnam, op . cit . , p. 5. 69 Quoted in Robert Morlan, "Life on the City Council: Realities of Legislative Politics," Capitol, Courthouse and City Hall , ed. Robert Morlan (4th ed.; Boston: Houghton-Mifflin, 1972), p. 212. 70 Charles R. Adrian and Charles Press, "Decision Costs in Coalition Formation," American Political Science Review, Vol. LXII (June, 1968), p. 556.

PAGE 93

74 71 Ibid ., pp. 556-563. 72 James D. Thompson and Arthur R. Tuden, "Strategies, Structures, and processes of Organizational Decision," Comparative Studies in Administration , e d. James D. Thompson (Pittsburgh: University of Pittsburgh, 1959), pp. 195216. 73 March and Simon, op. cit . 74 Eulau and Prewitt, op. cit . 75 Heinz Eulau, "Logics of Rationality in Unanimous Decision-Making," Rational Decisions, Vol. VII, Nomos , ed. Carl Friedrich (New York: Atherton, 1967), pp. 26-54. 76 Ibid . , pp. 26-30. 77 Ibid ., p. 45. 78 Ian Budge, "Consensus Hypotheses and Conflict of Interest: An Attempt at Theory Integration," British Journal of Political Science , Vol. Ill (January, 1973), p. 88. 79 Aage Clausen, How Congressmen Decide: A Policy Focus (New York: St. Martin's, 1973), pp. 12-35. 80 Theodore Lowi, "Four Systems of p olicy, Politics and Choice," published for participating universities by tue Inter-University Case Program, Syracuse, New York: November, 1971, p. 3. 81 Robert Salisbury and John Heinz, "A Theory of Policy Analysis and Some Preliminary Applications," Policy Analysis in Political Science , ed. Ira Sharkansky (Chicago: Markham, 1970), pp. 39-59. 82 Ibid ., p. 3. 83 Ibid . 84 Theodore Lowi, "Decision Making Versus Policy Making: Toward an Antidote for Technology, " Public Administration Review, Vol. XXV (May-June. 1970), p. 31. 85 Oliver P. Williams and Charles R. Adrian, Four Cities: A Study in C omparative Policy Making (Philadelphia: University of Pennsylvania, 1963), P23. 86 Ibid . , pp. 23-32. 87 Ibid ., p. 28. 88 March and Simon, op. cit ., p. 117. 89 Budge, op. cit ., p. 73.

PAGE 94

75 90 Roger Cobb and Charles Elder, Participation in American Politics : fhs Dynamics of Agenda-Building (Boston: Allyn and Bacon, 1972). 91 Clausen, op. cit,

PAGE 95

CHAPTER THREE THE DEMOGRAPHIC APPROACH TO FLORIDA MUNICIPAL POLICY During the last decade, the demographic approach has dominated research on public policy by political scientists. Most often concerned with financial resources, this approach uses a simplified systems model adapted from the early work of David Easton to explain variations in policy priorities among different governmental units. Its components are the now familiar concepts of Easton 's political theory: system, input, output and feedback. But within the demographic approach, these concepts have assumed new forms. Inputs, which in Easton 's work summarize a wealth of activities having their origin in the performance of the system, are typically conceived in more primitive fashion. They become the socioeconomic conditions which frame the context of a particular political system. Outputs, on the other hand, are usually measured by the financial expenditures of the governmental unit under consideration. And the political system itself is usually rendered in terms of its structural characteristics, such as the degree of party government or the type of government. As a rule, most authors writing within the demographic perspective have ignored feedback processes, and, rather, have assumed that outputs are registered implicitly in a political system's socioeconomic conditions. The interworkings of this approach, as depicted in the work of Thomas Dye, are straight-forward. A set of forces in the socioeconomic environment interacts with the political system to produce public policy outcomes. Accordingly, the determinants of public policy are to be found in the nexus of environment and political system. 76

PAGE 96

77 As a heuristic device, such a simple scheme has powerful appeal. It seems applicable to a host of problems and to any number of situations. For the researcher, however, this very generality is the source of many difficulties. The encompassing conceptual nature of the environment has lead many authors to comb census documents, extracting as many indicators of socioeconomic conditions as appears reasonable. Scores of indicators have been selected and collected on comparable units of analysis, usually American cities or states. And for each of the units of analysis, measures of policy output such as levels of municipal expenditure or per capita expenditures have been compiled. In an attempt to capture something of the political system in each of the units of analysis, researchers have gathered that information about political units which is most generally available. At the local level, this has included data on governmental type (whether city manager, mayor or commission form), electoral system (whether partisan or non-partisan), election districts (whether ward or at-large) , and the intensity of competition (turnover or party competition). In short, the approach demands a substantial data gathering effort with attention to a broad range of variables. The connection between variables is specified loosely. Following the lead of fiscal economists who have long attempted to explain the differences in the level of spending among American cities and states, most political scientists working within the demographic approach have resorted to statistical techniques for assistance in sorting out the influences on expenditures. The basic methodology is to apply statistical techniques of correlation and regression to these sets of variables. The policy output measures, the dependent variables, are "explained" to the extent that they are systematically related to political or environmental system variables. The

PAGE 97

78 explanation is a statistical one — it measures the percentage of inter-case variation in the dependent variable that is predicted by variation in the independent variable. Cither statistical manipulations can combine or control for the effect of other variables. 2 Early work using this approach, pioneered in political science by Thomas Dye, appeared to indicate economic and social factors in the environment were more influential in shaping governmental policies than were any of the political variables. Indeed, Dye argued most of the statistical relationships found between political variables and policy outcomes were a result of the pervasive influence of economic development on both the political system and the policies chosen in that system. 3 His research indicated that when political factors were controlled for, economic development continued to have a significant impact upon public policy. But when the effects of economic development were statistically controlled, political factors diminished in their importance for public policy. While there have been extensive criticisms of Dye's work and those who followed his lead, the contribution of the demographic approach should not be blurred by admissions of its difficulties. ^ A number of generalizations about the interrelation of environment, politics and policy can be drawn from the work founded on this conceptual scheme. In almost every instance, research has demonstrated large and dense populations correlate well with general city expenditures and with spending for specific municipal services. Large minority populations are associated with similar expenditure patterns. But it is population size which has the most systematic, independent effect on expenditures. As the size of a city increases, so does its expenditures. And this holds not only for total expenditures, but also for expenditures per capita. ->

PAGE 98

79 Authors such as Brett Hawkins have been persuaded such findings can be interpreted to bear upon the consideration of policy. He suggests that large, dense and heterogeneous socioeconomic environments generate demands for services and that city governments often respond favorably to such demands. 6 While this explanation would appear to neglect the manner in which socioeconomic factors are translated into action demands by a population, it does frame the larger concern standing behind much of the demographic analysis. Namely, which is more important in the determination of public policy, socioeconomic or political factors? Hawkins argues "that the impact of socioeconomic heterogeneity [on policy] is not materially affected by any intervening variable."' In other words, the political system, its structure and processes have little impact on the pattern or level of a municipality's expenditures. Not even the most ardent socioeconomic determinist denies the imporTtance of intervening political system variables altogether. For example, Hawkins writes: a high level of socioeconomic heterogeneity encourages the retention of politicized, responsive, group-accomodating institutions and influences other system properties (e.g., a greater sensitivity to problems on the part of official policy makers that in turn promotes a higher level of expenditures and affects the distribution of policy benefits)." Moreover, not all studies fashioned from the demographic perspective have rejected the importance of political factors on policy. Political factors have been found important in the adaptation of floridation proposals, in the responsiveness by cities to the demands of citizens, and even, on occasion, in spending priorities. 9 In short, the demographic perspective has not single-mindedly supported the impact of socioeconomic factors.

PAGE 99

80 It remains to be demonstrated whether Florida cities spend their revenues in a way which highlights their poltical institutions and practices or whether socioeconomic factors will prove more important. In either event, a systematic investigation of the perspective will allow for a determination of the significant factors bearing on public policies in Florida cities. It will facilitate an exposition of the worth of the demographic approach. Moreover, the results will condition further examination of policy as it is promulgated by municipal councils. These results can be incorporated into a research design in which council influence on policy can be studied. In the next few pages, I will specify the units, variables and linkages portrayed in Figure 3-1. [ENVIRONMENT] SES Factors [POLITICAL SYSTEM] Structural Aspects of City Government [POLICY] Expenditures Figure 3-1. The Demographic Perspective: Macro-phase Indicators The Units of Analysis The demographic model has had wide currency in the study of municipal expenditures, but its use has engendered an extended debate. Much criticism surrounding its application has turned on methodological difficulties. For example, it has been argued the reliance on correlations rather than regression coefficients, the severe problems of multicollinearity, and the cross-sectional nature of most studies, all act to weaken its

PAGE 100

81 findings. 10 However, many of the problems most frequently cited can be addressed by a careful choice of the units of analysis. In choosing only Florida cities as the site of my investigation, many of the confounding factors present in any comparative research project are minimized. The municipalities are subject to a common body of state law, and state requirements are relatively similar for all cities throughout the state. Moreover, to the degree urban environments vary greatly from state to state as the literature on political culture and state politics indicates, inter-state variations will be controlled. J1 There are additional advantages to conducting the research within a single state. In the demographic model, policy outputs are measured in terms of fiscal indicators. Research involving expenditures in the municipalities of a single state is unlikely to suffer from the same weaknesses as comparative studies involving cities in different states or regions. Costs should be similar across the sample of cities. Comparable services will require similar expenditures because the costs of living will be relatively uniform. And there is yet another advantage. Within Florida, municipal expenditures are standardized. In inter-state comparisons, expenditures are derived from aggregate figures compiled by different agencies using different procedures. Bookkeeping decisions vary; transfers are handled differently. This is not true in Florida where the Commission on Local Government, created by the Florida legislature in 1972, developed a questionnaire which formed the basis of Florida's Fiscal Reporting, and Uniform Accounting and Budgeting System. All the fiscal measures used in this study have been gathered and reported by municipalities on the basis of a common questionnaire.

PAGE 101

82 Detailed expenditure information is available on each municipality within the state, but it may not be appropriate to consider each one. In the state Florida, there are 383 municipalities. They range in population from the nearly 600,000 inhabitants of Jacksonville to the four residents of Marineland. Jacksonville covers 842 square miles, while Marineland occupies less than one square mile. 13 This range of variation is repeated on other variables and between other cities. It is not diversity itself that makes the comparison of cities as disparate as Jacksonville and Marineland problematic. Rather it is the realization that the process of urbanization introduces complexities into a community which may change its character (see Appendix B) . There is little reason to expect the effects of urbanization to be linear. In fact, the effects may be associated with thresholds; complexity may accrue before its effects are experienced. In many states, and in Florida in particular, municipalities have not been treated equally by the state legislature. The Constitution of 1885 under which Florida operated for more than 80 years extended broad powers to the legislature over local municipalities. ^ Under the guise of enactments, known as "general laws of local application" the legislature was able to specify quite clearly the cities to which a law might pertain. They did this by classifying cities on the basis of their population. Thus only cities of a certain size might be affected by a law. And while the Florida Constitution of 1968 did much to undermine the pervasive influence of such enactments on the authority of local governments, the effects of the old population acts remain in those cities which have not availed themselves of the home rule provisions in the new constitution. *-> Only three of the population groupings used in the 1972, Florida

PAGE 102

83 Census of Local Government Finances were considered in this analysis. They are intended to capture a degree of complexity in the socioeconomic environment and to control for the effects of Florida's traditional population laws. Seventy-eight cities were selected. According to the July 1, 1972, estimates of population made by the Florida Department of Administration, Florida cities formed five different population types. In this study, city groups with populations greater than 10,000 were selected: 1) 10,001-25,000; 2) 25,001-50,000 and 3) 50,001 and over. A list of the cities can be found in Table 3-1 along with the counties in which each city is located. The Dependent Variable The outputs of governmental institutions take many forms, but within the demographic model municipal allocations have most often been operationalized as financial measures. While such indicators do great violence to the richness of governmental action and to the actual performance levels that may be had with similar amounts of money, fiscal measures do . provide meaningful profiles of governmental priorities. Further, as Terry Clark has written, fiscal indicators have the following advantages: 1. They are based on a numeraire constant and comparable across policy categories, and, with corrections, cities and t ime . 2. They are cardinal measures, alleviating problems of establishing zero-points and comparable points. 3. Given current weaknesses of performance indicators and policy impacts in most cities of the world, fiscal measures remain, however temporarily, probably the least unsatisfactory general measure of governmental effort in different areas. 4. The fiscal level of governmental activity, at least for certain social groups, is itself a basic ideological concern. Left-right characterizations are hard to apply to mass populations, but such a fiscal dimension is a central component of most such ideologies.

PAGE 103

84 c o co u o J c 3 O CJ OJ QJ D x: cj CO CU PQ p a a o si TO 3 x: u to V B.H 3:0 O td to Oh pQ c u x: 00 3 o in O " X) ia a ti fc E t0 to 3 3 o c >-. u PQ pq TO CJ TO x: to m a ij oh tO tO <—l r-t 3 E > CJ CO 1-. 4-1 -H Oh PQ CO ex, O O c m •H to 6 i-t CJ to CO CO 30 3 O o X) 1-1 CJ 01 00 c to O M QMfJaiiUHp.faa.o CJ e to 2 u e e CO CO co 00 c H l-i D. CO U x: u to to E C to cj co x: a. 3 to OJ x: o x: to a CJ to PQ CJ PQ ^i H U a b to tO f— ( CL, •H to Z PL, T3 C x x m *J 4-1 T— ( u u ^ o o .-o n 2: o (0 J* a o -J O TO n c a o CO t-H E q. u n 000 x: -u a, to (X, c c to U -H H Oh Oh PL, x: CJ CO CJ PQ O c CO E > O -H Oh 05 (/) 1-1 3 CJ CO j_i 3 a < O, TO o cm c CO e CO 2 CJ CO CO CO CO CO CO H CO C E m E-i a) x: •-1 cj i-H CO •H CJ > PQ x: o CO cu PQ C cj ^ > M to CO -1 £ s: e-i CJ CJ 3 O 4-1 4-1 4J 4j 3 3 3 CJ TO c u sSsisSisiilSSSggsgingssssisssgsKErscs^s? OOOOOOOOOOOO c c o o x: x: u o to to aj a PQ PQ S E r-\ i-l CO CO Oh Oh CD W CJ E •o >H CO CI l-l CO > oi o to TO )-l CO 3 o J3 a CO OJ pq O 1-1 O CO -H 1h 2^^fiQ> u „^^« co a cj ^H to C J"i i-H v-l O < Oh OJ i-H O CO 3 M a > c to 3 o 005: to CJ pq E ^ Oh PQ pq to -a >-i 1-1 CO CO 3 > O CJ u cq PQ pq 0) c -o O CO 4-1 —I CO O 3 Oh O CJ U CO i-l 1< U H CO O CJ PQ pq pq O Xi CJ CJ CO a. to co to « cj cj u CO XI CJ to pq o CJ x: to CJ QJ co pq HI PQ TO CO CJ C -H O C-H 4J IH o o O c_> to a a o s CJ r-1 x: CJ CO CJ pq c >, -H a -a p QJ — < c 0) 3 « Q QJ TO 3 .-J x: o to QJ PQ C O 1 CO _CJ -H cd > Oh S 0) QJ i-H 4-1 CO 1-1 TO o c a. co tO 3 to 0;3 ?\ CO •-H CJ 35 3. X -o CJ QJ I H Bl CJ) Uh ^ T3 C CO i-H O J! en cu CO o a CO X) co co -J .J QJ C 4J 3 tO O GO jo n -I co Q) TJ o i— ( cm m i§§I§Iii!l§33gg33=3§Sgg g §ggggggg e gggggg

PAGE 104

85 5. Costs of government are important to minimize insofar as the basic goal is to provide citizens with abundant but inexpensive services. Precise information is essential if tradeoffs across categories are to be considered. 1 ' The difficulties of relying upon expenditure variables as measures of policy have been commented on by numerous authors. ° However, many of the troublesome aspects of such indicators are eliminated in this study. As I have already mentioned, the Commission on Local Government, created in 1972, initiated a fiscal reporting and uniform accounting and budgeting system for the minicipalities within the state. This system consolidates and standardizes the manner in which local governments throughout the state render a detailed account of revenues, expenditures and transfer payments. The questionnaire issued by the Commission solicits information in a highly explicit fashion and provides a system to eliminate duplicate accounting of interfund transfers with regard to revenues and expenditures. 19 In addition, for the 1972 state census, the state of Florida provided considerable assistance to each of the cities returning the questionnaire. As a result the data gathered in the questionnarie is the most detailed and most accurate information ever obtained on the fiscal activities of Florida municipalities — more accurate than the cen20 sus of government figures reported by the National Bureau of the Census. The Florida census of local government finances divides expenditures by municipalities in 32 categories. In the analysis presented in this chapter, each of these categories incorporates current operating expenses, capital outlays, and transfers to other governmental units. A list of the categories is provided in Table 3-2 . As with most demographic analyses involving such fiscal measures, municipal expenditures will be examined from several different perspectives. Total expenditures, per capita expenditures, and the distribution of expenditures across categories will be used to portray the outputs of each city.

PAGE 105

86 Table 3-2 Variables Contained in the Fiscal Reporting and Uniform Accounting and Budgeting System for the State of Florida Variable Mnemonic Variable FIN ADM CENCON GENBLD GOVTOT COURT POLICE CORR FIRE INSP SAFTOT WELF HEALTH HOSP URNRWL STREET PARKS NATRES LIB PARK AIR PORT REFUSE SEWAGE WATER ELEC GAS TRNSIT INTRST RETDBT 0TH1 OTH2 0TH3 Financial Administration Expenditures Central Control Expenditures General Public Buildings Expenditures Total General Government Expenditures Court Expenditures Police Activity Expenditures Correction and Detention Expenditures Fire Prevention Expenditures Protective Inspection Expenditures Total Safety Related Expenditures Public Welfare Expenditures Health Related Expenditures Hospital Expenditures Urban Renewal Expenditures Street and Highway Expenditures Park and Recreation Expenditures Natural Resource Expenditures Library Expenditures Parking Facility Expenditures Airport Facility Expenditures Port and Terminal Facility Expenditures Refuse Collection Expenditures Sewage Disposal Expenditures Water Service Expenditures Electricity Service Expenditures Gas Service Expenditures Transit System Expenditures Interest on Debt Retirement of Debt Miscellaneous Expenditures, category 1 Miscellaneous Expenditures, category 2 Miscellaneous Expenditures, category 3

PAGE 106

87 The Independent Variables The demographic approach assumes environmental variables influence political system variables and policy outputs. However, the conceptual status of the "environment" is quite vague within this approach. This poses special problems of operationalization. Most researchers, working within this perspective, have selected various socioeconomic variables for inclusion in their analysis. Typically, community characteristics have been captured in such socioeconomic indicators as total population, median income, median school years, percent foreign born, percent non-white, percent of families with income under §3000 and percent unemployed. 21 Yet the rationale behind the inclusion of these indicators, rather than others, has seldom been made explicit. The selection of environmental variables can proceed in two ways. First, it is possible to rigorously specify those components of the environment which theoretically represent inputs. This approach entails a retooling of the basic demographic model. Such a reworking would require much more specific hypotheses and a more detailed appreciation for the impacts of specific variables than offered in most demographic models. A second approach takes an inductive tact, and has many different forms. The most common of these is called the factor-analytic approach. This method typically involves starting with a large number of census variables for a number of cities and factor analyzing them (usually employing an orthogonal rotation) to produce "latent" dimensions. These dimensions, are then taken to represent the "environment" of the municipality. This approach suffers from several defects which have been amply documented in the literature. But perhaps the most worrisome

PAGE 107

88 difficulty involves the manner in which the original set of variables is selected. Since the final solution of the factor analysis is ineluctably tied to the variance and the associations of the variables included in the analysis, the selection of the variables becomes crucial to which environment" this inductive approach finds. The advantage of the approach rests in its ability to allow the researcher a margin of error in selecting the initial set of variables. But in the final analysis, this approach by itself is hardly a significant advance over arbitrary a priori inclusion of variables. There is a second inductive approach. This involves a detailed search of the literature on community characteristics in an attempt to define some consensus among observers on the types of variables that best capture the dimensions of urban places. Fortunately, several such studies have already been undertaken. In one study of the critical dimensions of community structure, John Tropman concluded four broad classes of variables have been shown to be important for describing the urban scene. These are 1) the size of the community, 2) the racial composition of the community, 3) the socioeconomic composition of the city, 4) the maturity of the city's growth. 23 Philip Rees has substantiated and refined this judgment in an extremely useful article. 2 ^ Rees' extensive analysis of the variables used in factorial ecology yielded the classification reproduced in Table 3-3. Rees' categories explicitly incorporate the findings of two generations of human ecologists who have ignored the physical environment , and who have accepted Park's assumption that a focus on aggregate community characteristics makes sense because, within a community, organizational structure represents

PAGE 108

89 Table 3-3 Classification of Variables Employed in Factorial Ecology 1. Socioeconomic Status Variables Population Variables (direct indicators of social status) Education Occupation Income Housing Variables (indirect indicators of social status) Quality Value of Rent Household Material Possessions Mixed Population and Housing Variables (for instance, the degree of overcrowding) 2. Family Status or Life Cycle Stage Variables Population Variables (direct indicators of family status — the Life Cycle Subset Age Family Fertility Marital Status Housing Variables (indirect indicators of family status) — the Urbanization Subset Type Age 3. Ethnicity or Minority Group Status Variables Racial Group Nativity Group National Group Linguistic Group Regional (Migrant) Group 4. Change and Mobility Variables Mobility Movement Rates Movement Classified by Origin or Destination Population Change 5. Scale Variables Population Area Population Density (may act as indirect family status indicator) Locational Measures (may act as indirect family status indicator, e.g., distance from city center) 6. Health, Welfare, and Social Problems Mental Health Physical Health Welfare Crime and Delinquency School Population Statistics (These are usually local government statistics) Source : Philip Rees, "Problems of Classifying Suburbs within Cities," City Classification Handbook: Methods and Applications , Ed. Brian Berry. New York: Wiley, 1972, p . 284.

PAGE 109

yu a successful adaptation to the demands of competing social environments . 2 ^ In featuring the socioeconomic aspects of the environment, factorial ecologists may have made serious conceptual mistakes, but their work falls squarely within the domain of the demographic model. The environmentally generated demands embodied in the demographic approach have a socioeconomic cast very similar to that used in factorial ecology. 25 In short, Rees' categorization is an apt beginning for the selection of environmental variables. In choosing the variables to be used to tap environmental demands, I originally gathered 103 variables for each of the 78 cities included in the study. Rees 1 categories were used to guide the variables selection process. However, variables were also included because of their widespread use in similar analyses. 27 For reasons discussed in Appendix C, the initial list of variables was pared to 73, forming the set of independent variables to be used in this study. This list, together with the original variables, can be found in Appendix C. 28 The Intervening Variables The demographic approach asserts the political system acts to transform demands generated in the environment into public policy. Yet the thrust of much of the work done within the demographic perspective is that "political structure and process have little independent explanatory power; they act as 'neutral transmission belts' converting resource capabilities into budgetary allocations." 29 Nonetheless, a significant proportion of the work done by political scientists dealing with public policy has been animated by the desire to demonstrate the influence of political factors on policy outputs. But in community studies relying on the demographic approach, this desire has been thwarted by a noticeable

PAGE 110

91 "lack of synoptic indicators of conversion processes. "30 Most researchers have focused on the governmental forms associated with reform or nonreformed governments. The reason for this focus is clear; it is only with respect to structural variables that data can be gathered conveniently on a large number of cases. While purely structural variables, such as partisan as opposed to non-partisan elections, may seem generally insensitive to great or sudden variations in policy expenditures, or to changes in the socioeconomic environment, such indicators have been shown to have systematic relation to 01 policy and demographic characteristics in several studies. Thus, in my examination of the expenditures in 78 Florida municipalities, I have included the most common measures of reform government as suggested in the Model City Charter of the National Municipal League.-^ 2 These measures are listed in Table 3-4 , In addition to these structural measures of the political system, three other indicators of the political conversion process were included in the analysis. The first of these is the community tenure of the city manager within each of the council-manager governments. Previous studies of manager tenure have shown this variable to be a potent indicator of the politics within the community. 33 Second, the city-wide turnover in council members is used to gauge the amount of competition for the city council in each of the 78 cities. Competition, even in nonpartisan municipalities^ has been shown to be a relatively reliable measure of the expenditures for certain types of services. 34 For example Eulau and Prewitt have shown electoral competition is related to the resource capabilities of a city;35 while Jerry Yeric has suggested that as a council member's tenure increases, his willingness to spend public revenues decreases. 36 The

PAGE 111

92 Table 3-4 Variables Relating to City Political Structure Variable Mnemonic Variable COMPET TURN 1NC0RP FORM ELECT MAYADM CNCLNO WARDS TERM PARTY Council Turnover (Terras /council member) City Manager Turnover (Years/manager) Years since Incorporation Form of Government Council-Mayor Commission-Mayor Council-Manager Commission-Manager Method of Election for Office of Mayor Popular Vote Commission Council Mayor Exercises Administrative Control Over City Operations Number of Council Members Method of Electing Council Members At Large Wards (Districts or Mixed) Other Length of Council Term (if split 2/4, then 2) Partisan/Nonpartisan Elections

PAGE 112

93 final structural measure is the number of years since the city was first incorporated by the state of Florida. Newly incorporated municipalities not only have new sets of political institutions, but are likely to still be in the process of developing the informal political mechanisms which impact upon public policy. Additionally, the newness of a city is quite likely to affect the nature of the problems considered most pressing in a city. These problems may well be independent of the socioeconomic context of the municipality, but may be associated with furnishing a range of municipal services in many areas for the first time. Hence, a city's incorporated age is very likely to have a relationship with expenditures. Specifying the Linkages The wealth of social science research employing the demographic approach is enormous. Thus it is difficult to summarize succinctly the specific relationships between variables established by numerous scholars. Indeed, much of the controversy over the findings has limited itself to commentary on the distinction between socioeconomic and political influences on public policy. The theoretical conceptualization of the possible relationships between variables has not proceeded apace of the concern for the general determinants of policy. Nonetheless, a number of generalizations can be rescued from the limited evidence presented in the literature. Following the argument presented by Aiken and Alford, there are five conceptualizations of community policy-making dynamics and their bearing on policy outputs which have dominated research within the demographic approach. Respectively, the foci of the five are 1) political culture or ethos, 2) concentration or diffusion of community power, 3) centralization of formal political structure, 4) community differentiation and

PAGE 113

94 and continuity, and 5) community integration. 37 Each of these foci have been used to tender hypotheses linking the measures already presented. I will sketch several of these hypotheses in the pages which follow. The values and beliefs held by citizens regarding the proper scope of governmental activity bias the types of policies pursued in the community. This thesis holds political culture as determinant of local policy. "The thesis is that voters holding 'public-regarding values' will be more supportive of policies which benefit the community as a whole than voters who hold 'private-regarding values! ' . . . Ethnicity, Protestantism, and income levels have been used as indicators of the degree to which there are 'public-regarding values. '"38 Protestants, so the argument goes, are imbued with an "ethos of personal responsibility,"™ in which the problems and morals of others in the community could be subsumed. Immigrants, on the other hand, are anxious for the benefits the political system could provide; their focus does not extend to the entire community. Thus, according to this view, there is the expectation that cities rating low on ethnicity variables, but high on SES variables such as income are likely to be public-regarding, and hence will evidence higher expenditures in functional areas relating to city wide amenities and personal social services such as housing, urban renewal, public welfare, and health care. Those scholars who have maintained that the concentration or diffusion of a community's power structure affects public policy have taken several different theoretical tacts, "but they share a common assumption that the greater the degree to which power is concentrated, the greater the degree of innovation."^ The belief standing behind this proposition suggests it is only when power is concentrated that the considerable obstacles standing in the path of expenditures for such nontraditional ser-

PAGE 114

95 vices as urban renewal or public welfare can be overcome. Testing this hypothesis in a large number of cities has been quite difficult; referents of power concentration are dependent on a number of additional assumptions-, if census-type information is to be used. Typically, researchers have assumed power is most concentrated where the proportion of individuals in white collar occupations is high, or the ratio of persons with incomes greater than $15,000 to those with incomes less than $3000 is high. The identification of such indicators is based on the notion that power is likely concentrated where there were also concentrations of professionals or wealthy individuals. Naturally, such an assumption need not hold. The focus on the centralization of power is similar to the one which features the centralization of formal political structures. Local governments which invest considerable administrative power in a central figure are thought to provide the conditions which promote innovative policy outputs. The indicators of centralization, it is expected, will monitor cities with relatively large expenditures in areas relating to the collective welfare. For most researchers, centralized local government means city-manager government, which is part of the reform package of public values, favored by groups with publicregarding values. The supposition is such institutions are "the political instruments of such values [that] should produce consequences similar to those of sheer demographic composition."^That is, city-manager governments should evidence higher nontraditional expenditures. Aiken and Alford suggest community differentiation and continuity best distinguish those cities with a high commitment to innovation from those cities which do not. ^ Large cities and cities of manifest maturity, it is argued, are likely to be more differentiated. This complexity is

PAGE 115

96 liable to establish the conditions and generate the demands for expenditures in nontraditional functional areas. Older cities are quite likely to have provided for most of the traditional caretaker services which citizens demand, and to have established the organizational climate in which new expenditures can take place. ^ Thus, scale variables, together with the age of the city, may be expected to correlate well with indicators of expenditures in functional areas relating to municipal amenities and personal social services. Several years ago, James Coleman argued that community controversy arose out of the cleavages created by the differences in economic classes, heterogeneous values, and dramatic population shifts which characterize a municipality.^ Such controversy, he wrote, could paralyze a community's policy making, forestalling expenditures on anything but caretaker services. While the sophistication of his argument goes beyond the use of expenditures as indicators of policy, it can be supposed communities where integration is high are more likely to engage in fiscal allocations favoring the collective welfare. Integration is operationalized using censustype data by simplifying the concept considerably. It has been taken to mean homogeneity and staticity. Hence, small populations, low ethnicity, low mobility, and small rates of change, together with low unemployment, are often used asindicators of community integration. 4 1 Demographic analysis has highlighted propositions such as those above in its sweeps through enormous arrays of comparative statistics on American cities. As is obvious, the operationalizations are often crude. Some variables, such as ethnicity, find their place in several of the hypotheses and often point toward different expenditure patterns. For example, the referents for community differentiation and community integra-

PAGE 116

97 tion can be relatively similar, yet the direction of the association with expenditure variables will be reversed. Of course, the most troublesome aspect of this observation is not that alternative hypotheses have been offered in the literature, but rather that they have each found some confirmation.^ 6 But there are reasons why this might happen. An investigation of the expenditure relations obtaining among the 78 Florida cities will illustrate the problems. Methods and Analysis The analysis presented here tests for systematic linkages between the independent, intervening, and dependent variables as specified by each of the conceptualizations presented above. Initially, correlation analysis will be used. And while the 78 cities in the study constitute the universe of all incorporated municipalities with populations of 10,000 or more, significance tests will be employed as a "rule-of-thumb measure of the real interest, substantive significance,"^ 7 of the relationship. Thus only associations which are greater than those which might occur by chance will be considered.^ 8 Aiken and Alford suggest "in the absence of any other systematic criterion to establish the strength of a relationship, statistical tests of significance can be used. "49 Correlational analysis has many disadvantages, not the least of which is the considerable difficulty attendant to apprehending large arrays. If the many variables involved in this study were to be presented, three matrices with 116 by 32 cells would have to be reported. Thus rather than reporting the results, I will offer some general observations on the nature of the findings , and then comment on each of the conceptual linkages. ->1 First, when associations less than .19 are deleted from the matrices,

PAGE 117

98 the tables are surprisingly bare. There are few associations between any of the variables which are significant. Twenty-two percent of the cells of the matrix concerned with expenditures by functional area are significant; about seven percent of the cells displaying the relationship involving expenditures by function as a proportion of the total expenditure in a city are significant; and fourteen percent of the per capita expenditure relationships are significant. Second, most of the coefficients are low. This is especially true of the coefficients relating per capita expenditure measures with the independent and intervening variables. The major exceptions to this rule come in variables detailing the scale of the city and those measuring the degree of ethnicity within the city. Third, there are some similarities among relationships across the tables of coefficients. That is specific independent and intervening variables tend to display associations with the same variables, regardless of the expenditure measure used. Again these variables tend to capture scale factors of the city or its subpopulations. Moreover, structural variables such as the number of council members, the form of local government, and the length of member terms also display similar patterns of association across the tables. Fourth, and perhaps the most evident, the tables are extremely difficult to interpret and digest. Even in their reduced form, the matrices present an overwhelming array of coefficients, over 4000 in all. Comprehending their meaning is extemely difficult. Nonetheless, the coefficients, taken in groups, can act to illuminate the various linkages already tendered. Political culture, it was argued, should affect the policies empha-

PAGE 118

99 sized in a municipality. In particular, publicregarding values signaled by the presence of low ethnicity and favorable socioeconomic conditions should be correlated with expenditures on such items as public welfare, public health, and urban renewal. For most indicators of political culture, such expectations proved false. Only total Negro population and number of Negro owner occupied housing correlate significantly with collectively oriented expenditures. None of the SES variables such as median family income or percent white collar were associated signifircantly with expenditure variables. Thus only those ethnicity measures which tap the absolute magnitude of the presence of ethnic groups within a city relate highly to expenditures, and these do so in a fashion quite opposite to that predicted by the culture hypothesis. The overall results appear insufficient to sustain the argument for the impact of political culture as drawn previously. Of course, culture may have a considerable impact; but the conceptualization offered above, based on the experience of the American northeast, does not appear to be transferable to Florida cities. This conclusion is not surprising. Florida cities are probably much different than their northern counterparts. At any rate, their growth rates are quite dissimilar; Florida municipalities have experienced dramatic growth in the post-war period. Indeed, it is extremely difficult to identify Florida cities in the following description of the difference in types of community ethos described by Richard Hofstadter: Out of the clash between the needs of immigrants and the sentiments of the natives there emerged two thoroughly different systems of political ethics. .. .One, founded upon the indigenous Yankee-Protestant political traditions and upon middle-class life, assumed and demanded the constant, disinterested activity of the citizen in public affairs, argued that political life ought to be run, to a greater degree than it was, in accordance with

PAGE 119

100 general principles and abstract laws apart from and superior to personal needs, and expressed a common feeling that government should be in good part an effort to moralize the lives of individuals while economic life should be intimately related to the stimulation and development of individual character. The other system, founded upon the European background of the immigrants, upon their unfamiliarity with authority, and upon the urgent needs that so often grew out of migration, took for granted that the political life of the individual would arise out of family needs, interpreted political and civic relations chiefly in terms of personal obligations, and placed strong personal loyalties above allegiloyj Is. 5 ' ance to abstract codes, law or mora] In Florida, the immigrants are themselves "Yankees" intruding on the traditional politics. In short, it does not appear as though ethnic cleavages should develop in the manner suggested by Banfield and Wilson. The hypothesis offered earlier, linking policy outputs to the concentration of community power, is given little support in coefficients between indicators. None of the referents chosen to test this hypothesis manifested significant associations with expenditures in non-traditional service areas such as urban renewal. While it might be argued the set of measures only poorly captures the concentration or diffusion of power in a community, the results do differ from those reported by Hawley and oth53 erswith reference to measures of education and median family income. Neither the percentage of individuals in white collar occupations or the ratio of individuals making $15,000 to those making $3000 are related to expenditures in the predicted fashion. Thus, although the conclusion must be cautiously drawn and tempered with a realization of the inadequacy of the empirical referents of the concepts, it appears there is no relationship between the concentration of power within a community and its policy outputs. Can political structure affect policy expenditures? Yes, if it is conducive to administrative arrangements which allow community opposition to expenditures in areas such as public welfare to be overcome. Reform

PAGE 120

101 governments, it is argued, manifest such arrangements. But this argument is supported only tentatively by the data. Per capita expenditures for items such as urban renewal are not associated significantly with most indicators of reform political structures. Only the length of a council member's term, the number of council members, and the form of government evidence significant relationships with the functional expenditure measures, Among these relationships, the most significant relationship (between health expenditures and the number of council members) accounts for only sixteen percent of the total variance. Political structural variables work better with measures of the distribution of expenditures in a city. Several measures correlate significantly with the hypothesized expenditure categories. Systematic variation shows up between the proportion of the total budget spent on non-traditional services and the following: the type of governmental form, the number of council members, the presence of at-large districts, and the length of the council member's term. However, the correlations are low and not consistently related to any of the non-traditional expenditure areas. Moreover, the direction of the various coefficients varies among the political structure measures. There is a negative relation between several expenditure areas and the number of council members, while the length of term correlates positively. This observation is symptomatic of the patterns of association. They are neither systematic nor large. Although the hypothesis concerning the centralization of political power may have some merit, the evidence produced here can not sustain it. Do cities which have had many years in which to grow and expand spend more for certain amenities and social services? The community differentiation and continuity hypothesis suggests they do. And in truth, several

PAGE 121

102 measures of continuity and differentiation do demonstrate the posited association. In particular, total population, total Negro population, and total Negro owner occupied housing consistently correlate with expenditures on public health care, public welfare, and urban renewal. The age of the city does less well as a covariant, but nonetheless does correlate significantly with per capita expenditures and the distribution of expenditures on items such as public welfare and urban renewal. The cluster of indicators used as surrogates for community differentiation and integration demonstrates a higher and more consistent set of coefficients across expenditure measures than any other group. But not all elements of the cluster perform equally well. Population density, for example, is not correlated significantly with measures of nontraditional service concerns. The key set of variables appears to be those indicative of the total population of the Black subpopulation. This finding is in keeping with the general importance given such variables in previous studies. ^^ The underlying assumption of the community integration hypothesis is that community conflict creates the circumstance whereby policy making may be paralyzed. As the argument proceeds, community homogeneity nurtures what conflict disrupts. Thus, in one form of the argument, homogeneity can be expected to favor expenditures on amenities and personal services. ' The data does little to sustain this view. Of the 16 measures selected as indicative of integrative communities, only three are significantly related to the predicted expenditures. Two of these — total negro population and total number of Negro owner occupied houses — were shown to be in a direction opposite to that hypothesized. Thus the integration hypothesis, as crudely measured by census-type variables does not find much support in the expenditure patterns of Florida cities.

PAGE 122

10 3 A Commentary on the Findings In the last few pages, I have played out the manner in which certain conceptual linkages can be investigated within the broad cast of the demographic perspective. Correlational analysis has lent support to one set of relationships. Namely, indicators of community differentiation and continuity bear a consistent relation to expenditures on nontraditional services. While these findings should be accepted cautiously in the absence of further analysis, it should be noted similar results dominate', much of the literature using the demographic perspective. But the similarity rests in the relationships observed between similar sets of indicators. Each indicator of differentiation and continuity did not perform equally well. It would be incredibly intemperate to conclude the conceptual linkages had been substantiated. This is the very problem with most presentations of the demographic approach. Its conceptual foundations are weak. Merely demonstrating a correlational connection between census-type variables is the coarsest form of data "bare footing," and is of little help in framing an explanation of why Florida cities exhibit certain spending habits. However, it is just this correlational connection the demographic approach yields. The approach thrashes correlation matrices in the hope a structure can be discerned. Yet, this is not the tone most researchers have adopted. As indicated earlier, the demographic approach has, more often than not, lapsed into a debate over the determinants of policy. Scholars have searched for sets of relationships among indicators and have been willing to accept patterns of coefficients as indicative of a causal temper. But this debate was over the influence of factors, broadly conceived, poli/

PAGE 123

104 tical versus socioeconomic. Once analysis attempted to draw the relationship between expenditures and certain causal factors more finely, the entire enterprise outgrew its basis. Shefter cut to the heart of the difficulty in his critique of Brett Hawkins' presentation of the demographic approach in Politics and Urban Policies : But the fundamental source of Hawkins' difficulties, one suspects, lies with the model of the political system which informs his analysis. The set of concepts which this model provides him, and the questions about the policy-making process which this model suggests to him, lead him to draw only the most abstract and uninformative conclusions from the existing literature in the field. The demographic approach cannot sustain the elaborate arguments which have been constructed from its use. The problem is twofold. First, the hypotheses flowing from the demographic perspective have been drawn too broadly and with little sense CO of process. The logical links between the concepts in the general hypotheses are often ad hoc. The mechanisms connecting them are not apparent. The second difficulty stems from the first. Namely, the operational indicators of the concepts involved in the hypotheses are often tenuous. Census variables have been adapted in the most off-hand manner. Nonetheless, the results of demographic analysis are interesting. They afford the researcher a keen sense of the interplay of certain indicators. Unfortunately, such results only muddle the status of conceptual linkages since multiple indicators have been used to portray each concept, sometimes with contradictory results. The lesson here is clear. It makes little sense to force meaning upon the relationships established between indicators. Rather, a perspective must be adopted which does not require the broad imputation of influences on the promulgation of public policy, yet which makes use of the information developed in the demographic approach.

PAGE 124

105 There are many different approaches to this problem. However, I believe the distinctions drawn by Margaret and Harold Sprout provide an 59 especially useful perspective on the matter. They argue the environment-system relationship can be thought of in several different ways and that each way has ramifications for the type of explanation which ensues. In particular, they suggest elements in the environment need not have a direct, conceptual relation to elements in the system. Rather, the milieu can be "conceived as a sort of matrix which limits the operational results of whatever is attempted." In this sense, environments are understood as constraints which mold and limit the form public policy assumes — they do not necessarily determine it. This is the possibilist conceptualization, and it bears directly on the type of analysis undertaken by a researcher. In the possibilist hypothesis, environmental limitations on accomplishment are assumed to be discoverable. But methodological discussion of possibilism is rarely explicit as to how one goes about discovering them. One may assume that a hypothetical omniscient observer could identify and delineate all the interrelations between milieu and environed unit. But, of course, no observer is in fact omniscient. The most that one can do is to frame hypotheses as to what environmental factors are significantly relevant to the action under consideration and how these set limits to the operational results thereof. Environmental possibilism, I believe, moves the efforts of those laboring within the demographic approach more squarely in the direction of significant research on comparative urban phenomena. For some time, scholars have been investigating the manner in which communities differ, and the way in which community characteristics act to constrain policy outputs. ^ 2 As early as 1943, Chauncy Harris published an article entitled, "A Functional Classification of Cities in the United States," which attempted to classify cities on the basis of their primary modes

PAGE 125

106 of economic activity. This effort was predicated on the notion comraunities could be differentiated by the "functions" they performed. This research, like that of the factorial ecologists, implicitly assumes community differences can be captured and registered in the attentive observation of a few characteristics. The taxonomies of factorial ecology are meshes designed to separate cities into relatively homogeneous groupings. The differentiating criteria are assumed to act as constraints on the pattern of relationships internal to each group. It is this underlying assumption, crucial to comparative research, that differences can be isolated and that they will provide meaningful distinctions, which embraces the link between environmental possibilism and the study of public policy. In this study, I will not deal with the many ramifications of the possibilist hypothesis. However, I will assume the actions of policy makers may be conditioned by their milieu; the environment can set limits on their actions. City council members, I believe, will be affected by the socioeconomic conditions which obtain in their constituencies. While I will not investigate the direct cognitive links necessary to establish the validity of the possibilist hypothesis, I will be guided by the concerns animating the hypothesis. In the next chapter, I will treat the socioeconomic factors shown to be important for expenditures as constraints whose effects must be controlled if the role of councils and their members in the formulation of public policy is to be investigated.

PAGE 126

107 Notes 1 A summary review of this literature can be found in Terry Clark, Community Power and Policy Outputs; A Review of Urban Research (Beverly Hills: Sage, 1973) and Charles 0. Jones, "State and Local Public Policy Analysis: A Review of Progress," prepared for delivery at Conference on Political Science and Local Government, American Political Science Association (Biloxi, Miss.: October 26-29, 1972). James Q. Wilson uses the term "demographic analysis" in the introduction to James Q. Wilson, (ed.), City Politics and Public Policy (Cambridge: Harvard University, 1968). 2 James N. Danziger, "Comparing Approaches to the Study of Financial Resource Allocation," Comparative Public Policy: Issues, Theories , and Methods, eds. Craig Liske, William Loehr and John McCamant (New York: Halsted, 1975), p. 57. 3 Thomas Dye, Understanding Public Policy (Englewood Cliffs, N.J. Prentice-Hall, 1972), pp. 243-263. 4 For a discussion of the difficulties associated with this approach, see Stuart Rakoff and Guenther Schaeffer, "Politics, Policy, and Political Science: Theoretical Alternatives," Politics and Socie ty, Vol. I (November , 1976), pp. 51-78. 5 See Brett Hawkins, Politics and Urban Po licies (Indianapolis: Bobbs-Merrill, 1971), pp. 65-727 6

PAGE 127

1C3 13 Florida, Commission on Local Government, 1972 Florida Ce nsus of Local Government Finances , Special Report 13-9 (Tallahassee, flaT: Commission on Local Government, September, 1973). 14 Staff Report, "The History and Status of Local Government Powers in Florida," Committee on Community Affairs, Florida House of Representatives, July 31, 1972. 15 Florida, Joint Legislative Management Committee, Index to Special and Local Laws: 1845-1970 (Tallahassee, Fla. : Division of Statutory Revision and Indexing, n.d.). 16 Florida, Department of Administration, Revenue and Expen diture :or City, County and Special District Fla.: Department of Administration, 1973) 17 Clark, op. cit ., p. 63 ^~~^~^ Count y and Special Districts Fiscal~Ye^r 1972 (Tallahassee. IS See Herbert Jacob and Michael Lipsky, "Outputs, Structure, and Power: An Assessment of Changes in the Study of State and Local Politics, The Journa l of Politics , Vol. XXX (May, 1968), pp. 510-538 and Clark, ibid . 19 Florida, Commission on Local Government, op. cit ., p. 5. 20 Interview with William Shelly, Staff Economist, Local Government Study Commission, November 10, 1973. 21 Hawkins, op. cit ., pp. 51-84. 22 Robert Alford, "Critical Evaluation of the Principles of City i 33 ^ Ca ^ 10n '' Cit 7 Classification Handbook: Methods and App lications. ed. Brian Berry (New York: Wiley, 1972), pp. 331-359! 23 John Tropman, "Critical Dimensions of Community Structure: A Reexamination of Hadden-Borgatta Findings," Urban Affairs Quarterly, Vol. V (December, 1969), pp. 215-232. " L r ., ^ P hill P Ree s, "Problems of Classifying Suburbs wihtin Cities," ^pL-glagsifj^tion Handbook: Methods and Applications , ed. Brian Berry (New York: Wiley, 1972), pp. 265-330. 25 Dye, op. cit ., pp. 254-257. 26 William Michelson, Man and His Urban Environm ent: A Socioloeical_ J App_roach (Reading, Mass.: Addison-Wesley, 1970), p. 19. 27 See the numerous studies included in Brian Berry (ed.) City aassmcationjlg ndbook: Methods and Applications (New York: Wiley -Ly/^ as well as Jeffrey Hadden and Edgar Borgatta, American C ities (Chicago: Rand McNally, 1965) and Richard Forstall, "Ecm^Tc Classification of Places over 10,000, 1960-1963," The Municipal Year Book 1967 fnhirPonInternational City Managers' Association, 1967), pp. 30-48

PAGE 128

109 28 Variables within the same category found to be highly intercorrelated (r i .70) were identified, and then, through a process of random selection, a proportion of the variables equal to half of those so intercorrelated with a category were eliminated from further analysis. Sharkansky and Hoffert recommend this procedure in "Dimensions of State Politics, Economics and Public Policy," American Political Sci ence Review (September, 1969), pp. 867-879. ' 29 John Kirlin and Steven Erie, "The Study of City Governance and Public Policy Making; A Critical Appraisal," Warner Modular Publications Reprint (Andover, Mass.: Warner Modular Publications, 1973), p. 7. 30 Jacob and Lipsky, op. cit . , p. 519. 31 Clark, op. cit ., pp. 57-67. 32 Model City Charter prepared by Committee on Revision of the Model City Charter (5th ed.; 7th printing: New York: National Municipal League, 1957) . 33 Gladys Kammerer, Charles Farris, John DeGrove and Alfred Clubok, C ity Managers in Politics: An Analysis of Manager Tenure and Termination (Gainesville, Fla. : University of Florida, 1962). 34 John Fenton and Donald Chamberlayne, "The Literature Dealing with the Relationships Between Political Processes, Socioeconomic Conditions and Public Policies in the American States: A Bibliographic Essay," Polity^ Vol. I (Spring, 1969), pp. 388-404. 35 Eulau and Prewitt, op. cit ., p. 241. 36^ Jerry L. Yeric, "City Managers, City Councilmen and Public Budgets," prepared for delivery at Annual Meeting of Midwest Political Science Association (Chicago: April 29-May 1, 1976). 37 Micharl Aiken and Robert Alford, "Community Structure and Innovation: Public Housing, Urban Renewal and the War of Poverty," Compara tive Community Politics , ed. Terry Clark (New York: Halsted, 19747 pp. 231-287. 38 Ibid ., p. 244. 39 Richard Hofstader, The Age of Reform: From Bryan to F.D.R, (New York: Vintage Books, 1955), p. 205. 40 Aiken and Alford, op. cit ., p. 249. 41 Ibid ., p. 255. 42 Ibid., p. 256. 43 Lawrence Mohr, "Determinants of Innovation in Organizations," American Political Science Review , Vo. Ill (March, 1969), pp. 111-126.

PAGE 129

110 44 James S. Coleman, Community Co nflict (New York; Free Press 1957). 45 Aiken and Alford, op. cit . 46 Ibid . 47 Danziger, op. cit ., p. 63. 48 The .05 level of significance was used; a coefficient greater than or equal to r = ±.19 meet this criterion. 49 Aiken and Alford, op. cit . Aiken and Alford write tests that are appropriate despite the population being examined. First, they argue, census observations are liable to considerable errors in measurement (see p. 234). Second, the selection criterion is particularly arbitrary. It is extremely likely cities manifesting significantly different patterns of expenditures have been included in the population, while cities of more complexity have been deleted. Third, the present list of cities is itself a sample cross-sectional of these same cities when looked at along such dimensions as time. 50 For a discussion of this problem, see Charles Cnudde and Donald McCrone, "Party Competition and Welfare Policies in the American States," American Political Sc ience Review, Vol. LXIII (September, 1969) pp. 860-862. " 51 The tables, together with the reduced list of independent variables, can be obtained from the author. Separate tables were generated for total expenditures, per capita expenditures and expenditures by function as a proportion of the total expenditure in a city. 52 Hofstader, op. cit ., p. 9. 53 Amos Hawley, "Community Power and Urban Renewal Success," American Journal of Sociology , Vol. LXVIII (June, 1963), pp. 422-431. 54 See Aiken and Alford, op. cit . and Hawkins, op. cit. 55 Ibid . 56 Hawkins, op. cit . , pp. 61-84. 57 Martin Shefter, "Review of Politics and Urban Policies by Brett Hawkins," American Political Science Review , Vol. LXIX (March, 1975), p. 274. 58 Charles Lave and James March, An Introduction to Models in the Social Sciences (New York: Harper and Row, 1975), p. 40. 59 Harold Sprout and Margaret Sprout, "Environmental Factors in the Study of International Politics," The Shaping of Foreign Policy , eds . Harold Jacobson and William Zimmerman (New York: Athertcn, 1969) /pp. 43-

PAGE 130

Ill 60 Ibid ., p. 49. 61 Ibid ., p. 50. Deo-f.-fn! m l° V eX3 "! ple ' See Terr y Clark (ed.), Cmmnuni^Structu^ and D££ision_Mak 1 ng: Comparative Anal I ses (San Fra^Ii^7^hl^dlIrTT%8) . „., " Chauncy Harris, "A Functional Classification of Cities in the United States," Geographical Review, Vol. XXX (January, 1943) , pp. 86-99.

PAGE 131

CHAPTER FOUR CLUSTERING FLORIDA CITIES Most studies conducted within the demographic perspective have concluded environmental factors such as socioeconomic conditions correlate higher and more consistently with expenditure measures than do indicators of political arrangements such as governmental structures. And further, most studies have found environmental factors remain important even when structural effects are controlled. As a way of understanding municipal policy formation, the demographic approach is of little use, but its findings can be applied to an investigation of the manner in which city councils help shape policy. Clearly, environmental conditions may constrain the actions of local councils in some fashion. This chapter is devoted to an analysis of 78 Florida cities, their socioeconomic characteristics, and spending habits. It represents an effort to identify cities with similar socioeconomic environments; cities in which councils face relatively similar constraints. Indirectly, I am attempting to limit the variations in socioeconomic conditions facing the members of disparate councils so as to facilitate an examination of the link between council politics and municipal policy. Socioeconomic conditions, it is thought, should not so limit the policy options of council members as to result in a common response by the councils of the same socioeconomic cluster. Instead, even councils within the same socioeconomic grouping should evidence distinct policy differences. Considering the Operational Environment In the last chapter, I argued that the demographic approach, while 112

PAGE 132

113 convincing at the most general level, does little to advance our understanding of urban policy making. In particular, I suggested it could not sustain a detailed investigation of policy making. Nonetheless, its findings are important. Specifically, the demographic perspective cautions that socioeconomic characteristics should be conceived as constraints upon urban public policy. In some sense, the actions of council members should be limited by their context. If this position is correct, then it is imperative that any research design aimed at the detailed consideration., of public policy incorporate environmental factors bearing upon council members and their policy decisions. If socioeconomic characteristics are the occasion of policy, controlling such conditions should allow for a consideration of distinctively political factors shaping public policy. Unfortunately, there are many aspects of a community's circumstance which are encompassed by the conceptual notion of environment. In the demographic approach, environmental factors are usually limited to socioeconomic indicators. However, many socioeconomic variables correlate poorly with any measure of policy performance. Such realizations caution the acceptance of any indicator or set of indicators for environmental factors. But despair is not the remedy; the specification of the environment can proceed along several paths. The paths may be many, but all have a similar form. In order to acquire information about the relation between two or more properties in an empirical design, several procedures (perhaps assuming different methodological forms) are necessary. One needs an observed partition on at least one property, a design partition on at least one other property, and randomization of the other variables. 1 In fact, meeting all

PAGE 133

114 these conditions is quite difficult. But Heinz Eulau has recommended a quasi-longitudinal, quasi-experimental design dealing with local policy which fulfills most of the requirements. His design constant becomes the hypothesized uniformity of the region in which his sample of cities lie; the observed partition is each city's degree of urbanization as measured by its population size; his design partition is the policy environment of each city as measured by its expenditure ratios for development planning and city amenities. 2 Eulau is suggesting, in fact, the manifold aspects of a city's policy-making environment can be captured by an emphasis on three major features of the municipality— its physical, socioeco^ nomic, and policy environments. Eulau' s argument is in keeping with the thrust of the possibilist position. He explicitly recognizes environmental factors can impact upon policy, but accompanies this recognition with the realization that the environment is too complex to be included, in toto, in any research project. Thus he posits a set of factors which are significantly relevant for the making of local public policy. Unfortunately this design cannot be easily adapted to the present research situation. Part of Eulau' s design depends upon restricting the design constant to a small region, the area around San Francisco. 3 Extending this procedure to a larger and more historically diverse geographical area, such as the state of Florida, would subject the design to the vicissitudes of many confounding factors. Suppose,, however, that Eulau 's insights are extended to municipalities in Florida; that his conception of the conditioning influence of the physical, socioeconomic environment on policy making is a useful statement of the possibilist position. How can this approach be adopted for research in Florida's municipalities? First, it should be recognized

PAGE 134

115 that if the physical, socioeconomic environment of a city does condition its policies, then it can be expected cities with relatively similar physical, socioeconomic circumstances will have relatively similar policies, ceteris paribus . It is just elements of the policy making process itself which are not likely to be equal. Thus differences in the policies of relatively similar cities can be traced to differences in the way policies are made. Specifically, in this research, the differences may be manifestations of the contrast in politics between their councils. Eulau produced groups of relatively similar cities by examining cities within the same region and within population groups of varying size. However, there are many different ways in which such cities can be identified which are consistent with the possibilist position and the empirical -. cBndKona in Florida. For years geographers and other scholars have attempted to produce classifications of cities. Lewis Mumford generated _ a. simple scheme almost forty years ago/ More recently, classifications have made extensive use of contemporary computer technology and easily available software packages to process enormous amounts of data in accordance with various clustering techniques. Such techniques can operate upon census-type variables and comprehend the considerable diversity of cities and variables. And when every city lies within a single state, a host of variables not contained in census tables, but common to most municipalities within that state, are also controlled. In this manner the thrust of Eulau 's design can be preserved. Perhaps the most common technique used to produce clusters of similar cities is factor analysis. Typically, since the results will be used in further mathematical manipulation, researchers will use an orthogonal rotation to achieve some form of simple structured when a varimax ro-

PAGE 135

116 tattoo is employed, each of the factors extracted is given its maximim intestability. 6 Each of the factors or dimensions which emerge from this analysis captures some aspect of the socioeconomic environment. On each of these factors, a factor score can be computed for each city. These scores represent a composite of the relative importance of each of the variables for charactering a city and its magnitude for that city. The factor scores represent a reduced space in which each city occupies a unique element. The scores can be used as criteria on which cities can be grouped with a number of techniques. These include the following: Q-factor analysis, discriminant analysis, and any of a number of clustering alogarithms such as Ward's. 7 While there are several means of grouping cities, in each a common problem presents itself. Namely, how are groupings to be produced that are homogeneous and different from one another? This problem is hardly trivial, and for each solution difficulties will remain. 8 However, for ease of computation, tests of homogeneity, and group difference, one technique can be recommended— discriminant analysis. 9 The Establishment of Environment The use of factor analysis in city classification has been subjected to rigorous critique." Yet many of the difficulties are associated with the use to which scholars have put the technique. It is generally assumed the "goal of both the social scientist and the policy maker is ultimately to discover the causes and consequences of observed characteristics of cities. "11 However, it does appear factor analysis can be profitably US ed in order to gain control over selected features in a research design. Clearly, errors will be made when relying upon a small set of factor scores to characterize aspects of an entire city. "Factor

PAGE 136

117 scores are like aerial photographs, which do not always provide the necessary detail for fine analysis but nevertheless afford a useful topology of the landscape." 12 The point remains: a comparative research design must control for elements of the units under comparison, and factor analysis is a systematic, if flawed technique for obtaining control over a large number of factors. 1 -^ In order to specify the highly salient aspects of the socioeconomic environment, the product moment intercorrelations of 73 variables from the six categories described in the previous chapter were calculated (see Appendix D) . A principal components solution was produced yielding 14 factors. The first 12 factors "explain" approximately 87 percent of the variance. A varimax rotation of these factors was performed; the results of this procedure can be found in Appendix D. The 12 substantive factors or "latent dimensions" describe the environmental constraints of the 78 cities. While the dimensions have emerged through the internal logic of the procedure, they can be given empirical meaning by examining the degree to which sets of variables are involved in a factor pattern. This process, performed in Appendix D, yields 12 factors whose labels have been designated in accordance with their empirical involvement with a series of variables and their use among factorial ecologists. The factors, given in terms of their importance in the solution, follow: Factor I Functional City Size (SIZE) Factor II Stage in the Life Cycle (LIFECY) Factor III Socioeconomic, Income Status (INCOME) Factor IV Ethnicity (ETHN) Factor V Governmental Occupation and Institutional Connection (GOVTEM) Factor VI Young, Suburban Concentration (YOUTH) Factor VII Unemployment (UNEMP) Factor VIII Manufacturing (MANUF)

PAGE 137

118 Factor IX Veterans (VETERN) Factor X Income of Black Families (BLACKI) Factor XI Black Blue Collar Workers (BLUE) Factor XII Commuting (COMMUTE) The difference in the amount of variance accounted for hy the factors falls gradually from 25 percent for factor I, 17 percent for factor II, 11 percent for factor III, and 8 percent for factor IV to nearly 12 percent of the total variance for factors VIII through XII. Do the factors extracted from the varimax rotation capture the "essential dimensionality" of the original variables? Unfortunately, it is difficult to know. There are numerous criteria which may be used. However, the most widely "used rule today is that proposed by Kaiser (I960): to retain only those components corresponding to eigenvalues of the (complete) correlation matrix that are greater than unity. »U Yet I have deviated from this guide. Only 12 of the 14 factors with eigenvalues greater than one were selected. The f inal two were deleted, not only because their contribution to the "explained" variance was small, but because their substantive interpretation is quite tenuous. Of course, interpreting factors is generally difficult, but the problem is compounded when the entire character of a factor rests on a single high loading. In short, attempting to give the remaining factors meaning stretches their empirical worth beyond their conceptual soundness. The extracted dimensions are requisite to further analysis. They can be used to generate factor scores upon which a discriminant analysis can operate to form distinctive groupings. 15 However, more extensive analysis is contingent upon the nature of the factors themselves. The factors should behave as environmental constraints in the same fashion as the original data.

PAGE 138

119 The Reliability of the Factor Solution If the factors obtained from analysis of the socioeconomic data are to be used in grouping the cities, their purpose must be kept firmly in mind. The factors are used to specify the environment. But the solution produced by a factor analysis is always problematic; there is little way to know the structure in the original data has been preserved. Thus the new factors must be checked to ascertain whether they function as measures of the environment in the same fashion as the original variables. Rather than duplicating the extensive correlational analysis undertaken earlier, I adopted a modified procedure. The number of expenditure measures was reduced through factor analysis in order to facilitate manipulation of the measures in a regression equation. While factor analyzing expenditure data runs the risk of distorting the data in such a way as to alter the nature of the linkages, it will make testing environmental linkages manageable. And the occurence of the same linkages will do much to establish both the validity of the environmental factors and the expenditure factors. Since the expenditure factors will be employed later in the analysis, this will afford an opportunity to gauge their worth. Of course, it is possible similar linkages could reproduce themselves by chance. Thus the possibility of distortion was guarded against by using two of the sets of expenditure measures. Two factor analyses of expenditure measures were performed. The first involves measures to total expenditure in each of the functional categories; the second deals with the percentage of total city expenditures in each functional area. Following the procedures developed with respect to the socioeconomic data, a varimax rotation of total expenditures was performed on the prin-

PAGE 139

120 cipal components solution. The analysis yielded six factors having eigenvalues greater than one. These six factors accounted for 83.7 percent of the variance in the original matrix. The factor analysis is reproduced in Table 4-1 . It is dominated by the first factor which accounts for little more than half the variance of the matrix (50.4%), and far outstrips the importance of the second factor (13.3%). The first factor can be identified by the emphasis on human resource expenditures (HUMRES) . It is dominated by such variables as public welfare, urban renewal, public health care, and natural resource expenditures. The second component, designated as service expenditures (SERVIC) , involves such variables as expenditures on fire protection, safety inspection, refuse collection, water provision, and police protection. The third factor has near zero loadings on all variables save two, water terminals and expenditures for the provision of gas. These are two specialized functions, but ones within the traditional purview of municipalities. I characterize this factor as specialized older public works (WORKOL) . The fourth factor is especially difficult to interpret since its highest loadings occur on variables designated "other." These categories cover a plethora of city expenditures. They represent the latitude of most municipal budgets, from maintenance of horse tracks to relocations funds for families residing in publicly condemned housing. This factor also correlates with library and park expenditures. Consequently, this dimension was called library and discretionary (LIBOTH) expenditures. The fifth factor is named newer public works (WORKNW) . The variable highlighting this factor is expenditures on public transportation; however, sewage and urban renewal expenditures also are highly involved. The sixth and final factor loaded most highly on expenditures for electric

PAGE 140

Table 4-1 Factor Analysis of Expenditure Variables: Varimax Rotated Factor Matrix 121

PAGE 141

122 utilities. Thus, while this dimension includes moderate relationships with hospital and interest expenditures, the factor was designated as electric utilities (ELEC). 16 A second factor analysis was performed on the percentage of total expenditures in each functional spending area. Similar techniques yielded a completely different set of factors. Thirteen factors having eigenvalues greater than one, accounted for 73.7 percent of the variance in the original data. This solution was less concise (more factors) and less successful (less variance accounted for) than the analysis of total expenditures. Table 4-2 depicts the analysis. The first factor is comprised of several variables whose pattern appears to suggest expenditures on a city's poor (POOR). Public welfare, public health care, and urban renewal expenditures, all correlate highly with this factor. However, expenditures on corrections and courts are also highly involved. Given the skewed distribution of social classes affected by the judicial system, the factor's characterization as indicative of the priority given a municipality's poor does not seem misleading. The second factor correlates most highly with expenditure concentrations on police, protective inspection, and safety items. These are designated as police powers (POLICE). The characterization of the third factor is more difficult. The percentage expenditures on hospitals and electric utilities is negatively correlated with this factor, while fire and sewage expenditures are positively correlated. Cities have long assumed such functions, however in Florida, these observations indicate a specialization goes on. Those cities providing their own electricity will tend to spend similar shares of their revenues on hospitals. In the absence of electric generating capacity, more traditional functions

PAGE 142

123 1

PAGE 143

124 Conununality

PAGE 144

125 will receive the allocations. This suggests the factor be designated as traditional shares (TRAD IT ) . Expenditures allocated to streets dominate the fourth factor; expenditures on water provision detract from this factor. There appears to be no obvious reason for this trade-off, and hence I will settle on a name which captures the exchange, water-streets (WATER). Variables associated with a city's transportation provisions, such as expenditures on the transit authority and parking, suggest a transit (TRANST) denotation for the fifth factor. A municipality's desire to preserve latitude in its expenditure distribution constitutes the sixth factor. The focus is on discretionary (DISCR) expenditures. The high loadings on government related expenditures indicate the seventh factor might capture a city's propensity to allocate monies for the government's administration and maintenance (GOVT). The eighth factor identifies a trade-off between expenditure priorities in libraries and port facilities, and hence is designated the port-library exchange (PORT) dimension. The moderately high loading of natural resources on the ninth factor suggests a characterization a s natural resources (NATRES). An inspection of Table 4-2 will reveal the remaining factors are associated highly with just a few variables. In each case, the factor has been named after the variable loading most highly on the dimension. Respectively, the factors are named refuse (REFUSE; #10), airports (AIR; #11), parks (PARKS; #12), gas (GAS; #13). The factor scores generated for each of the cities can be used to examine the adequacy of the factor analysis in reducing the number of variables used to describe a city's circumstance. Specifically, they can be used to check the linkages hypothesized previously (see Chapter three). To the degree the data reduction is successful, the form of the

PAGE 145

126 linkages should be similar to that found among the individual variables. I will check for the linkages by examining first, the set of correlation coefficients between measures, and then, a series of regression equations displaying the relative importance of the measures for different expenditures. The intercorrelation matrix of the twelve environmental factors, the ten political structure measures, and the two sets of expenditure factors will not be presented here in order to shorten the length of this investigation. Rather the results will be summarized in discursive fashion. 17 As in the previous chapter, the empirical surrogate of community differentiation suggested by Aiken and Alford, city size, correlates significantly with total expenditures in areas such as human resources and allocations for the poor . A city's functional size also correlates highly with expenditures delineated in the service factor and the so-called traditional share of the budget. However, functional size has a low correlation with expenditures on newer public works . The larger the city, the more expenditures a city is likely to make in areas covered under the transit factor. Beyond these relationships, a city's functional size apprehends none of the other expenditure dimensions. It is worth noting, however, that black income is likely to be lowest in those cities making expenditures on human resources or for the poor . (The direction of the coefficients is negative, but the values are not significant.) The direction of this relationship re-enforces Aiken and Alford 's conception of the impact of differentiation on municipal policies. To the extent the earlier environmental-expenditure relationship is found in the factor scores produced by the varimax rotation, other relationships are also dramatized. For example, the younger the age compo-

PAGE 146

127 sition of the city's residents, the less likely the city is to make expenditures on libraries and parks or to hold back discretionary funds. But such cities do have a tendency to allocate funds for electricity generation.. Cities scoring high on the ethnicity factor (primarily those cities in southeast Florida) appear likely to engage in discretionary funding and to support libraries. While cities in which there are concentrations of young people are unlikely to spend on libraries, an examination of the correlation coefficients alone indicates a moderate association with relative expenditures on police functions. Intuitively, this finding appears as reasonable as the relation between government employment and expenditures on electric utilities. In the former case, the young might be expected to commit more crimes and thus need more police; while in the latter, governmental personnel will be increased by the staffing of municipally owned utilities. Among the political structural variables, the one that bears the most significant relationship to the two sets of expenditure factors is the number of council members. The number of members is positively and strongly correlated with both human resource expenditures and spending with some emphasis on the poor. This is at odds with the findings reported earlier. Previously, council size was found to discourage expenditures on public welfare, urban renewal, and health care. This discrepancy is difficult to explain. In fact, this is the most noticeable failure of the factor reduction. Other aspects of the relationship between environment, structure and expenditures have been preserved. The negative relationship between a city's age and police expenditures remains, as does the direction of the relation between welfare expenditures and the form of government. In short, this major problem, although worri-

PAGE 147

128 some because of the size of the coefficients involved, does not contravene the worth of the entire analysis. It merely suggests caution. In the previous correlational analysis, the number of variables and the complexity of the possible relationships eliminated the possibility of estimating the most powerful constraints on municipal expenditures. Attempting to specify relationships through regression equations would have introduced additional difficulties. Aside from the sheer number of variables, the substantial multicollinearitvamong independent variables would have yielded parameter estimates of extreme volatility. 18 The use of factor scores does much to reduce both of these problems. Table 4^3 depicts the beta weights of regression equations where the dependent variables are the expenditure measures and the independent variables consist of all the socioeconomic factors. For each expenditure area and type, the entire set of SES factors have been forced into the equation pursuant with data search techniques. 19 As with the correlation analysis using factor scores, functional city size once again proves to be the most important socioeconomic constraint. It dominates the equations "explaining" human resources , service provision, and expenditure allocations favoring the poor. Moreover, only those equations have r-squares of greater than .35. Each of these equations account for more than half the variance in the expenditure factors. In short, to the degree any of the socioeconomic factors are effective in capturing the pattern of expenditures, city size is the best single predictor. Although functional city size is clearly the most important constraint among the set of socioeconomic dimensions, it is not the only significant factor, nor is it the most important factor for each expen-

PAGE 148

129

PAGE 149

130 diture area. Despite the relatively small amounts of variance explained in the other equations, each of the remaining SES factors are involved significantly in the explanation of at least one expenditure area. The only exception to this rule is black income , the tenth factor in the principal components analysis. In other words, each of the socioeconomic factors do serve as constraints in some fashion. Thus, it can be argued, further analysis should include the several factors; each has a substantive contribution to make. Only the inclusion of the factor called black income is of marginal importance. Political structural variables are much less successful in accounting for municipal expenditures than SES measures. In fact, none of the equations into which all the political structural variables has been forced account for more than 42 percent of the variance in the expenditure measures. Moreover, the individual variables themselves do poorly in explaining the expenditure patterns. Nonetheless, the importance of the respective political structural variables parallels that found in the correlational analysis of the raw data. The number of council members proves an important, if negative, indicator of expenditures in human resources. And once again, the age of the city bears a significant relation to several of the expenditure categories. Isolating the impact of political, structural, and socioeconomic variables i n distinct regression equations is instructive of the relative importance of political or socioeconomic components for expenditures. However, such a presentation cannot settle the question of the relative potence of the two types of constraints. Clearly, this is an important part of establishing the reliability of the various factors. Yet, forcing 20 variables into a single regression equation for each expenditure

PAGE 150

131 area and type would not solve the concern. As an alternative, a multiple partial correlation coefficient was calculated for each expenditure area, controlling successively for SES factors and political structural variables. This was done by using the b-weights of the most significant predictors of each expenditure category to generate the variance accounted for by these predictors. This variance was then used in an analysis partialling out the effects of any number of socioeconomic or political structural variables. Each multiple partial correlation coefficient was then compared against the multiple correlation coefficient on the same set of variables to measure the relative depression of the correlation coefficient when the variance of either the environmental or political structural variables is extracted. The greater the depression of the coefficient, the larger the effects of the controlled variables. Table 4-4 depicts the results of using regression estimates to generate values for both socioeconomic and political structural measures, and employing them, in turn, to calculate multiple correlation and multiple partial correlation coefficients. The Delta-C (AC) column indicates the level of depression in the coefficient when controls are added. Of the 18 expenditure variables, 13 are depressed more by SES variables than by structural measures. This occurs despite the possibility the overall larger multiple correlation values of the SES-expenditure relationships stand a greater chance of being affected by controls. This is because the variance pool at stake is greater, and absolute changes in magnitude are likely to be larger. For two of the expenditure dimensions, structural variables have the greater effect. In both cases, however, the magnitude of difference may be attributed to rounding error. ^ u In sum, the evidence does point toward the relative importance of socioe-

PAGE 151

132 en •u c QJ -H O •H C4H U-t QJ o u o •H QJ SH o u p. qj S C o H en en Q) U &, CJ Q CJ > •H 4-1 rt T— I a a en w o 4-1 CO CJ H o CO co 0) r4 a) Q cd •H 4-1 r4 CtJ cd o rH -H QJ CM P. rJ u o H O CJ 3 a QJ tn O 4-J c o •H tJ c p! o •H 4-1 03 U CJ H Yl tH CJ CH ft ^ in H M QJ 4J O O rH O O 3 C 4J o a QJ -H CJ H 4-1 tH a cj u •H H t) 4-> QJ U-i i— I V-l <4-l 3 rJ QJ S O O C-> CJ u 3 M 4-1 en cj en QJ en C 4J o c •H CJ 4J -H a, ra u tH rH -H 4-1 QJ CH rH J-l M-J 3 U QJ 2 o o u c_> CO en QJ U aj Q O 3 M 4-J en U S o CO CJ en CJ CJ 3 M 4-1 en en QJ en U S cu 3 TJ 3 CJ o a s w rH

PAGE 152

133 conoraic factors over political structural variables in a majority of expenditure categories. In re-examining the linkages between the environment of socioeconomic influences, political structural variables, and measures of expenditures, an attempt has been made to determine how faithful the reduced space is to the original data. The analysis was designed to establish the reliability of the factor solutions before using the results in subsequent analysis. While the results are hardly conclusive , the analysis once again identifies environmental socioeconomic constraints as preeminent. Moreover, the factor analytic solutions appear to preserve the character of the original data set. Overall, it appears reasonable to proceed with the classification of the cities using the factor scores already generated. City Classification: Socioeconomic Environment Cities can be grouped using discriminant analysis, but the technique requires a preliminary specification of the number of groups to be considered, and a tentative specification of the members of each group. Something must be known of the way in which cities cluster along particular criteria before refined classification can continue. The a priori grouping of cities can proceed in a number of ways. In the absence of explicit theoretical criteria, two methods were used. First, a scattergram of each city along the SES factors was examined for similar structural profiles. And in fact, visually sorting through the 78 cities revealed at least six distinct profiles. However, the profiles were quite complex, and for several cities, inclusion in a single group was problematic. Thus a second method was used to complement the visual inspection. A Q-technique factor analysis of the intercorrelations of

PAGE 153

134 the city factor scores on the socioeconomic measures was performed in order to generate the "types of cities" which appear to exist in the original set of factor scores. Accepting the notion that there were six types of cities in Florida, a quartimax rotation of a principal components solution was used to identify the factor space in six dimensions. These six factors accounted for 70 percent of the total variance of the factor score matrix; no single dimension dominated the solution. The contribution of the factors to the variance accounted for in this solution range from 14.3 percent to 8.8 percent. The loadings of each of the cities on the six factors were inspected to determine the extent to which each city was involved in a particular factor or city type. 21 This initial analysis provided the basis for a discriminant analysis of the Florida cities using the socioeconomic environmental constraints generated earlier. For each city, the posterior probability of a city's group membership and its distance from the group centroid was calculated. These findings enabled a city's membership in the a priori grouping to be evaluated and the sense of group membership refined. Where the initial grouping was felicitous, the distance of a group member from its centroid was less than that from the centroids of the other groups. Additionally, the posterior probability of a city's membership in a group was high. The final "assignment" of a city by the discriminant analysis will not differ from the initial assignment when these conditions hold. The assignments based on the visual sort and the Q-factor analysis proved extremely good. Correct assignments marked 96.6 percent of the cases. The final assignment of cities, and the list of environmentally constrained groups, is reproduced in Table 4-5. 22

PAGE 154

135 Table 4-5 Socioeconomic Groupings Used Throughout Study : Groupings of Florida Cities Over 10,000 Population Based on Discriminant Analysis of 73 SES Variables Reduced By Varimax Factor Rotation into 12 Scores Per City (Primarily 1970 Data) 005 Bradenton 008 Clearwater 014 Deland 016 Dunedin 022 Gulfport 029 Kissimmee 031 Lake Worth 032 Lakeland 033 Largo 034 Lauderdale Lakes 036 Leesburg 045 New Smyrna Beach 056 Pembroke Pines 058 Pinellas Park 067 Sarasota 071 Tamarac 074 Vero Beach B 003 Belle Glade 019 Fort Pierce 026 Homestead 059 Plant City 066 Sanford 077 Winter Haven 004 Boynton Beach 012 Daytona Beach 013 Deerfield 015 Delray Beach 017 Ft. Lauderdale 018 Ft. Meyers 023 Hallandale 025 Hollywood 027 Jacksonville 040 Miami 047 North Miami Beach 051 Opa Locka 052 Orlando 061 Pompano Beach 062 Riviera Beach 065 St. Petersburg 072 Tampa 075 West Palm Beach A D 011 Coral Gables

PAGE 155

136 Regardless of the statistical worth of a classification, its empirical form should make some intuitive sense. In other words, to a well travelled and keen social observer, the six socioeconomic environments should appear reasonable. Of course, it is unlikely such an observer exists. The environmental constraints specified above, and the number of cities studied, reflect a complexity which would defy the most detailed "windshield" survey. As an alternative, I turned to an intuitively appealing geographical notion; namely, that cities embedded in a metropolitan region should appear in the same grouping quite often, if they are contiguous. This is especially the case with cities in southeast Florida and on the Gulf coast. Here the geographical distinctions between cities as governmental bodies need not sustain differences in socioeconomic conditions. In many instances, even neighborhoods overlap the political boundaries. A contiguity analysis of the six socioeconomic groups confirms such expectations. Table 4-6 presents the results of an examination of the fashion in which cities in each socioeconomic cluster are themselves contiguous. Notice, in each group, several contiguous cities are found, while the classification scheme does differentiate between cities in a metropolitan region. For example, cities such as Largo and Gulf port are located in the Tampa-St. Petersburg SMS A, and they have been assigned to a socioeconomic grouping quite distinct from that into which Tampa and St. Petersburg fall. It is instructive to mark the manner in which the clusters apprehend cities in the same region. The cities of Lauderhill, Plantation, Sunrise, and Margate are suburbs of Ft Lauderdale. The northern suburbs of Ft. Lauderdale, Wilton Manors and Oakland Park, demonstrate different

PAGE 156

137

PAGE 157

138 socioeconomic characteristics from those of the west. The latter fall in SES group D, while the former are found in group E. The classification scheme has differentiated the country club cities of the west from the older, more socially mixed cities north of Ft. Lauderdale. Yet it has also placed the country club suburbs into a single cluster. In short, the contiguity analysis lends support to the classification scheme, and its worth in capturing meaningful socioeconomic differences. If the classification is meaningful, is there any way in which the distinctive socioeconomic environments can be identified? For convenience, and to emphasize the artificial nature of the city groups, I labelled the groups alphabetically in Table 4-6. A more meaningful perspective on the groups can be purchased in an examination of the discriminant functions used to differentiate cities. The most important factors involved in separating cities were functional city size, ethnicity, socioeconomic-income status and stage in the life cycle. These factors can be represented in two dimensions as depicted in Figure 4-1 . Here the ordinate is a combined income-stage in the life cycle variable, while the abscissa is a size-ethnicity dimension. At least in two dimensions the difference between the group A and E cities is difficult to detect; they are located squarely in the center of the figure. However, group F appears in the upper half of the income-stage in the life cycle dimension, while the group B cities are in the lower half. The size-ethnicity dimensions divides the group C cities from those in group D; group C has a centroid below that of group D. While it must be remembered the two-dimensional representation simplifies the criteria used in separating the cities, it does provide a handle by which the groups might be further identified. For example,

PAGE 158

139 o

PAGE 159

140 group F appears to contain high SES, demographically young cities. Typically, group B cities would appear just the opposite — low income, demographically older cities. The cities of group C are, on average, the largest, most ethnically diverse cities, while in group D cities tend to be similar and more ethnically homogeneous. To the degree the two dimensional representation divides the cities of groups A and E, the distinction seems to be in the income-stage in the life cycle components of the cities. Group E has a higher income, younger demographic cast than those of group A. Of course, such summary identifications may be misleading. They suggest a reality I would like to avoid. Hence throughout the remaining analysis, I will continue to use letter designations to differentiate between groups. However, where necessary, and to avoid the tedium of alphabetic identification, I will prick the readers memory with references to the dimensional considerations given above. The two-dimensional representation does appear to capture the most salxent aspects of the socioeconomic circumstance of the 78 Florida cities, or at least, its prescription is not strikingly at odds with my own impression of the many cities. Without engaging in a post hoc justification of the groups reminiscent of natural area analysis, I would like to suggest the clusters are certainly plausible. 23 Consider group C. It contains the most populous cities in the state; cities where ethnic diversity is the greatest. Even smaller cities such as Opa Locka and Daytona Beach have surprising concentrations of different ethnic groups. Group E consists primarily of retirement communities which have sprouted up relatively recently. Palm Beach, long a resort for the wellto-do, and Naples, a west coast resort noted for its shelling, are exceptions to this pattern. However, even these cities share the relative affluence common to the other cities in group E. j

PAGE 160

141 Group F contains cities of special interest, which I will examine more closely throughout this study. According to the two-dimensional representation, these are the most favorably endowed cities in the state — their residents have high incomes and are well educated. An examination of the cities reveals why this is the case. Several cities have universities; the largest universities in the state are in Gainesville and Tallahassee. The professional, well educated individuals in these cities are matched by similar aggregations assembled in the area of the NASA complex at Cape Kennedy and Merritt Island. Rockledge, Melbourne, Cocoa, Cocoa Beach, and Titusville are hosts to scores of skilled technicians, engineers, and scientists. Bartow is the county seat of the largest county in the state (by area) and brings together a substantial number of white collar employees in a growing governmental complex. In similar fashion, Ft. Walton Beach, adjacent to Eglin Air Force Base and Special Weapons Testing Area, is the gathering point for highly skilled professionals and retired military people on substantial government pensions. Cape Coral is perhaps the most anomalous city in group F. It is a retirement community, experiencing phemonenal population growth. The town was the site of a former development scheme that laid waste to nearly a hundred square miles of Florida scrub lowland. Here, huddled in one corner of the immense development, scores of professional, retired people labor to revitalize the landscape and to build a city out of the web of canals and asphalt. Physically, the city is much different from the other cities in group F; socially, it has much in common. The cities in group B are peripheral to the dominant urban nexus in Florida. Cities such as Belle Glade, Homestead, and Sanford lie on the outskirts of major urban centers; Plant City and Winter Haven are

PAGE 161

142 older cities embedded in the orange country where growth has passed them by in favor of Lakeland. All the cities have considerable agricultural connections, although Fort Pierce is rapidly expanding and diversifying its economy. Many of the cities in Group D also lie on the periphery of major urban centers. As shown earlier, few lie in large metropolitan complexes. Of course, cities such as Miami Beach, Miami Springs, North Miami, and Coral Gables are part of the Miami complex and it may be difficult to see what they share with communities such as Lake City and Key West. But these Miami cities are in a period of decline in the face of a new suburban challenge; the areas typically house families of moderate income. Even when the older affluent sections of Coral Gables are considered, the overall cast of the cities is relatively the same. The cities in group A are more difficult to categorize. I am tempted to say the commonality lies in their prominence as sites for baseball spring training. In fact, there is something to this notion since many of the cities have been relatively large (10,000 to 25,000 people) for many years and have not experienced the explosive growth of many Florida communities. However, cities such as Tamarac and Pembroke Pines hardly fit this pattern. In the end, the two-dimensional representation appears to have captured the most salient commonalities, namely, the way the cities seem to fall in the middle range of all Florida cities on most socioeconomic measures. City Classification: Expenditure Types The research design I have adopted in this study conceives of environmental constraints as sufficient conditions for the commission of public policy. The public actions of municipalities, it suggests, are an intricate blend of many forces, only a few of which are likely to be ap-

PAGE 162

143 prehended in the mesh of census-type variables. Nonetheless, aggregate measures of the socioeconomic environment can be used to isolate and control a portion of the factors confounding the study of public policy. Constructing artificial, but homogeneous and distinct, clusters of cities sharing similar socioeconomic circumstance is one method of conditioning a more detailed investigation of public policy. Before shifting our concern to a more intensive study of the forces shaping public policy, it is worth remembering the stimulus for the present research design. It originated in the conceptual muddle of the demographic approach. Socioeconomic variables were closely linked to expenditure patterns. This recollection prompts an interesting speculation. If the socioeconomic clusters were used as the a priori grouping for expenditure profiles of the many cities, would cities with similar socioeconomic circumstances occasion similar spending patterns? The demographic perspective argues the possibility of just this happenstance is considerable. If local political machinations matter, the chances are more remote. In any event, the groupings of cities by expenditure patterns will provide a dramatic portrayal of the environment-expenditure relationship. Taking the six SES groups as the preliminary assignment of group membership, a discriminant was performed on the two types of expenditure measures. Total city expenditures and relative expenditures by functional category were examined using the factor scores generated earlier. The final assignments derived in each analysis are shown in Table 4-7 . A quick Inspection of the various socioeconomic groups reveals that more often than not the final assignment of group membership differs from the original classification. The exception, proving this rule, is

PAGE 163

144 Table 4-7 Summary Listing of Final Assignments by Discriminant Analyses of Cities on Total and Relative Proportions of Expenditures in Each Functional Category Final Assignment roup Total Expenditure Relative Expenditure Group Final Assignment Total Relative Expenditure Expenditure A adenton earwater Land me din ilfport simmee ke Worth keland rgo uderdale Lakes esburg |hw Smyrna Beach mbroke Pines nellas Park rasota amarac sro Beach B lie Glade Pierce mestead ant City nford Inter Haven )ynton Beach lytona Beach ierfield Beach ;lray Beach Lauderdale Meyers illandale )llywood icksonville [ami :th Miami Bch >a Locka rlando )mpano Beach .viera Beach :. Petersburg impa :st Palm Beach E D E E E A B B E E B E E E F E B E B B E E E F A E E C A E C C C D E C E D C C C C D C F A B B B D E B A A F C E B C B B F A D F D F D C D D C C C C A C C D C C C D Coral Gables C Key West E Lake City E Miami Beach D Miami Springs E Miramar E North Miami A Oakland Park E Ocala B Ormond Beach E Panama Beach E Pensacola D St. Augustine E Wilton Manors E Winter Park E E Casselberry E Lauderhill E Lighthouse Pt. E Margate E Naples E North Palm Beach E Palm Beach E Plantation E South Miami E Sunrise E Bartow A Boca Raton D Cape Coral E Cocoa A Cocoa Beach E Ft. Walton Beach E Gainesville B Hialeah C Melbourne A Rockledge E Tallahassee B Titusville E D F D D C A F F B F C D C B D E E E E A E E A E E A F F F F F B A C F B D Correct Assignments 29.5% 43.6%

PAGE 164

145 found among the expenditure patterns of the cities in group E. In this instance, ten cities have relatively similar expenditure patterns associated with similar socioeconomic environments. It should be noted, however, that many other cities with different socioeconomic environments share this expenditure pattern. The ten cities in expenditure group E are less than 25 percent of the total number of cities falling in group E as result of the final expenditure assignment along expenditure criteria. The magnitude of the difficulties associated with the use of the SES groupings as an a priori solution to description of expenditure types can best be demonstrated by determining how frequently the initial and final assignments were the same. For total expenditures by functional expenditure category, the SES clusters do an extremely poor job of predicting the expenditures of a city. In only 29.5 percent of the cases does the strict socioeconomic environment link to expenditures hold. Even at that level, the Wilk's lambda indicates the group centroids are not significanlty different from one another to consider the differentiation of expenditure groups meaningful. 2 ^ The relationship between socioeconomic conditions and relative expenditures is more meaningful than that considered above, but, even in this case, the predictive abilities of the SES classification are still remarkably low. The SES clusters act to classify 43.6 percent of the relative expenditure pattern correctly. The inadequacy of the socioeconomic grouping as preliminary clusters for expenditure patterns is most demonstratively shown when city expenditures are differentiated in the same fashion used to produce the SES clusters themselves. Starting from factor score profiles, and a Q-factor

PAGE 165

146 analysis employing a quartimax rotation, a priori groups were derived for both sets of expenditures. A discriminant analysis based on these results yielded groups considerably different from the SES clusters. The six patterns found in the SES groupings are not repeated in the expenditure groupings. Rather four distinct forms emerge; and these forms cannot be produced by merely lumping together groups appearing in the analysis performed on the SES groupings. Table 4-8 portrays the final assignments of cities on the basis of their total expenditures in the several functional categories. The discriminant functions indicate the cities in groups B and C are considerably different from one another. This distinction is drawn by their loading on the electric utility and service factors. Group C which contains several of the large cities in the state is heavily identified with the provisions of traditional services. Cities such as Gainesville, on the other hand, engage in expenditures on public utilities. By far the great portion of Florida cities have relatively similar expenditure profiles. Forty-six cities fall into group D which represents a mix of expenditures across several functional areas. This is a significant departure from the six clusters of SES cities. In fact, when expenditures are grouped in this fashion, using an a priori starting point derived from only a consideration of expenditures, the final assignments match the initial assignments 87.2 percent of the time. This is substantially better than the grouping based on the SES groups. The discriminant analysis of the relative expenditure factor scores was based upon six groups or patterns discerned in an inspection of the scattergrams of the 78 cities on the scores. The clusters calculated on this basis do not overlap to a significant degree with those coming out

PAGE 166

147

PAGE 167

of the socioeconomic analysis. Instead, the distribution of cities across the groups is more uniform. The number of cities in each cluster ranges from 10 to 17. Once again, the so-called retirement cities of SES group E cluster together, but surprisingly, high SES cities such as Gainesville display allocation priorities similar to those of Ft. Pierce. Such clustering stems from the importance of high allocations in the police dimension. There are other examples of cities from vastly different socioeconomic groups having similar expenditure priorities. Consider Winter Haven and Boca Raton. Both appear in the same expenditure cluster, yet Winter Haven was identified as a city having relatively poor families, while the families in Boca Raton were rather favorably endowed. An inspection of Table ^9 shows the elements of several of the expenditure groups cut across the SES clusters. In short, there is no apparent relationship between the SES groups and the relative expenditure patterns. 25 The alert reader may have noted there is a slight difficulty in concluding the relationship between expenditures and socioeconomic conditions is not isomorphic. Namely, the groups have been formed from different pools of variance. Previously, Box's M was used to check upon the equality of dispersion in the various groups. However, in comparing SES groups with expenditure groups no such check was made. In fact, such a comparison would be exceedingly difficult to accomplish. A complex transformation would be necessary to normalize one group's variance to the variance in the other across groups. This is a problem I have not been able to solve directly. However, the problem can be handled indirectly. Suppose we examine a single socioeconomic grouping. If expenditures formed the same or similar profiles in each city throughout a SES grouping, it would be conWW

PAGE 168

149 Table 4-9 Grouping of Florida Cities Based on Discriminant Analysis Of Percent of Total Expenditure Variables By Functional Area Largo Pembroke Pines Gulfport Opa Locka Miramar Naples Plantation Wilton Manors Hallandale Daytona Beach Dunedin Cocoa Delray Beach B New Smyrna Beach Bartow Vero Beach Leesburg Kissimee Lake Worth Lakeland Ft. Pierce Homestead Ocala Gainesville Tallahassee Sarasota Belle Glade Pompano Beach Ft. Lauderdale North Miami Jacksonville Tampa Orlando West Palm Beach St. Petersburg Miami Springs Panama City Melbourne Hialeah D Miami Bel and Hollywood Clearwater Ft. Meyers Riviera Beach Winter Park Pensacola Miami Beach Coral Beach Lake City Titusville E Lauderdale Lakes Tamarac Sunrise Casselberry Lauderhill Lighthouse Point North Palm Beach South Miami Palm Beach Margate Sanford Pinellas Park Plant City Deerfield Beach Boynton Beach Ormond Beach Key West North Miami Oakland Park Cocoa Beach Rockledge Cape Coral Ft. Walton Beach Boca Raton Winter Haven St. Augustine

PAGE 169

150 vincing evidence of the connection between socioeconomic circumstance and expenditure policies. To the degree distinctly different expenditure patterns surface within a single socioeconomic group, a different case is supported. At best, socioeconomic conditions provide the occasion for similar expenditures; it helps establish a circumstantial case for the conditional and non-necessary effect of socioeconomic environment . " This particular solution turns out to have an additional advantage. It helps re-enforce the ability of the research design to test for the effects of the policy making within relatively similar socioeconomic milieu. One of the primary difficulties of demographic analysis is that it often relies on expenditures as measures of public policy. Obversely, most studies of public policy examining the dynamics of policy making have ignored expenditures as meaningful expressions of policy priorities. Establishing expenditure patterns within controlled socioeconomic environments affords the opportunity to relate the political aspects of a community to its nominal expenditure priorities. At the aggregate level, political influence can only be inferred. There are six SES groups, but not all were analyzed for internal differences in their expenditure patterns. Group B with only six members was considered too small for meaningful statistical analysis. Examining each of the remaining groups would have been an enormous task. Thus I selected three groups. Group F, the high SES cities, was selected because it so clearly emerged from the discriminant analysis as a viable and cohesive group. Further, I thought that among such cities I might be able to find clearly differentiated expenditure patterns. If socioeconomic environments are merely conditional aspects of policy, then the m

PAGE 170

151 presence of many professional groups, different forms of governmental activity (federal, state, and county interpenetration) , and the higher interest in politics which accompanies individuals of high socioeconomic status should make for increased competition in the political arena, and hence, increase the probability that expenditure patterns might differ. 27 Group C \vras chosen for similar reasons. Size is often used as a surrogate for differentiation and the presence of diverse and competing groups. Thus in group C, I believed expenditure differentiation should also occur. In group D, I had reason to believe otherwise. The member cities form a group of low SES environments; the cities are of moderate size. They represent an extreme opposite of group C. Hence, if an undifferentiated pattern of expenditures were to occur, group D appears to nurture favorable conditions. The analysis was performed as before. Expenditure factor scores for the entire set of cities were inspected and likely group clusters noted within each of the SES groups. These together with the quartimax rotation performed earlier guided the formation of a priori groups for each of the SES groups under investigation. Discriminant analyses were then performed on each set for total expenditures by functional category. The results can be found in Table 4-10 . In each case, three distinct expenditure groups appeared. Surprisingly, the patterns found among the cities of group C are the most mixed, while the patterns in groups F and D are rather unambiguous. In all cases, the groups are found to be statistically different. In other words, the expectation that group D might not manifest distinctive expenditure patterns finds little support when a discriminant analysis is performed on its members. Even if this finding is an artifact of the operation being performed, it is nonetheless

PAGE 171

152 Table 4-10 Discriminant Analysis of Percentage of Total Expenditures By Functional Category Within Selected SES Groups City Name Cocoa Beach Rockledge Cape Coral Bartow Ft. W a i ton Beach Melbourne Cocoa Titusville Boca Raton Hialeah Gainesville Tallahassee Rank Distance

PAGE 172

Table 4-10 Continued 153 City Name Rank Distance From Centroid of SES Group Posterior Probability Membership SES Group Expenditure Group Posterior Probability Expenditure Group Hallandale

PAGE 173

154 of substantial interest that differences do appear. An examination of the discriminant functions can lend meaning to the findings. The aspects of the discriminant function which most differentiate the expenditure patterns in SES group F are service expenditures and newer public works. Gainesville and Tallahassee are both a considerable distance from the other cities in the SES group. They score highly on the newer public works dimension. On the other hand, the set of cities including Cocoa Beach, Rockledge, Cape Coral, Bartow and Ft. Walton Beach are distinguished by their expenditures on the service dimension. The remaining cities display an expenditure pattern that appears to be a composite of these two dimensions. Cities such as Melbourne, Cocoa, Titusville, Boca Raton, and Hialeah evidence a mixed pattern of expenditures, the spending priorities are a balance of service and newer public works. Of course, these interpretations rely on the involvement of expenditure factors in the discriminant space, and they do not catch the full impact of the various factors in the discriminant solution. The service dimension acts to differentiate cities in group D as well. Along with human resource factor, the provision of service dominates the discriminant functions. While the group of cities which includes Wilton Manors loads highest on the human resource dimension, the difference among the expenditure groups on this factor is small. The real difference is made by the service dimension. Miramar and Pensacola are the cities most oriented toward service expenditures, while Key West, Panama City, Coral Gables and Miami Beach are less inclined to spend such sums on service factors. The large cities of group C manifest a different pattern. Electric utility expenditures and specialized older public works act as the most

PAGE 174

155 powerful factors in the discriminant functions. Ft. Lauderdale, Boynton Beach, and Miami are not involved in electrical generation. Moreover, their expenditures on hospitals, a component of the electric utility factor, are typically less than those of the subset including St. Petersburg. Jacksonville and Orlando own and operate their own generating systems, and this helps differentiate their spending patterns from other cities in this socioeconomic group. And although none of the other cities clustered together with Jacksonville and Orlando engage in such utility operations, they do score highly on the utility factor because of expenditures in areas subsumed under this label. The cities of Riviera Beach, West Palm Beach, Hollywood, North Miami Beach, St. Petersburg, and Ft. Meyers load highly on the specialized older public works factor. This is indicative of their operation of water terminals or port facilities, and their involvement in the provision of gas. This factor, more than any other, acts to separate these cities from the group including Jacksonville. The most optimistic interpretation of these findings argues for the differences between relauively similar socioeconomic municipal settings in their expenditure policies. A less sanguine perspective might question this attempt to circumvent the problem of demonstrating the equality of dispersions between SES groups and expenditure clusters. In truth, the technique is problematic. Unfortunately, it is the only available handle on this formidable problem. Thus, a caution is in order, but the tenor of this investigation points towards the importance of factors not included in the socioeconomic environment. The socioeconomic environment of a city is certainly not a necessary condition for the pursuit of a policy. It may well be a sufficient condition. However, the nature of this sufficiency is left unspecified. 4

PAGE 175

156 Summary I began this chapter »ith an assumption: socioeconomic characteristics were conceived as constraints upon urban public policy. This position, it was thought, could torn the basis of . ^^ ^^ ^ at the detailed investigation of policy formation involving the city council. If socioeconomic characteristics are the occasion of policy, controlling such conditions should allow for the consideration of the distinctively political, factors shaping public policy. To this end, I classified 78 Florida cities with popalatioas of 10,000 or more into six distinct and relatively homogeneous clusters. These clusters will be used in an analysis of the manner in which city conncils help mold municipal public policy. Of course, it is not entirely obvious that the relationship between socioeconomic factors and public policy, marked by the demographic perspective, should be abandoned so easily. To check this presumption, 1 took advantage of the socioeconomic groups as a means of ferreting the oonnection between socioeconomic factors aad a city's expenditures. A s-gle SES cluster should manifest a single expenditure pattern. In fact «* of the SHS groups investigated displays a varied set of expenditure patterns. Beyond lending credence to the assumption standing behind this chapter, these results set the stage for an i„ q „iry into the manner of Politics which might produce such differences. I" the next chapter, I win begln to examine ^ ^ ^^ ^ the confluent molding municipal policy making. , „ lu focus on ^ „_ oris throughout Florida and their members.

PAGE 176

157 Notes 1 See Donald Campbell and Julian Stanley, Experimental and Quasi-Experimental Designs for Research (Chicago: Rand McNally, 1963). ' ~~ ' ~ 2 Heinz Eulau, "Policy Making in American Cities: Comparisons in a Quasi-Logitudinal, QuasiExperimental Design," General Learning Press Reprint (New York: General Learning, 1971), p. 3. 3 Ibid. 4 Lewis Mumford, The Culture of Cities (New York: Harcourt, Brace and Jonovich, 1938) . 5 R.J. Rummel, "Understanding Factor Analysis," Journal of Conflict Resolution, Vol. XI (December, 1967), pp. 444-480. 6 Harry Harmon, Modern Factor Analysis (2nd ed. ; Chicago: University of Chicago, 1967), p. 311. y 1 Richard L. Sutton, et al. , "American City Types: Towards a More ^rf^ tlC Urban Stud Y>" Urban Affairs Quarterly , Vol. IX (March, 1974), pp. 369-401. ~ 8 Ibid . 9 See Maurice Tatsuoka, Multivariate Analysis: Techniques fo r Educat zonal and Psychologica l Research ^ (New York: Wiley, 1971). Discriminant analysis proceeds m the following manner: working from an a priori determined numoer of groups and group members, a linear combination of the original predicator variables can be found which exhibits large differences in the group means. Unlike principal components analysis, however, the various dimensions in discriminant analysis are not likely to be orthogonal, but rather represent a best solution (see Tatsuoka, p. 163). In determining the initial groupings of cities for a discriminant analysis, the researcher assigns to each city a probability of being in a particular grouping. This probability along with the positions of each city relative to the others on a linear function enables one to determine the group to which the city most likely belongs. In short, given an a priori probability of group membership and a set of discriminant functions, the posterior probability of group membership can be calculated (see Tatsuoka, p. 230). Following ^his process, the homogeneity and difference of each of the groups can be tested using variants of the multivariate analysis of variance, namely, Wilk's Lambda, ana box s M. If a grouping proves to consist of a set of statistically similar cities, and _ is simultaneously, statistically dissimilar from each other grouping: then a distinctive environment has been established for each city. Each group ot cities can be said to have distinctive and different environments

PAGE 177

158 10 Robert Alford, "Critical Evaluation of the Principle of City Clas•fication," City Classification Handbook: Methods and Applications , ed. Brian Berry (New York: , Wiley, 1972), pp. 311-359. 11 Ibid . , p. 331. 12 Terry Clark, "Urban Typologies and Political Outputs," City Classi ficat ion Handbook: Methods and Applications , ed. Brian Berry (New York: Wiley, 1972T7~P^" 164 13 "Factor analysis of cities," Robert Alford has written, "suffer from a number of arbitrary research decisions. These arbitrary decisions concern (a) the unit of analysis, (b) the particular cases selected, (c) the attributes or variables chosen for inclusion, and (d) the factor names that are assigned" (fn., Alford, op. cit . , p. 332). It may appear many of these decisions are necessarily arbitrary, in any research effort; but, in fact, Alford has a particular sense in mind. By arbitrary, Alford means atheoretical or without conceptual justification. In this study, many of these decisions have already been made; but in each instance, a justification was advanced. Municipalities with 10,000 or more in population have been included to insure a certain degree of complexity. Six groups of socioeconomic variables have been selected as representative of the theoretical range of factors included in municipal environments. 14 Tatsuoka, op. cit . , p. 147. 15 The factors themselves tell us something about the types of environmental constraints to be found when the socioeconomic environment of a city is conceived of in a specific manner. However, these functions are only a prelude to further analysis, which is predicated on the notion that the cities can be represented in the dimensionality of the factor solution. Factor scores are the manifestation of this transformation. They are composite scores for each city on each factor derived by considering the factor loadings of each variable on a factor and the original value of the variable (see Harmon, op. cit .). As noted in the interpretation of the various factors, dimensions account for differential amounts of the variance of the original data. If the raw factor scores were to be used in calculations, the assumption would have een made that the various dimensions were equally important. To avoid this ifficulty, the factor scores were weighted by the square roots of their respective eigenvalues. This weighted set of factor scores assigns more importance to city size than to the factors such as the Income of Black Families. 16 Following the techniques already outlined, factor scores were comPu ed for each of the seventy-eight cities. As before, these factor scores ere then weighted by the square root of their respective eigenvalues. These res were then ready to be used in an examination of the environmental conamt linkages. Notice this procedure lends considerable weight to human source expenditures and very little to electric utilities. ,.„ Tne results of the analysis may be obtained from the author upon request. u-n Ronald Wonnacott and Thomas Wonnacott, Econometrics (New York: Wlle y, 1970), PP . 59-61.

PAGE 178

159 19 Estimates are thought to be significant if the coefficient is at ast twice its standard error. Equation constants are not reported in this • stance, nor are the F values because the purpose of the analysis is to check tn e pattern of linkages not to explain expenditures in detail. Only ten qps factors were used in order to parallel the analysis of political structure ariables, to avoid over loading variance. 20 Negative depressions occur in five instances indicating controls acted to increase the association. While anomalous, these changes are uniformly small, and within the tolerance of error. Nonetheless, these results a re symptomatic of the difficulties involved when using a sub-set of variables in deriving the multiple correlation coefficient. Thus the results should be taken as tentative evidence for the dominance of socioeconomic measures. 21 Q-factor analysis was not used as the single classification technique since a good portion of the variance will normally be neglected or lost in factor reduction of the space. Moreover, an inspection of the factor solution suggests the factor-city involvement is not unambiguous. As with most factor solution, interpretation of factor is still difficult. Thus the Qtechnique was used as a preliminary sort — as a means of checking upon the visual groupings and providing a set of cities for further analysis. 22 A successful classification procedure will produce groups that are internally homogeneous and yet distinct from one another. In order to check the validity of the classification of cities into different socioeconomic groups, two tests were performed. Box's M was calculated in order to check for equal group dispersions, while Wilk's Lambda was used to test the significance of the overall difference among the seceral group centroids. For a discussion of these tests see William W. Cooley and Paul R. Lohnes , Multivariate Data Analysis (New York: Wiley, 1971), and Tatsuoka, op. cit . For both tests, the statistics demonstrate the groups to be both homogeneous and distinct. Wilk's Lambda = .0126, with 10, 5, 72 degrees of freedom, and F = 9.344 significant at the .005 level. Box's M = 0.391, F = .0093 significant at .05 level. 23 Paul Hart, "The Concept of Natural Area," Studies in Human Ecology , ed. George Theodorson (Evanston, 111.: Row, Peterson, 1961), pp. 104-107. 24 See Cooley and Lohnes, op. cit . , p. 12, for a general discussion of the problem. The Lambda statistic is not significant at .01 level with F = 1.239 and 30, 270.00 degrees of freedom. 25 The relative expenditure grouping was derived from factor scores and scattergram inspection. In fact, this a priori grouping proved 91.1 Percent successful in predicting the expenditure groups. Moreover, the groupings do have a certain statistical validity. Consider the following: Wilk's Lambda = .007, F = 9.090 with degrees of freedom = 65, 287.49. 26 Of course, this tact has its difficulties. Specifically, if any set °i cities is examined closely, they can be differentiated in some fashion. A Pool of variance can be carved in any number of ways. Thus, there is the danger hat in looking for differences, I am creating the situation in which they must e found. Later in this work, I will examine this possibility by assuming another perspective on the problem. See Chapter Nine.

PAGE 179

160 27 This theme is developed throughout empirical literature on political behavior. See, for example, Lester Milbrath, Political Participation (ChicagoRan d McNally, 1965). -

PAGE 180

CHAPTER FIVE FLORIDA COUNCIL MEMBERS AND THEIR COUNCILS Unlike demographic analysis, exchange theory focuses upon individuals and the choices they make. Each of the over 400 council members who sits on local councils and commissions in Florida cities with populations numbering more than 10,000 participate in the formulation of policy. Exchange analysis begins with a consideration of their attitudes, their attributes, and their behavior. A questionnaire was sent to each council member in 78 cities. It attempted to monitor a member's background characteristics, ideological predispositions, goals, and perceptions about different facets of council life. Each member was asked to comment on significant community problems, important community groups, council performance, the interrelationships among the members, and the decision-making processes used by the council. Additionally, members had an opportunity to discuss matters of policy which had appeared on the council's agenda in the year before the questionnaire was administered. These responses form the basis of my analysis. They also provide a description of council members and council politics in 61 Florida cities. There are more than 2000 cities in the United States with populations greater than 10,000; the 61 studied here are but a small proportion of those. Where is the meaning in such a narrowly conceived examination? It lies in the responses of the many council members; it emerges as an image by which the entire enterprise is framed. Drawing the pro161

PAGE 181

162 file of the council members and their councils provides the parallax necessary to gauge members' relationship to the national norms. Thus, before beginning the analysis promised in the previous chapter, I will re-tell the story written by council members across the state. I will use such items as age, sex, occupation and the like as parameters to judge this sample. I will introduce selected characteristics to be used throughout subsequent analysis. Finally, I will describe the manner in which councils in the six socioeconomic clusters on these characteristics also will be considered. In particular, special attention will be given to the SES group consisting of the most affluent cities — SES group F. The Typical Florida Council Member By and large, the Florida council members are not dissimilar from their colleagues throughout the nation. Members are predominantly white, male, businessmen in their forties. There is no statistical difference between the proportion of professionals and businessmen in the Florida sample and that appearing in a National League of Cities survey released in 1974. 1 The fraction of council members in their forties is not dissimilar from the national average. Florida members are much better educated than those sampled by the National League of Cities. But among Florida council members, graduate educational experience or graduate degrees are less often found than in the national sample. Among those members responding to my questionnaire, there are two notable differences from the results of the national sample. First, there is a greater proportion of women in the Florida sample (9.3%) than in councils throughout the United States. Second, there are more blacks in the Florida councils (7.4%) than might be expected.

PAGE 182

163 Like their national counterparts, Florida council members wish to continue in office. When asked about the possibility of not running for any future office and of returning to private life, only 25 percent indicated this would be their most likely decision. And although a substantial proportion (17.2%) signified they intended to run for a state or national affice, a majority of members (55.5%) attested to their devotion to local politics. They indicated a willingness to serve for two or more future terms of the council. This willingness coexists with the feeling on the part of over 42 percent of the members that being on the council had hindered the pursuit of their private occupation. Florida council members handle citizen complaints over matters prevalent in cities nationwide. Zoning appears with street and highway matters as the two most frequently mentioned areas of individual citizen contact. As in the national sample, garbage and sewer complaints, as well as property taxes, figure highly in the day-to-day activities involving a council member with his constituents. However, Florida members are contacted more frequently on matters of planning, area growth and public utilities than the national sample would lead one to suspect. This difference may reflect the timing of the Florida questionnaire. Administered in late 1974 and early 1975, the energy crisis forced a sharp rate of increase in public utilities and this prompted a severe public reaction. Florida members differ with their national colleagues in their evaluation of the importance of the council for the initiation of policy. Over 58 percent of the members responding indicate the council has a strong or very strong role in the initiation of policy. Moreover, most members believe their ability to generate new proposals, independent of

PAGE 183

164 the manager or mayor, is quite good (56.8%). In short, Florida city council members believe their council to be a potent institution in the community. Indeed, the council is listed as the most important group or organization in the community by nearly 60 percent of the council members. This evaluation of the council's importance in urban decision making hardly accords with much of the literature on community power. However, at least one comparative study of the various forces working in urban politics has discovered the council can be, on occasion, the pre-eminent o group xn city decision making. Florida council members heartily ratify this conclusion as witness the ranking arrayed in Figure 5-1. Only the manager/mayor compares favorably to the influence of the council in the evaluation of council members. But the city manager and appointive city officials are important to members as information sources. In some respects the evaluation by Florida council members of significant urban problems is very similar to that of those council members surveyed by NLC. Nationwide community development concerns are viewed as pressing problems, and in Florida, aspects of this broad category such as area growth and zoning are identified as troublesome. Human resource matters typically draw little attention from members across the country; this is the case in Florida as well. Health care, racial and ethnic relations, and poverty are ranked among the least serious problems facing Florida cities. This neglect is graphically portrayed in Figure 5-2 . One of the most significant differences between the perceptions of council members throughout the country and those in Florida comes in their awareness and sensitivity toward problems involving growth, plan-

PAGE 184

r 165 4.5 SELDOM IMPORTANT 3.5 OCCASIONALLY IMPORTANT 2.5 VERY IMPORTANT 1.5 School Employees Motel/Hotel Associations Bar Association •Labor Unions Welfare Organizations Health Interests TV/Radio Stations Minority & Ethnic Groups Church Leaders Retail Merchants Municipal Employees Bankers & Executives of Financial Institutions 'Newspapers Real Estate & Land Developers 'Chamber of Commerce 'Heads of Local Government Agencies Manager/Mayor City Council figure 5-1. Mean Scores For All Council Members on Evaluation of Community Influence of Groups and Organizations

PAGE 185

166 N0T SERIOUS 3.0 2.5 FAIRLY SERIOUS 2.0 1.5 VERY SERIOUS 1.0 JL Racial and Ethnic Relations -Poverty -Health Care and Hospitals Urban Redevelopment & Housing Recreation & Entertainment -Local-Federal Governmental Relations •Industrial & Economic Development -Education •Local-State Governmental Relations, Transportation & Tax Base Traffic 'Employment Opportunities •Public Utilities •Governmental Finance Crime Environmental Concerns Zoning •Area Growth Figure 5-2. Mean Scores For All Council Members on Evaluation of the Seriousness of Community Problems

PAGE 186

167 ning, environmental concerns, and natural resources. In a survey of Florida citizens undertaken just eight months before my questionnaire was administered, environmental concerns, overpopulation, and excessive growth were among the most frequently cited problems facing the state. Only the energy crisis and inflation came close to matching the importance of these issues. For the most part, this study is not concerned with the direct connection between the beliefs of constituencies and the positions of their representatives. In later chapters I will attempt to draw a rough connection between citizen action and policy responses. But the immediate point is somewhat different. It is that Florida council members evaluate the problems facing them somewhat differently than their colleagues across the country. However, the members of the 61 Florida city councils do share many of the attributes of council members found in the NLC survey. Thus, for all their differences, Florida council members do not appear anomalous; they roughly match their nationwide counterparts. Florida Council Members: Characteristics and Perceptions The exchange theoretical fragment offered in Chapter two was intended to detail selected analytical features of council life bearing on the formulation of public policy. These features focus on the individual council member and his response to the external and decision-making costs attendant to his tenure. Aside from individual background characteristics such as age, education, and occupation which capture a member's life experiences in an indirect fashion, my earlier argument suggests such personal characteristics as predispositions and goals should be important in fashioning a member's policy positions. Beyond these, the exchange theoretical argument holds that the patterns of interaction established

PAGE 187

168 among members are also important for policy making. I consider several measures of such council characteristics as viewed by the members in this section. Of course, members are also affected by factors outside the council itself. Each member's perception of the influential groups in his community together with his reliance on a particular set of information sources acts ±s> shape his policy positions. This section introduces measures designed to tap such perceptions. These considerations were reflected in the questions asked of Florida council members. Their responses formed the basis for an extensive set of scales dealing with member characteristics and perceptions (see Appendix A) . The cognitive and affective aspects of a member's perception and understanding of his environment are structured by his predispositions or "primitive beliefs." These predispositions can act to fundamentally change the manner in which the council operates. 3 In order to gauge the predispositions of each of the city council members in the Florida cities, each respondent was administered a series of items related to various components of their beliefs. Four separate predisposition measures emerged -from an analysis of the council members' answers (see Appendix A). These scales measure political, social, religious, and total conservatism. The political conservatism scale includes member feelings on items such as welfare, socialized medicine, and disarmament. Florida council members are, by and large, conservative. Almost 50 percent gave responses best classified as conservative. A little less than a quarter of the council members can be identified as liberals on this political dimension. These results can be seen in Table 5-1 where political conservatism is listed by the mnemonic CONSPOL. 4

PAGE 188

169 CJ c QJ rJ

PAGE 189

170 0) 3 3 H 4J c o u I (1) cd H >1 a B QJ 3 o1 QJ M a) 23 r-( CU > m Or^-rnomcNroci^Oi-H >>. QJ r-\ 23 4J

PAGE 190

171 o s CO 3 cr QJ S-i Pm (^iD^onoinovOsfw i— i i^ m o -JWiOHNoinoMnMnNvfvoo oj qj

PAGE 191

172 o 3 Q) 3 C7l 0) J-4 Ph 3 > J OO^COrHCM NOn On rH * h in no rH CO cm cm > QJ Z CO !>> E -H o to O T3 tf E >H cj rH QJ U < CO O C r^> 0) rH 4J 4J M-H c o QJ 3 0) cr +j QJ -H M 3 FH o* O CO N to E >> 0) > QJ CO z s c QJ CO CO U . -H O 14-1 rH 00 C 01 3 CT i— I QJ QJ S-J CO Pn S o 13 ~* m o r-cm e 3 •H 13 -3 O QJ rJ S •H VO -3" O lo H hj a 3s ONO H oo r^ -ooo E 3 . • H -£ & T3 6C O QJ -H >-J S m — < a> o CT\ vO .33 QJ 4-1 o to QJ E •H •U H QJ QJ E > o a) co Z -a QJ 3 3 3 O O I 3 O z w o H IH J P-4 CO o Q M H M (J Ph CO 5S Ph H J Ph CO z o u Q M o 5 W z o rJ H Z w CO z o u -a co H QJ E CO Z CO •H U CO >

PAGE 192

173 While political predispositions are likely to be a powerful component in a member's predispositions toward public policy, there are other beliefs which will likely impact upon policy decisions. The moral cast in which public policy is often promulgated augurs for a conception of a member's predispositions which includes his orientation to religious matters. This is especially true in Florida, a state cleaved by significant denominational differences. The Christians of north Florida and its panhandle are typically Baptist, while southeastern Florida is predominantly Roman Catholic. In central Florida, no religious denomination claims a majority of the churchgoers in any county, but Southern Baptist adherents outnumber those of any other Christian church. These cleavages, while pronounced, appear to have little impact on member predispositions. Although the respondents to the questionnaire come from cities throughout the state, the members exhibit a consistently conservative religious bent. As seen in Table 5-1 , 60 percent of all council members were rated as conservatives on the measure of religious conservatism (CONSEFL) . Only a small portion of the members evinced liberal responses (see Appendix A for discussion of scale) . If the council members indicate a distinctly conservative turn on religious and political matters, their attitudes on social issues such as divorce and different types of population control are considerably more liberal. Fully 72.9 percent of all council members indicated acceptance of such practices and were classed as liberals on the social conservatism scale (CONSSOC). However, it should be noted that this liberalism is limited to social areas in which Protestant theology has sanctioned such activity. Examining related items which were not included in the social conservatism (CONSSOC) scale, such as reactions

PAGE 193

174 toward parental authority and evolution theory, reveal a slightly different orientation. Almost every member affirms the notion of parental authority, while little more than half "believe in" evolution theory. In short, the items included on the social conservatism scale appear to tap aspects of members' belief systems on which there are socially acceptable answers . Social conservatism is a scale with little variation. For this reason, its use throughout this study will be quite limited. The political, social, and religious conservatism scales were derived from items included in a more general list associated with an often-used scale (see Appendix A). However, in computing these scales considerable information, provided by the deleted items, was lost. Moreover, these scales emphasize distinct aspects of each member's belief system rather than capturing an overall evaluation of its content. The summary or total conservatism scale was constructed to remedy these difficulties (see Appendix A). When the scores of the members are arrayed on this measure, slightly more than 46 percent are rated as conservative. Forty-three percent are classed as moderates, and only a small proportion (10.7%) are liberals. As I have constructed it, this scale (CONSTOT) indicates a more thorough-going conservatism among Florida council members than registered on the political conservatism (CONSPOL) scale. While the proportion of members classed conservatives is relatively the same on both scales, the numbers of liberals identified by the summary measure drops noticeably. Only a little over ID percent of all members are classed as liberals on the summary measure. A member's predispositions may well affect interpersonal relationships among the council members. However, predispositions alone are not sufficient to affect the manner in which members interact. The exchange

PAGE 194

175 analytic view developed earlier argues individual behavior is conditioned through strategic evaluations of the context and the possibilities for certain rewards. Other factors may act to determine costs in the decision-making situations of those emanating from external sources. These factors emerge from the interaction of the members over time and include the way in which members take sustenance from their environment, and the patterns of interaction which have been worked between the members. However, gauging the character and relative importance of each factor for a given council is quite difficult. In this study, I have relied upon the perceptions of the members themselves. These perceptions are assumed to capture a wealth of previous experiences with each council's actions. And while it can be supposed no member will accurately detail council life in a fashion identical to a fellow member, each individual's perception provides a bit of valuable knowledge. According to exchange theoretical considerations, the way in which a member views his fellow members is likely to be significant precisely because it is upon such perceptions that his behavior within the group will proceed. In order to measure selected aspects of council life bearing on policy making, several questions were asked of each member. From their responses, seven scales dealing with different council relation characteristics were constructed. These scales include measures of council participation, stratification, polarization, intimacy, autonomy, viscidity, and pressure perceived emanating from the community (see Appendix A) . The city council members believe most members participate actively in the affairs of the council and that the council is seldom without significant amounts of business before it. Over 70 percent of all members rated the involvement of fellow members as high. This result is reflec-

PAGE 195

176 ted in the council participation. This participation is not necessarily the product of similar minds working toward a common goal, however. While the viscidity— the degree to which members of the council function as a unit— reported by the members if fairly high, this harmony is punctuated by a substantial number of members who perceive their situations differently. The registered amount of council polarization— a divergence or ambiguity in member goals— is considerable. Almost a quarter of all members report high degrees of polarization (POLARIZE). This polarization is accompanied by reports of considerable stratification among the members. The stratification scale measures the differentiation of members according to experience or influence criteria. Nearly 20 percent of all members register high degrees of council stratification (STRATIF) . City councils are often depicted as small groups of individuals working intimately with one another in league with potent groups within the community on problems of mutual concern. 6 In Florida, this image is not as well defined. Only 4.7 percent of the members report intimacy (INTIMACY) among the members is high. This low figure may reflect a reluctance of Florida council members to associate closely with their colleagues for fear of violating the state's strong open meetings law. However, it should be noted that the great proportion of members report moderate levels of familiarity, and that 40 percent of all members could agree to items identified with council intimacy. Forty-seven percent of the Florida council members could either agree or strongly agree to items indicating the council was under considerable pressure from the community (PRESSURE). For 17 percent of the members, this pressure was partially determined by their agreement that there were individuals in the community who determine what the council does.

PAGE 196

177 Yet most council members were moderate in their appraisal of community influences. Indeed, most members (65.2%) believed the autonomy of the council from the influence of community interests was considerable. A majority of members rate their council's autonomy (AUTONOMY) as high. Repeatedly, council members affirm this judgment and bear witness to their own ability to influence policy independent of community forces. Community groups and organizations can be important on occasion, but their importance pales before the influence of governmental agencies within the community. The 18 community groups and organizations which each member was asked to evaluate comprise five distinct types of community organizational factors (see Appendix A). Governmental agencies are rated as the most influential (GOVTINF) in community decision making. A mere fraction of the council members (0.8%) see governmental influences as seldom important, while a convincing majority (62.4%) avow otherwise. Business (BUSINESS) and media (MEDIA) oriented groups are evaluated substantially below governmental groups in community influence by the council members. People-oriented groups (PEOPLE) are rated as occasionally important with about the same frequency as business and media groups, but their ability to dominate the perceptions of the city council members appears small. Less than five percent of all members indicate people-oriented groups such as teachers' organizations or welfare groups are often important in influencing community decision making. As small a proportion as this is, it still exceeds the number of members suggesting groups such as unions or local bar associations are important. Such groups have small influence (SMALLINF) in the eyes of most members. Council members not only perceive governmental agencies as the most important influences in community decisions, they also report governmen-

PAGE 197

178 tal information cues (GOVTQ) as their most important sources of information about affairs in their cities. Citizen cues (CITIZENQ) are afforded a certain credibility as an information source by a considerable number of the council members, but for over 18 percent of the members, they are seldom important. Nonetheless, most council members maintain citizenoriented activities occupy .a considerable portion of their time (see CITIZEN scale) . Citizen activities such as providing assistance on specific problems may occupy a good portion of a council member's time, but the problems which council members perceive as most serious typically have less to do with human resource issues (HUMPROB) (14.9%) than with community growth (GROWPROB) (45%) . The problem perceptions of the council members vary considerably from issue category to issue category. And although growth problems are clearly seen as the most serious, no problem area is without those who suggest it is important. In each case, nearly half the members are willing to accede the problems are, at least, fairly serious. However, members indicate financial (FINPROB) problems loom as most serious for their communities once growth problems are accounted for. Perhaps, it is such a broadly conceived set of concerns which, more than anything else, stimulated members toward disagreement. Eighty-six percent of all members report frequent council disagreements over particular issues appearing on their agenda (see S PL IT PAR) . But many council members apparently would disagree with this simple pronouncement. For them, council disagreements, more often than not, have their roots in the character of persistent patterns of council behavior (see SPLITGEN) . On the other hand, members report that disagreements resulting from pure happenstance or idiosyncratic conflict are virtually nonexistent (2.3%). Nor do members suggest conflicts are avoided to preserve the working har-

PAGE 198

179 mony of the council. The amount of reported bargaining is quite low (see AVOIDCON) . Rather they claim that agreements over policy issues is the product of hard work in fashioning a form of amicable agreement (see CONSENT) or is the natural outgrowth of shared consensus (see ALLONE) . Surprisingly, most members suggest the informal ties that are established with continued service on the council do little to promote agreement on policy issues (see INFORMAL) . Overall, the council members paint a rather distinctive picture of themselves and city councils throughout the state. But the portrait of councils as intimate groups, agreeing on issue after issue, in the face of overwhelming community forces is called into question by these responses. Florida members insist councils are important; agreement is not ubiquitous. Moreover, over three-quarters of all members believe their council performs its functions either well or quite well. These statements may appear self-serving, and this view may not withstand further scrutiny such as bivariate analysis. However, among council members who indicate a propensity to disagree on almost any given issue, there is agreement. that, more often than not, city councils are potent. The Characteristics of Florida City Councils Sixty-one city councils are represented in the responses of the 173 council members. These councils form 61 different contexts for member behavior. The many contexts introduce an interesting dilemma into the analysis. The survey data are at once subject to reductionist and constructive difficulties. Taken as a complete data set with each member considered apart from his particular council context, as done above, a systematic bias can creep into the analysis. Certain contexts are over-represented. If, on the other hand, the responses themselves are to be treated

PAGE 199

180 in some statistical fashion in order to construct meaningful measures of council proclivities for each context, then problems can arise associated with the way in which the responses are put together. This is especially true when each context consists of just a few possible responses as is the case with councils. The possibility of random error or idiosyncratic responses affecting the picture of any given council is magnified by the small number of possible responses. This problem is aggravated by the differential response rate of the 61 councils represented in the sample. Only one member responded from 11 different councils; in 27 councils three or more members responded. Constructing the properties of a context from a single member's responses requires certain assumptions be made about the weighting of individual responses and the validity of engaging in constructive exercises. Eulau and Prewitt write, "Computation was used to develop councillevel measures whenever the individual interview could be meaningfully transformed through mathematical manipulation." 8 The point behind such tactics lies in obtaining a view of council properties. This could be done through indepth, extensive observation by skilled researchers using well-established criteria. However, even this mode of research would entail considerable problems and run the risk of researcher bias. Of course, constructing council properties through computation involves the same danger, but this risk should decrease as the number of respondents increases. 9 Moreover, the emphasis placed on strategic interaction within exchange theory lends some credibility to an attempt at pooling the responses of the members of a council. There is a sense in which the "typical" member is revealed in the arithmetic mean, or divergence of opinion is captured by standard deviations. The perceptual, subjective emphasis of interpersonal relations makes such additive assumptions reasonable.

PAGE 200

181 The data used in this section attempt to draw the broad outlines of council characteristics. Profiles are constructed from the responses of individual council members. Council-level information is presented by the SES groupings determined in Chapter three, each of which forms a relatively homogeneous socioeconomic milieu. If the demographic model examined in the previous chapter were correct, differences in council characteristics might exist and bear some theoretical relation to the character of the several groups. Otherwise, the overlap between the councils of different clusters might be substantial. In addition, I will report the aggregated figures for the 12 councils in SES group F — the high SES cities. In Chapter nine, I will examine in more detail the notion that councils with similar socioeconomic characteristics display different political characteristics. In the intervening chapters, I will investigate with particular care the patterns manifested among the councils of SES group F because the high response rate of council members in this group (81.6%) lends the constructed council measures increased reliability. When a one-way analysis of variance is performed on the many measures of council member perceptions and predispositions, as aggregated to form constructed councils, the socioeconomic effects of the six distinct clusters can be gauged. At least, it can be determined whether the socioeconomic groups help differentiate each of the sets of councils on the political measures. According to the demographic model, the means of the six groups on such measures should be different from one another. A quick glance at Table 5-2 shows the means of the councils in each of the clusters are not typically different from one another at the .05 level. Of the 34 measures, only five are separated by the socioeconomic groupings. The means of the groups overlap considerably. This suggests, for these political measures, the demographic model may not work well.

PAGE 201

182 1.500

PAGE 202

183 Table 5-2 Continued Variable Name F-Value Significance of F G. Decision Modes Generalized Policy Disagreements (SPLITGEN) 1.406 Idiosyncratic Policy Disagreements (SPLITIDO) 1.018 Particularized Policy Disagreements (SPLITPAR) .130 Bargaining or Logrolling (AVOIDCON) 1.796 Informal Patterns of Agreement (INFORMAL) 1.000 Consensual Patterns of Agreement (ALLONE) 1.237 Amicable Agreement (CONSENT) .986 .237 .389 .999 .128 .427 .304 .999

PAGE 203

184 Socioeconomic effects do appear to distinguish councils on a few measures. For example, the religious conservatism displayed by members of councils in the various groups are different from one another. A multiple classification analysis of this finding shows it is in the low income, demographically older, and more rural cities of group B that religious conservatism is most marked, while in the group F — high SES cities — this effect is most depressed. Socioeconomic effects are also registered in the perceptions of members with respect to the influence of business and typically small influence type groups. Business groups are seen as most important in group F cities, while in the relatively affluent, service-oriented suburban communities of group E business groups are afforded much less importance. Groups such as labor unions, bar associations, and motel/hotel associations which are normally accorded little importance by members throughout the state are most important in the larger metropolitan cities of group C. Somewhat surprisingly, the cities in group B, the low income, rural communities have members which evaluate these groups highly as well (second ranked deviation) . Once again, however, it is the councils in group E which extend least importance to such groups. Only two other measures are systematically differentiated by the effects of socioeconomic conditions. The perception of human resource problems by council members is greatest in the large cities of group C. In fact, these cities account for most of the recognition of such problems. And again, it is the cities of group E who have members most insensitive to this measure. The registered polarization of council members almost follows this pattern, but here the order of SES cluster contribution changes slightly. The group F cities, the high SES cities, evidence the

PAGE 204

185 most polarization, but the large cities of group C follow closely in member perception of polarization. The cities of group E manifest little polarization, but on this measure, the low income cities of group B evidence the least polarization. Although most councils, when they are grouped according to the socioeconomic criteria, do not manifest statistically different features on political measures, differences do exist. And these differences can be displayed in convenient fashion using graphical techniques. For example, the councils in the cities of group B were shown to be the most religiously conservative in the multiple classification analysis. As characterized in the discriminant analysis, these are the cities with the lowest scores on the SES variables. However, the conservatism of the councils in group B is more thorough than the analysis of variance would indicate. An examination of Figure 5-3 reveals SES group B councils are typically the most conservative when compared to the other five groups. This is not incredibly surprising. Conservative attitudes have typically been linked to rural, low income areas. But the nature of this link has taken several conceptual forms. Only in linear development theory is there an ineluctable relationship between the socioeconomic status of a community and the attitudes of its leadership. And, in fact, the rankings of the various groupings do differ slightly across the scales. For example, the high SES cities of group F have the most religiously liberal councils, but politically the councils in other groups are more liberal. Of course, a pattern can be detected in the figures on council predispositions. The functional size of the city appears related to the attitudes of the members of councils in the various socioeconomic groups. The largest cities found in group C, are ranked as most liberal on the

PAGE 205

186 h pq w Q < fn u \ I I I 1/ H O H CO z O u ttl a. O u o w CO M-l o pq W < Q U w J W Pi CO Z O u •H CJ o CJ u o pq p W fK < u \ w \ \ \ o CM CO z o u en aj rH CO CJ CO e CD •H U CO > U Q) c o o a o en C CO a) S aj N > H Eh w CO o u hJ pq u o z d) S-i 3 00

PAGE 206

187 political conservatism scale and on the summary measure. They stand just behind the group F cities on the religious conservatism scale. The diversity evinced in these cities may favor the election of liberal council members. But in Florida, this conclusion is mitigated by the religious conservatism which marks most of the council members throughout the state. Nonetheless, those cities which have the smallest functional size are relatively conservative across all measures. SES group E cities, perhaps best thought of as planned retirement communities, are uniformly conservative and lack diversity within the community. Socioeconomic conditions appear to affect the types of groups which come to be viewed as important for community decision making. This was the case with business groups and organizations such as labor unions, for members of SES group E who rejected the notion such groups might have import in their cities. This observation and the findings on member predispositions in the several clusters suggest a relationship between the predispositions of a council and the manner in which it takes sustenance from certain community groups and organizations. Figure 5-4 displays a pattern indicative of the form the relationship might take. The SES group E cities exhibit a rather uniform conservatism. Further in relation to council members of other groups, SES group E, the planned retirement communities, have councils where the influence of community groups, any community group, is supposed to be quite low. On the other hand, the more liberal councils of the high SES cities in group F score each of the groups as relatively influential. Of course, the willingness of the members of councils in the group F cities to cite the importance of community organizations and groups

PAGE 207

188 O u to ex p o M CJ CO w co m o •H U c o u o fn U < pq Q W \1// \ Ph m q u < w i / y/ \ w l-J p-< o w PM CO CO w z M CO !=> « 03 (U H cd a CO 0) CJ C a) a a o S-i o Pi o e a) a id a) N •H H rt 6 o Z i-J hJ H < Z 13 >H O OS Pj > M < H O z o i u-i QJ U D GO •H

PAGE 208

189 should not be considered independent of the relative influence assigned each type of group within a SES cluster. That is, although the councils of the high SES cities extend business, people, and governmental agency groups more importance than any other group of councils, these evaluations have to be considered with respect to one another. How important are different groups in the same socioeconomic cluster of city councils? In Figure 5-4 , the scores of the councils in the six clusters have been normalized to provide a means of judging the relative importance of different community groups within the same socioeconomic group. Notice that governmental agencies are seen as the most important community forces in each of the SES clusters. However, while business groups are typically the second most important set of forces in the SES groupings, this relation does not hold for each set of councils. In particular, media groups are extended considerable importance by the councils of groups A and C. Further, the differences between the scores on these normalized scales show the councils of group C to be especially attuned to group influences in general. The range of evaluated importance for the groups depicted in Figure 5^ is quite small in group C; people-oriented organizations are afforded especial attention. In summary, it appears that for each of the SES groups, governmental agency influences are viewed by council members as most important for community decision making. People-oriented groups are given the least credence as potent community forces. However, there are some differences in relative rankings across SES groups. For example, in group D cities business groups edge out media groups in importance, while this situation is reversed for group A cities. In group C cities, the largest cities in functional size, business, people, and media groups are given remark-

PAGE 209

190 ably similar ratings. Governmental institutions are the most important organizations in the planned retirement cities of group E, but group influences in general are not cited with the frequency noted among the other SES clusters. People-oriented groups are burdened with a particular insensitivity in the cities .of group E. The council members in the various SES groups differ considerably in their perception of the problems facing their cities. While these differences are statistically significant in separating the group means only with respect to human resource problems, other notable differences appear between group means on other problem measures. This should not be entirely surprising if it assumed many problems confronting councils will be a function of the set of objective conditions within a city. Socioeconomic factors might well capture such conditions, and thus become surrogates for the problems themselves. For example, an examination of Figure 5-5 shows human resource problems are viewed as most serious by councils in group C, the largest functional size cities. Group C cities also rank highest on the perception of serious financial problems. Both these findings appear reasonable. The cities of SES group C have significant proportions of poor; the retirement programs and bonding projects of these cities pose serious challenges for the financial resources of these cities in the future. Moreover, most of the cities in this group are administering ad valorem tax rates at or near the limit allowed by state law. Hence, their ability to respond to increasing financial demands is severely limited. Council members in SES groups E and F indicate growth related problems pose the most serious problems for their communities. In light of the explosive growth experiences of a number of the cities within these

PAGE 210

191 h o W Ld > CO PQ O Pi Ph O Pi o PQ O Pi Ph > O O PQ S CO O H Pi W CO CD Pu 3 O u o CO w CO o H o § o o n o M-H w QJ H cd o CO C o H j-l pu QJ o H a) Ph e 0) tH ,a o Ph o c cO QJ a) N H iH I ft O z I 0) 60 •H Fxh

PAGE 211

192 SES clusters during the sixties, these perceptions are understandable. For example, in Sunrise the population grew from 7403 to 13,456 in just two years between 1970 and 1972. 13 Other cities such as Gainesville, Melbourne, Ft. Walton Beach, and Cape Coral are experiencing similar growth. Severe service provision problems are noted in two SES groups — A and F. The energy crisis of 1974 had forced the costs of electrical generation to new highs, and this affected cities with public utilities greatly. Leesburg, Kissimmee, Lake Worth, Lakeland, New Smyrna Beach, Bartow, Gainesville, and Tallahassee provide for the provision of electricity to their citizens and use the revenues from service provision to augment their general revenue funds. The dramatic price increases of 1974 could not be passed through to local customers without severe political costs. Local revenues suffered. The energy crisis also affected an increase in the costs associated with water services in cities such as Cocoa. For the members in these cities, service provisions became a major headache. Governmental problems were considered most serious by the council members in group D, cities small in functional size and having relatively old populations. These problems include the relations of the local community with state and federal agencies as well as transportation and traffic problems. As a rule, city councils throughout the state report growth-related problems as the most serious issues confronting their communities. Only the councils in group A report another set of concerns— service provision —to be more important. However, in the cities of group C, the functionally large cities, the range of problems facing the cities and their relative importance for the councils evidence a pattern different from that of the other SES clusters. The councils of this group are likely

PAGE 212

193 to have members who suggest the problems facing their cities are many and serious. Table ^3 presents the normalized scores of the six sets of SES clusters on the council relations scales. The greatest differences among the SES groupings appear on the polarization (POLARIZE) scale. Members in the councils of group F, the high SES cities, are more polarized than any other grouping. They are also less willing to work toward a single goal than any of the other groups. In addition, the group F councils report themselves subject to the most intense community pressures of any group, and indicate the greatest degree of stratification along experience or inter-personal influence criteria. Finally, the highly endowed cities of group F have councils which manifest little intimacy among members and which report relatively lower amounts of autonomy than other council clusters. This pattern of relations squares well with the idea that affluent, highly differentiated cities will produce political environments in which elected representatives will be responsive to a plurality of interests. In some respects, the council members in SES group B, the low SES cities, are quite the reverse of members in group F. They report high degrees of intimacy and considerable autonomy from community influences. On the other measures, however, these cities' councils do not register extreme scores. Rather the cities have moderate amounts of polarization and community pressure. On these measures, the cities of SES group E score lowest. Despite the fact that these planned, retirement communities have experienced great growth in the last decade, this growth has not been translated into political cleavages on the councils. Perhaps this is because their growth has been primarily derivative of the re-

PAGE 213

194 gions in which they reside. Being at the periphery, they have missed many of the problems usually associated with growth. This interpretation, at any rate, is buttressed by the problem perceptions of the council members; the issues acting to divide the councils are less severe than for other communities. Table 5-3 Group Council Means on Council Relations Scales for SES Groups: Normalized to Range 1-10 Scale

PAGE 214

195 actions are not subject to undue control by community groups. Council m embers believe their councils can be relatively autonomous; the scores on the autonomy scale all fall in the top tertial of possible values. The City Councils of SES Group F The demographic model posits the direct relationship between socioeconomic factors and policy outputs. However, in the last chapter it was shown that cities with relatively similar socioeconomic environments exhibit different profiles of expenditures. This provides some evidence that the link featured in the demographic model may be considerably more j -, n t^ fartpolitical factors may intervene complex than that model allows. In fact, political in some crucial way. The socioeconomic milieu of the Florida cities was held constant through the creation of distinctive SES groupings in order to investigate the fashion in which political forces help condition public policy in local governments. Any one of the six SES clusters might have been investigated. However, I hoped to choose cities in which the political actions of the council could impact upon public policy. In addition, the character of the cluster of cities had to be amenable to a practical research effort. The cities of SES group F satisfy both these criteria. First, the socioeocnomic endowment of these cities is such that they are liable to be marked by considerable social differentiation and economic diversity. Moreover, the character of the population, the presence of several different professional groups, different forms of governmental activity, and the higher interest in politics which typically accompanies individuals of high socioeconomic status should make for increased competition in the political arena. An examination of the council characteristics scales suggests this view may be correct. At least, relative to the

PAGE 215

196 councils of the other SES groupings, political competitions appears more sharply drawn in the high SES cities. There is another distinct advantage to group F. It has 12 cities of moderate size whose councils are also characterized by a moderate number of members (Ft. Walton Beach has the largest council). Thus, the scope of the research effort involved in visiting each of these cities and attending council meetings was relatively manageable. The only limiting factor present in the intensive study of this set of cities was the great distance which separates them from each other. For example, it is over 650 miles from Ft. Walton Beach to Hialeah. 14 While the mean score of the councils in group F tested comfortably in the midrange on the political conservatism scale, there is considerable variation among the cities in the cluster on this dimension and others (see Figure 5-6 ) . On each of the scales, however, Cocoa Beach has the most conservative members. Its consistent pattern is matched only by the scores on these dimensions for the councils representing Cape Coral and Ft. Walton Beach, although Rockledge's members evidence a persistent conservative bent. Cocoa, the city just across the Hubert Humphrey Causeway from Cocoa Beach, is rated to have a much more liberal council than its neighbor. In fact, Cocoa's members have the most liberal ranking on the religious conservatism scale. Gainesville has the council with the most liberal political predispositions. But while Gainesville is among the most liberal councils on social matters and on the summary measure of conservatism, it ranks among the most conservative on the religious conservatism scale. Boca Raton displays a similar inconsistency. It is rated as the most liberal coun-

PAGE 217

198 cil on the summary index and scores low on the social and religious conservatism scales. However, its political predispositions are moderate. Hialeah, the largest city in group F and a city located in the heart of the Miami-Dade county metropolitan region, might have been expected to have a council whose members were liberal. Instead, the Hialeah council members take moderate stances on the controversial items used to gauge their predispositions. Once again imputing secularization to a community's leadership on the basis of the differentiation hypothesized to accompany size proves a mistaken tact. Governmental influences dominate the members' perception of those forces within their communities excerising significant impact on decision making. Yet when the patterns of perceptions are investigated more closely, it becomes clear there is substantial variation from city to city. This is graphically portrayed in figure 5-7 . In Gainesville, for example, while governmental agencies are rated as the most important organizations within the community, their pre-eminence is hardly clear. Media, business, and people-oriented groups, all are extended credence as potent organizations within the city. Media groups are actually rated more highly than governmental agencies in Tallahassee. Business groups are rated almost as highly as governmental influences in Ft. Walton Beach, Cape Coral, Bartow, Cocoa, Titusville, and Boca Raton. Perhaps the most surprising evaluation of group influences is found in the perceptions of the Hialeah council members who rate people-oriented interests as less important than that coterie of groups designated as small influence organizations. This pattern is repeated in Cape Coral, but is more understandable in that city. Hialeah is a city whose demographic make-up is in the process of substantial change; it is the largest

PAGE 218

199

PAGE 219

200 city in group F. Its high SES neighborhoods are gradually being abandoned before an influx of Spanish-speaking people. It is a city where people-oriented groups have made significant demands upon the city council in recent years and have affected the traditional dominance of the mayor in local politics. These factors make the evaluations of the council members suspect, but Hialeah, nonetheless, follows the pattern seen in other cities in this group of placing great import in the role of business and governmental organizations. Cape Coral's members manifest opinions which appear more reasonable. Cape Coral is a recently incorporated municipality growing out of a prolonged effort to develop a planned community near Ft. Meyers. For a small city of 16,000 (although it is growing rapidly), its boundaries take in 105 square miles, an enormous area. The developed portion of the city lies huddled in one corner of a vast plain of inter-woven blacktop streets and as yet unused canals. The rapid changes it is experiencing are only reaffirming its high SES character. But the changes are only beginning to have effect on the political actions of the council. Peopleoriented interests have had neither the time nor the reasons to form. The cities in SES group F may share relatively similar socioeconomic environments, but the council members of the various cities perceive problems differently. Of course, there are some similarities. Every city, with the exception of Cocoa and Bartow, has members who indicate growth-related problems are the most serious problems facing their communities. However, Gainesville council members report their concern for growth is matched by the financial problems they see facing their city. Gainesville, together with Boca Raton, shows a distinctive concern for human resource problems. In Cocoa Beach, the members rate financial and

PAGE 220

201 governmental problems at about the same level as growth problems. Service provision problems challenge growth problems as the top priority in Boca Raton. These observations are depicted graphically in Figure 5-8 . The members of the city council in Bartow provide the most anomalous picture of the problems confronting their city. They rate governmental problems as most important, followed by service provision problems. Growth problems rank third. In a sense, though, this pattern fits very we'll with the circumstances of the Bartow members. Located in the largest county in the state (Polk), Bartow, the county seat, has not experienced the growth of its neighboring communities. Both Lakeland and Winter Haven have expanded much more rapidly than Bartow, the traditional center of the county. Increasingly, county actions are made with respect to the interests in Lakeland. Moreover, like several other cities in Florida Bartow was caught in the energy crisis of 1974. For years it had taken monies from its electricity generation and put them into the city's general fund. In the wake of the oil embargo, the council was faced with difficult decisions about the rate structure of their utility and about the very future of the utility. The problem perceptions of the council members are quite complex and not easily understood. For example, Boca Raton's members ranked human resource problems and service provision problems quite high in the array of problems confronting their communities. However, in Boca Raton, the importance of growthrelated, human resource and service provision problems are inter-related. Boca Raton's attempts to limit its growth are wellknown. 15 But the coalition working for limiting growth in Boca Raton is polyglot; its motivations are diverse. At least one member of the council suggested the limitation on growth was supported by certain segments

PAGE 221

r 202 0) HI W i-l C CO o OC 4-1 U to 01 pa C iH O CO 4-1 U Ol ra C 0> 2 T! "° .5 > ^ ° J= >-CG j:>aic .n i-H 4-1 O O -H c nj iU-HOOrtj-ira'H u to £ O •-H U 01 O 3 o u CO U to 01 m J= 0) c i-H -H 10 O to .H 01 4-/ M «H « H O > C 01 0) o h in ai J-> i-i 0] CO to ,n *h to xj toow3pito>-coi ™^ n a a) hi o & o u U w CO U o «4-l en (U H CO cj 00 c O •H U a CD o !-l a) Ph e (D rH ,Q O u Ph C o Cfi e cfl a) S rH -H CJ § O -a ai N •H CO O 1-4 a w on w \ \ \y ^

PAGE 222

203 16 of the community on racial grounds, excluding minorities. 1 As a consequence, minority groups and other concerned interests were demanding a reformulation of the growth limits. Boca Raton had been taken to court by such interests to percent its use of the city's water service monopoly as a cudgel with which to restrict the development in the area. The controversy over this aspect of service provision had dominated council meetings in the year before the questionnaire was administered. Boca Raton's city council was rent by the controversy.Virtually every vote taken in the year proceeding my questionnaire tallied at three votes against two. The effects of this situation can be seen in Figure 5-9. This polarization of the council's members is reflected in the council's high score on the polarization scale. The breakdown in council relations is confirmed in the low rating given by the members on council intimacy and the degree of council viscidity. Moreover, according to the council members the community was exerting considerable pressure on them. As expressed by the members, the council's autonomy from groups within the community was quite low. The entire community was involved in the controversy, and this involvement was reflected in the council itself. Cape Coral's council is the polar extreme of Boca Raton. The polarization and stratification of the members in Cape Coral is relatively low. While not necessarily close personal friends, the members in Cape Coral indicate their consensus on goals is high. Moreover, the council is able to work in an atmosphere of relative aplomb. Community groups and individuals exert only moderate amounts of pressure on the members. The members perceive themselves to be an autonomous body where public decision making can take place in a "calm atmosphere." "The members are all working for the good of the community, and the people know it. The members are all experienced and everyone trusts they will do good." .17

PAGE 223

204

PAGE 224

205 cd U o u 0) face cd o o o PQ O U u cd 0) PQ C O O 0) H ^cas-i r cicg >cncd^:=)QJ(^ Cfl 03 Q) 0) Cfl O r-j O 3 £ t— I t-l ,Q ^ cd 04-J-HCflr-ltHUO O -H cd -H cd CD q o PnOHOffiHSPiW H u CO u cd 0) PQ a o u rH cd B

PAGE 225

206 Not every council is as harmonious as Cape Coral, although Bartow, Cocoa, and Ft. Walton have similar patterns (see Figure 5-9). Both Melbourne and Rockledge appear more like Boca Raton. However, for several cities, no consistent pattern emerges from this simple examination. Gainesville and Tallahassee both assume moderate values on most of the council characteristic scales. While these university cities evidence similar degrees of council stratification and polarization, they differ on the other scales in no consistent fashion. Summary This chapter has been devoted to drawing the picture of Florida councils and their members as it emerges in the responses of members from 61 different cities. Typically, Florida council members do not appear to differ substantially from their counterparts across the nation. Their personal attributes, their evaluation of community problems, the types of citizen complaints, and their use of time is similar to that of council members throughout the United States. However, Florida council members are very likely to see themselves as quite influential in the community, and to view local government itself as the most potent force in the shaping of community policy. By and large, Florida council members are conservatives, politically and religiously. Perhaps this contributes to their relative neglect of problems concerning human resources and their concern for financial problems. However, there is no clear link, at this point, between political predispositions and the perception of problems. Indeed, growth-related problems dominate the concerns of Florida council members. In any event, most council members believe their council is performing well and serving citizen needs.

PAGE 226

207 The councils on which the members serve have their own political characteristics which can be gauged through the responses of the members. If the demographic model developed in the previous chapter was correct, then it could be expected these characteristics would be systematically related to the socioeconomic milieu of the council themselves. In general, this relationship did not hold. Rather, political characteristics appear to exist independently of the socioeconomic conditions. However, a few variables such as religious conservatism, business group influence, small group influence, council polarization and the existence of human resource problems are related to the socioeconomic environment of the council. While the differences between all the socioeconomic clusters are not statistically significant across most political characteristics, certain patterns do emerge from an examination of the group means of the councils in each of the six SES groupings. The low SES cities are typically the homes of the most conservative councils, while in the functionally large and high SES cities the councils tend to be more liberal. Business groups have the most influence in the low SES, rural cities, and in the high SES cities of group F. The governmental agencies' influence shows a similar pattern. But people-oriented groups are more likely to be influential in the functionally large or the high SES cities. Councils in these cities also are more likely than most other SES groups to recognize human resource problems. However, it is growth-related problems which dominate the concerns of Florida city councils. In order to investigate the effect of political characteristics upon public policy, the raging effects of socioeconomic forces need to be controlled in some fashion. The SES groupings provide one method of providing for a comprehensive look at political variables. The cities of

PAGE 227

! 208 group F, the high SES cities, show different patterns on each of the political variables. Cocoa Beach is the most conservative council in this group; Gainesville and Boca Raton are the most liberal councils. In the most liberal councils, human resource problems are extended considerable import; in Cocoa Beach, they are relatively unimportant. Cape Coral has a council in which disagreement and polarization among the members is low; in Boca Raton, they are habitual. Throughout this chapter, patterns among variables were treated in anecdotal fashion. Numerous relationships may have suggested themselves to the reader. In the next chapter, I will trace several of these connections as they bear on the development and persistence of council characteristics. Those links help set the stage for an examination of the way in which councils frame their decisions and determine policy within local government.

PAGE 228

209 Notes 1 National League of Cities, "America's Mayors and Councilmen: Their problems and Frustrations," Political Power and the Urban Crisis , ed . Alan Shank (3rd ed. ; Boston: Holbrook, 1973), pp. 153-170. 2 Peter H. Rossi, Richard A. Berk and Bettye K. Edison, The Roots of Urban Discontent: Public Policy, Municipal Institutions, and the Ghetto (New York: Wiley, 1974), p. 34. 3 Robert Putnam, The Beliefs of Politicians: Ideology, Conflict and Democracy in Britain and Italy (New Haven: Yale University, 1973), p. 5. 4 A discussion of the items used to generate this scale can be found in Glen D. Wilson and John Patterson, "A New Measure of Conservatism," British Journal of Social and Clinical Psychology , Vol. VII (1968), pp. 264269. 5 Douglas Johnson, Paul Picard and Bernard Quinn, Churches and Church Membership in the United States, 1971: An Enumeration by Region, State and County (Washington, D.C.: Glenmary Research Center, 1974), pp. 37-42. 6 See, for example, Robert L. Morlan, "Life on the City Council: Realities of Legislative Politics," Capitol, Courthouse, and City Hall: Readings in American State and Local Politics and Government , ed. Robert L. Morlan (3rd ed.; Boston: Houghton-Mifflin, 1972), pp. 213-215. 7 Heinz Eulau and Kenneth Prewitt, Labyrinths of Democracy: Adaptations, Linkages, Representation, and Policies in Urban Politics (Indianapolis: Bobbs-Merrill, 1973), p. 39. 8 Ibid ., p. 53. 9 Ibid . , p. 54. 10 Norman Nie, et al. , SPSS: Statistical Package for the Social Sciences (2nd ed.; New York: McGraw-Hill, 1975), pp. 416-417. 11 For a discussion using this view, see Gabriel A. Almond and G. Bingham Powell, Jr., Comparative Politics: A Developmental Approach (Boston: Little, Brown, 1966) and Richard Appelbaum, Theories of Social Change (Chicago: Markham, 1970). 12 However, it is not particularly useful merely to compare the scores of a single set of councils on a given measure against those of the same set of councils on another measure. In particular, I will not assume that there is some sort of factor at work in a given cluster of cities which makes their response rates to similar questions dissimilar other than the fact that there may be real differences in the perceptions of the members of the clusters. In

PAGE 229

210 ur vey research, it is assumed that similar questions will evoke similar ' gponses where the underlying forces are the same. Assuming otherwise is a Insularly desultory procedure which, if pursued consistently throughout the na lysis, would contravene the epistemological foundations of survey research. t will argue that responses can be compared, not only between different ca les for the same cities, but also across SES groups and scales. 13 Bureau of Economic and Business Research, University of Florida, Flo rida Estimates of Population, July 1, 1972: State, Counties and Municipal ities (Gainesville, Fla. : Bureau of Economic and Business Research, 1972). 14 Summing the scores of each council within each SES group to produce group scores is a useful technique. It can help point to meaningful differences between the various SES groups. However, this technique may overlook the significant internal variance of a group's council scores (the analysis of variance technique overcomes this objection), and increases the possibility that single member's opinion will affect the representation of an entire SES grouping. An investigation of SES group F will remedy these shortcomings since over eighty percent of the members of the councils in this cluster responded to the survey. It will provide a refined look at a comprehensive set of responses in a SES grouping whose expenditure profiles have already been established (see Chapter Three) . 15. See, for example, Ira Sharkansky, The United States: A Study of a Developing Country (New York: McKay, 1975), p. 44. 16 Interview with councilmember, number 130221, December 10, 1974. 17 Interview with councilmember, number 0260613, January 17, 1975.

PAGE 230

CHAPTER SIX PERSONAL ATTRIBUTES, CONTEXTUAL FACTORS AND COUNCIL RELATIONS In the Labyrinths of Democracy , Eulau and Prewitt are concerned primarily with the properties of city councils. Such properties, they argue, help shape the form of municipal policy. But because Eulau and Prewitt begin their analysis at the level of the council itself, they neglect the preferences, aspirations, and beliefs of council members and the impact of such factors on public policy. Exchange analysis counsels us to begin the study of choice with the individual; to focus on his personal attributes. The triple bonding of which Eulau and Prewitt speak, "of councilman to councilman, councilman with council, and council with citizenry," has its basis in the members themselves. 1 The bonds forged by the internal and external constraints on member behavior are conditions by each member's personal predispositions, life experiences, and personal goals. Each member's understanding of his social setting will affect the costs which mark any choice situation. A member's personal attributes and perception of various contextual factors should bear upon the relations which develop among the council members and each of these factors should help determine the policies a council pursues. In this chapter I will consider the manner in which aspects of each member's cognitive structure interrelate. Initially, I will treat the manner in which the member's personal attributes are systematically linked to one another. Then I will examine the pattern of association between 211

PAGE 231

212 personal attribute variables and the members' perception of various contextual factors. Finally, as depicted in Figure 6-1 , I will treat the fashion in which personal attributes and contextual factors are related to various measures of council relations. This analysis will be undertaken at three levels. First I will treat the responses of all the council members in the sample. The responses of members will be taken as subjective patterns of cognition. Second, I will look at the attributes of the councils. The aggregated responses of the members will be considered as objective reflections of the conditions and beliefs which dominate the council as a whole. The patterns emerging from this investigation will be contextual; their divergence from the structural features of the individual members' patterns will be taken as indicative of the effects of council membership. Finally, I will briefly treat the councils of the 12 cities in SES group F, the high socioeconomic status cities. These councils will provide a detailed picture of the relations among the members as they become manifest within a relatively homogeneous set of socioeconomic factors. They will also allow us to consider the manner in which differences among the members condition the relations among the members. Members and the Form of Their Cognitive Structures The public policy which emerges from city councils is likely to be shaped by many different forces. But analytically, at least, it is useful to examine this relationship from another direction. That is, the policy under consideration can be said to galvanize forces differentially to produce political situations. In this view, the politics attendent to the debate and promulgation of a policy are contingent on the nature of o the policy itself.

PAGE 232

[PERSONAL ATTRIBUTES; 213 [CONTEXTUAL FACTORS]

PAGE 233

214 This view has the great advantage of pushing policy to center stage. However, this perspective has a highly reactive thrust. It is evident no one policy is ever considered in vacuo . The expectations of the actors, the means by which decisions are reached in a group setting, and the possibility of conflict itself are conditioned by the evolving circumstance of the council and each of its members. In short, while aspects of the policies themselves may trigger reactions within the council, these reactions are founded in the attitudes of the members and their developed sense of what policy-related actions mean within the council. This is to say, the various factors constituting council life and the promulgation of policy are interdetermined in some fashion. However, the nature of this interdeterminancy can be conceived in two ways. It is possible to view each member as conditioned by his context. This view contends that the connection between beliefs and attitudes is lost to the welter of circumstance and is beyond the researcher's pale. Alternatively, it is possible that council members enter council service with well-developed views of the world and attitudes toward other people and that these attitudes fundamentally color their view of council life. These beliefs and attitudes can be considered to form a member's cognitive structure. The subjective component of a member's responses may be highly resilient to the features of the context. But the constellation of factors interrelated in each member's cognitive structure may not be unique. Empirically, definite patterns may exist because of the manner in which strategic interaction helps narrow the choices available to members and helps shape their choosing mentality. City councils are composed of members; each member brings to the council a set of personal attributes. Fundamental to these are his life

PAGE 234

215 experiences. These experiences may condition his predispositions or personal goals. They might also affect the manner in which he relates to his colleagues or how he views groups within the community. But life experiences are fundamental. While differences in the life experiences of members can be expected to generate differences which may surface in the relations among the council members, life experiences have their separate validity as factors shaping a member's behavior. Are a member's political predispositions related to his life experiences? A substantial number of political observers think the answer is yes. Time and again, an individual's political beliefs have been linked to personal characteristics such as age and education. Typically, older individuals (for whatever reason) evidence more conservative political beliefs than younger ones. Similarly, less educated individuals have been shown to evince more conservative beliefs than more educated ones. Moreover, an individual's secondary group associations, his status, or his occupation have been shown to have importance for personal predispo3 sitions in numerous studies. In order to discover if these same patterns obtain among the Florida council members, I used each member's age, education, occupation, and number of years of council service as indicators of his life experiences. Kendall's tau-c, and where appropriate, Cramer's V were then calculated to establish the degree of association between life experiences and the 4 various indicators of a member's political predispositxons. As seen in Table 6-1 , most of the relationships between measures of life experiences and the political predisposition measures or conservative-liberal scales are statistically significant. Not surprisingly, a member's age (AGE) is associated with his score on the political conser-

PAGE 235

216 vatism scale (CONSPOL) . However, this relationship is most dramatically illustrated in the association between age (AGE) and the summary measure of conservatism (CONSTOT) . The older the member, the more conservative his predispositions tend to be. The association between political conservatism (CONSPOL) and education (EDUC) displays a direction similar to those found in earlier studies as well. The greater a member's educational achievements, the less conservative he is likely to be on all measures of predisposition. Moreover, this relationship is fairly strong across most measures. Only the link between religious conservatism and education is marginal. Notice though that a member's age does not correlate positively with the measure of social (CONSSOC) conservatism. This reflects the generally liberal attitude of all members on social matters. Table 6-1 Kendall's Tau-c Between Conservatism Scales and Member Life Experiences CONSPOL CONSREL CONSSOC CONSTOT EDUC -.473 -.143 .239 .462 AGE .180 .252 .154 .239 YREXP .199 .366 -.088 .276 OCCUP a .276 .179 .186 .291 a. OCCUP associations measured with Cramer's V; OCCUP proceeds as follows: professional, managerexec, real estate/sales; crafts; retired. (Italicized relationship significant at .05 level) The longer a member has served in a council (YREXP) , the more likely he is to be conservative. This conservative bent is not merely a political conservatism, but rather exists with respect to the religious and summary (total) measures. On social matters however, experience on

PAGE 236

217 the council appears to have little effect. What relationship is found between these two variables is negative. Of course, it should be remembered that the variation of member scores on the social conservatism scale (CONSSOC) is small. This decreases the usefulness of this measure since the measures of association are sensitive to column-marginal disparities. For this reason, much of the analysis to follow will omit consideration of the social conservatism scale (CONSSOC). Table 6-2 demonstrates there is a relationship between a member's occupation and his predispositions. On the summary measure of conservatism, each occupation category is revealed to be predominantly conservative. The notable exception is the professionals. Remembering that the summary measure consists, in large part, of the attitudes of members on the religious and political conservatism scales, this moderation by the professionals is entirely understandable. A significant portion of the professionals are liberals politically, while almost half evidence conservative religious sentiments. The associations between council member life experiences and political predispositions generally conform to those noted by other researchers. This finding is heartening in two respects. First, it lends some credibility to the measures being used and the nature of the sample under consideration. This is because the link between life experiences and predispositions as operationalized here is among the most persistent in social science. Had the relationships between the measures of life experiences and predispositions not been consistent with those demonstrated by other researchers, the worth of this research enterprise would be immediately suspect. Second, and perhaps more importantly, it allows us to draw upon the insights of other researchers who have examined similar re-

PAGE 237

218 Table 6-2 The Relationship Between Measures of Conservatism and Occupation (N = 167) Summary Measure of Conservatism (CONSTOT) 3 Occupation

PAGE 238

219 lationships. Verba and Nie, for example, have demonstrated in Participation in America that political beliefs can modify the normal workings of the standard socioeconomic model of participation. It is axiomatic among social scientists that high socioeconomic status is associated with participation in political activities. However, Verba and Nie found conservative beliefs contribute to participation of Republicans in a greater manner than one would expect on the basis of their socioeconomic status alone. The authors suggest that among Republicans "political beliefs play a much more independent role in recruiting citizens to actiQ vity than, is the case among Democrats.' Among council members, predispositions may intervene in a crucial fashion between life experiences and participation in council activities. This is the lesson to be drawn from the work of Verba and Nie. However, it seems clear from the exchange theoretical notions developed in Chapter two that each member's actions are likely to be conditioned by the system of empirical beliefs, expressive symbols, and values which help define the situation in which choice must take place. It also involves each member's personal goals. Richard Fenno, it was noted earlier, has suggested personal goals can condition behavior within a legislative body." In particular, I consider the effects of three types of goals, namely, whether a member has ambitions for public office beyond the (IRUN) council, whether a member derives personal benefit in a personal occupation from his council position (HELPED), and finally, whether a member's ambitions center on his future role in the council itself (FUTURE). While a member's goals may be critical in determining his perception of council relations or in forming his orientation toward community groups, they might also be linked to his life experiences. An examina-

PAGE 239

220 tion of Table ^-3 demonstrates this is true, at least to a limited extent. Older members have less concern for continued service on the council. More experienced members, those with the most years on the council, indicate they have been helped in their private occupations by their tenure on the council. Statistically, the other interrelationships between life experiences and member goals are not significant. Nonetheless, it is instructive to note the negative relationship between educational achievement and responses indicating a council member had been helped in his private occupation through his service on the council. Table 6-3 Kendall's Tau-c Between Measures of Personal Goals and Life Experiences

PAGE 240

221 willingness or desire to run for state, county or national office, and almost half the retired members indicate imminent retirement from the council. Perhaps because they are retired from their private occupations, most retired members suggest their tenure on the council has had no effect on their private activities. Contrariwise, those members employed in manager-executive positions indicate being on the council bas helped them in their private occupations. Slightly over 60 percent of these individuals report they have either been helped or helped a great deal by their council service. It is professionals whose responses indicate they have been adversely affected by membership on the council. Only 36 percent write they have been helped or helped a great deal by their tenure on the council. However, both craftsmen and sales people report council activities have hindered their private occupations in some way. In each case, about 50 percent of those responding report such circumstances. Although professionals report their private occupations have been hindered by being on the council, they nonetheless indicate in great numbers (51%) their willingness to continue in the council for two or more terms. Indeed, each occupational group has a majority of its members suggesting their willingness for continued service. The only exception to this proposition comes among the retired members who report in over 60 percent of the cases that future service is unlikely. As a rule, it appears a member's life experiences are related to his personal goals. However, in no single case, is the association extremely strong. Clearly, other factors may enter into a member's determination of and commitment to a set of personal goals. One likely factor which is suggested in Verba and Nie's work is the political predispositions held by a member. Some social psychologists have suggested some «««

PAGE 241

222 individuals are more risk accepting than others, and that this attribute can be linked to a broad number of personal beliefs.10 In this sense it can be supposed political dispositions might alter goals, making the risktaking of changing office more acceptable or re-enforcim; predilections to remain on the council. Yet this argument can be turned around as well. Conventional wisdom and the example of national politicians like Fulbright and Johnson on matters of race suggest that political predispositions can be altered or set aside for ambition and aspirations. In either case, an associational analysis can determine the manner in which predispositions and goals are linked in the responses of Florida council members. Table 6-4 displays the association coefficients between political predispositions and member goals. Each measure of conservatism associates negatively with intentions to serve additional terms on the council. However, only religious conservatism and the summary measure associate at levels which are statistically significant. The more liberal a member is on these measures the greater the chance he is willing to continue his service on the local council. Conservatives, on the other hand, indicate a consistent willingness to either run for higher office or retire (for the purposes of further analysis, the IRUN measure was dichotomized; the greater the score on IRUN the more other-council oriented the member) . Notably, political conservatism was not associated significantly with the future orientation of the members. Mere political conservatism does not appear to bear significantly on a member's aspirations for office. However, political conservatism is linked with perceptions that holding office had helped a council member in his private occupation. Conservatives are more likely than liberals to report they have been helped by

PAGE 242

223 their tenure in office. Social conservatism bears no significant relationship to any of the measures of member goals. Table 6-4 Kendall's Tau-c Between Measures of Conservatism and Members' Goals

PAGE 243

224 the importance of various groups in local decision making, information sources, and community problems bear a consistent relation to the personal attributes of the members themselves? For example, are liberal members more likely to perceive human resource issues as serious in their community than conservatives? The received knowledge about the elements of an individual's cognitive structure suggest they would. The pattern emerging from an examination of the association coefficients between the members' personal attributes and their perceptions of contextual factors describe a picture of council members with a certain coherence in their cognitive structures. Liberal members are young, educated, attuned to human resource problems, and willing to accede importance to people-oriented groups in the decision-making process. But liberals are in the minority. Politically conservative members are older, more experienced, less educated, attuned to business groups, and not intimately concerned with any special type of community problem. Without dwelling upon the theoretical rationale behind the consistency of members' cognitive structures, their presence can be given some empirical validity. The association coefficients (Kendall's Tau-c) can be pieced together into a network of bivariate relationships which summarize in a convenient fashion the recurring relationships found among the variables. Although these bivariate relationships include colinear influences, they capture the sense of the relationships between the contextual factors and personal attributes without reproducing the detail of the many associations. Figures 6-2a through 6-2e depict the relationships as they appear in a network constructed to include re-enforcing patterns. As a convention, each of the figures is presented in terms of the positive associations between salient features of a member's cognitive

PAGE 244

(a) [PEOPLE]' [citizenq; [HUMP ROB] l(-)CONSPOL]' 225 ;[EDUC] (b) [PEOPLE] [GROUPQ] [CONSREL] [GOVTQ] '[AGE]^y [YREXP] (c) (d) [GROUPQ]^ '[BUSINESS] [CONSPOL] [ GOVTQ ] [EDUC].[GROWPROB]. [AGE] -[(-)CONSPOL] (e) [EDUC] ! •[SERPROB]> [(-)CONSTOT] [GOVTINF] Figure 6-2. Closed Networks of Bivariate Relationships Centering on Measures of Problem Perceptions and Political Predispositions for All Members

PAGE 245

226 structure. They portray only the re-enforced aspects of the structures and should not be mistaken for path diagrams. Unlike balance models or affective-cognitive structures found in the social psychological literature, dissonant elements or associations are not shown. The emphasis is on those measures which are most intimately linked together. In this sense, it varies dramatically from an analysis attempting to tap the latent structure among all aspects of a member's cognitive structure. The networks, I repeat, serve only as a convenient visual guide to the nature of the bivariate associations between the measures. One of the most notable aspects of an inspection of the patterns of associations is the lack of any re-enforcing components related to the perception of financial problems in the community. This is at odds with the received wisdom which attributes considerable importance to financial matters in the conservative constellation of beliefs. Perceptions of the importance of business groups are linked to the view financial problems are important and serious, as is the perceived importance of governmental information sources. However, these measures are not linked to one another. Their only association is had through the mediary of political conservatism. But being political conservative is no assurance of recognizing financial problems. While financial problems are not linked into complementary structures, there is a well-formed network involving the perception of human problems. As seen in Figure 6-2a, such a perception is associated with being a liberal, being well-educated, taking cues from citizens and extending importance to people-oriented groups in community decision making. Each of these factors is intimately and positively related to one another. The network of associations involving the perception of governmental

PAGE 246

227 problems is more complicated and less well formed (see Figure 6-2b). Notice that each element does not re-enforce each of the other elements. Also the number of nodes acting within a distinct sub-set of the network has increased. This structure incorporates the religious conservatism variables whose particular manner of association with contextual factors desires special mention. When the relation between members' predispositions and perceptions of influential groups is investigated, members who are the most politically conservative are seen to have substantial orientations toward business groups, while liberals are significantly oriented toward people-oriented groups. But this pattern is reversed for religious conservatism. There is a slight but still significant relationship between religious conservatism scores and responses indicating the importance of peopleoriented groups (Tau-c = . 124). The relationship between religious conservatism and business groups is negative though not statistically significant (Tau-c = -.107). Members with high scores on the religious conservatism scale are likely to perceive typically uninf luential groups such as the Bar Association, labor unions and motel/hotel associations as important within their community. Moreover, those members most imbued with religious conservatism view governmental agencies as least important in community descision making. This is particularly unusual since the dominance of such governmental groups typifies the perceptions of council members (see Chapter five) . This pattern of association raises a question. What is the nature of this religious conservatism? It is related to political conservatism (Tau-c = .370), but is different enough to alter the relationship between the perception of important groups and predispositions. The scale itself in in I | IIIIIIM1H— iiiiiii III I I KiHIHH

PAGE 247

223 consists of responses to controversial issues such as the death penalty, school prayer, inborn conscience, military discipline, right-to-work laws and the need for chaperones. These items have no precise interpretation. But perhaps the feeling of the measure is evoked in a quote of George Hoar resurrected by Emment Hughes in drawing one distinction between the Republican and Democratic parties: The men who do the work of piety and charity in our churches... the men who own and till their own farms the men who went to war and stayed all through the' men who paid the debt and kept the currency sound and saved the nation's honor. . . by the natural law of their being find their place in the Republican party. ~While~ the old slave-owner and slave-driver. .. the criminal class of the great cities, the men who cannot read or write... by the natural law of their being find their congenial place in the Democratic party. 11 The religiously conservative Florida council member feels a similar pulse. In sum, this conservatism can be sketched as a "God-fearing and self-exalting kind of Calvinism, a reverence for virtues supposedly rural, and a respect for 'sound' fiscal policy ardent enough to equate solvency and sovereignty. "12 The religious conservative shares much with the Republican Will Rogers described in 1928: "He wants politics to be known as a sideline. He is sorter ashamed of it." Underlying this view of society and the state is the notion government is meant to serve the people, but that the people should also help themselves. It is a more traditional type of conservatism, one that cherishes the small farmer and the small business. It is differentiated from contemporary political conservatism in its greater mistrust of government or, at least, the political side of government. Politics in government Is an aberration which should and can be avoided. Government is a matter of administration. There is a strong Calvinist sense of the orderly dispositions of matters. In .any ways, this is the traditional form of

PAGE 248

229 Southern conservatism, a conservatism that sweeps across the broad Baptist-dominated, Bible-belt of the South. Religious conservatism plays an interesting role in the network involving governmental problems for it is the node of triads which might otherwise seem unlikely correlates. Given the known link between attitudes classed as liberal and the attribution of community decision-making influence to people-oriented groups, one triad of this network appears unstable. Namely, the perception of people-oriented groups PEOPLE) as influential is linked with the cited use of group information cues (GROUPQ) and the recognition of serious governmental problems (GOVPROB) in the community. This triad neglects the typical association found between people-oriented attitudes (PEOPLE) and years of council experience (YREXP) ; it also overlooks the general relationship between religious conservatism (CONSREL) and political liberalism (CONSPOL) . Political liberals are seldom the experienced council members. Hence the link between people-oriented members (PEOPLE) and experienced members (YREXP) does not appear. Instead, religious conservatism provides the confounding link. This discussion is meant to illustrate a common problem with the networks. They do not necessarily form as stable regularities as their portrayal would suggest. They are visual aids. And in keeping with the simple nature of this analysis, such potential instabilities must be kept in mind when considering their representation of cognitive structure. An assymetric network of associations is depicted in Figure 6-2c centering on the role of conservative political attitudes (CONSPOL) . Here two distinct triads exist. The first includes the perception by political conservatives of the importance of business groups (BUSINESS) and their reliance on group information sources (GROUPQ). The second aspect

PAGE 249

230 focuses on the manner in which older members tend to rely on governmental information sources (GOVTQ) . This is a triad in which older members and those relying on governmental bodies for their information do not perceive business groups as important in the community. When viewed from this perspective, there appear to be at least two forms of conservative attitudes among the Florida council members. The first form of political conservatism has an ideological edge; it lists toward the business community and relies on business groups for advice. The second form appears associated with the maturation of the member and his reliance on information sources developed in connection with his role in municipal government. Growth related problems are the most serious problems facing the more than 60 cities represented in this study. However, the perception of such problems is strongly linked only to the amount of educational achievement on the part of the member and his degree of liberalism (on the CONSPOL and C0NST0T scales especially). This relationship obtains despite the relative dominance of conservative attitudes of all forms (except CONSSOC) among the members at large. Liberalism, whether political, religious or summary, is shown to be linked to the perception of service provision problems in the communities. However, as portrayed in Figure 6-2e, such problems are also linked to the general notion that governmental bodies are influential in the decision making of the community. Since the governmental influence measure (GOVTINF) includes the heads of local governmental agencies and municipal employees, this relationship is not too surprising. The measure of service provision problems (SERPROB) deals primarily with public utilities. In 1974, the managers of public utilities were hard pressed by the rapid

PAGE 250

2 31 increase in fuel prices to maintain their rate structure; the cost of utility provision was a serious problem to such agency heads. Thus it appears reasonable to suppose a connection between perceptions of service provision problems and the influence of governmental agencies handling such problems. Cognitive Structures and Council Relations Regardless of a member's personal attributes, regardless of his perception of the many contextual factors bearing on the council, he shares with his fellow council members a seat on the council. Per force, he must interact with his fellow council members. In Florida the largest city council has 19 members. Most councils have only five. The members are thrown together and they must somehow get together. The relations that develop among council members need not be amicable. Each member's personal attributes and perceptions of contextual factors may influence how relations among the members appear. In the following pages, I will systematically investigate this proposition. Once again I have assumed a member can be stripped of the context of his council, that his perceptions are totally subjective. In attempting to describe the relations that appear among members, I relied upon Eulau and Prewitt's suggestive illusion of "triple bonding." 13 Each council member, in this view, stands in relation to every other council member, to the council as a whole, and to the citizenry. I constructed seven scales (see Appendix A) in order to tap these aspects of a member's tenure. These measures, together with their mnemonics and definitions, are listed in Table 6-5. Overall, the associations between council relations and aspects of a member's cognitive structure tapped by the measures of personal attri-

PAGE 251

232 butes and contextual factors are many. Only the perception of financial problems (FINPROB) , future (FUTURE) goals, social conservatism (CONSSOC) and belief that the media (MEDIA) is important in community decision making are not linked with several of the measures of council relations. Table 6-6 reveals the manifold complexity of the interrelationships. Table 6-5 Council Relations Scales Used to Tap Each Element of Eulau and Prewitt's Concept of "Triple Bonding" Element in Triple Bonding I. COUNCIL MEMBER TO COUNCIL MEMBER :i. COUNCIL MEMBER TO COUNCIL III. COUNCIL MEMBER TO CITIZENRY Scale (Mnemonic) — Definition 1. Intimacy (INTIMACY)— Familiarity between members of council 2. Viscidity (VISCIDTY) — Degree to which council members work together as a unit 1. Participation (PARTICPT) — Each member's involvement in council activities 2. Polarization (POLARIZE) — Degree of difference in members' general goals 3. Stratification (STRATIF) — Degree of difference among members in terms of their influence and experience 1. Autonomy (AUTONOMY) — Degree to which council is able to pursue its objectives independent of community or individual citizen influences 2. Pressure (PRESSURE)— Degree to which council is under community pressure and this pressure reflected in council behavior

PAGE 252

233

PAGE 253

2 34 It is the politically conservative members who most often respond that participation (PARTCPT) by all members on the council is high. Of course, those scoring highly on the summary measure (CONSTOT) make similar observation. But those registering highest on the religious conservatism scale (CONSREL) make no such observation. Notably this is the only council relation with which religious conservatism is not significantly involved. The perception of extensive council participation is also unusual in the manner in which the influential group and problem perception scales are associated with it. To the degree groups are seen as influential, participation is rated as high; to the degree problems are detected in the community, participation is rated as low. When bivariate associations alone are considered, this scale exhibits peculiar properties. Later, when the network of re-enforcing relationships are examined, these properties will largely disappear. The view that each member is involved in the council's activities does not necessarily coexist with the perception of harmonious and familiar relations among the council members. The association between participation (PARTICPT) and member intimacy (INTIMACY) is quite low (see Table 6-7). Moreover, reports of intimacy among the members appear related to slightly different aspects of the members' cognitive structures. The longer a member has been on the council, the more intimate he perceives relations to be. And while conservative attitudes in politics are associated with perceptions of intimacy, conservative religious views are also linked to such perceptions. Perceptions of intimacy differ from those of participation not only in this respect, but in the types of group influences and the problem perceptions with which they are associted. Business groups and governmental problems are cited as important

PAGE 254

2 35 where intimacy is high. People-oriented groups and growth problems appear to discourage reports of intimacy. The viscidity recorded in the responses of the members is linked to both the measures of participation and intimacy. This is reflected in the association coefficients presented in Table 6-7 . There is then some evidence that a member's ability to work with his fellow members is linked to his view that participation in council activities and the intimacy among the members are high. These relations re-enforce one another in a manner John Hays might have described like this: "He trumped Death's ace for me that day, and I'm not goin' back on him." But beyond the reenforcing aspect of other aspects of council relations, viscidity is recorded as high where the members tend to be older, more conservative individuals who have come to rely on governmental information sources. Viscidity may be encouraged by intimacy and participation but it is not the same social phenomenon. Both stratification (STRATIF) and polarization (POLARIZE) are involved in similar associations with elements of a member's personal attributes and his perception of contextual factors. In both cases, younger members see greater divisions on the council. Liberal beliefs are the correlates of such cleavages, as is a people-orientation. Where members rely on governmental information cues both stratification and polarization are low. The perception of governmental problems acts in like fashion. And reports of service provision problems are contingent to division of the council. There are differences in these two dimensions, however. More experienced members perceive more stratification, while more educated members are likely to say their councils are polarized. Among members who _^di

PAGE 255

236 discern a polarization on the council, group information sources (cues) are not important. Stratification is reported least by those members who rely on citizen information cues. The set of council members which sees their councils as polarized are also likely to suggest their communities are faced with serious growth-related problems. Table 6-7 Spearman's Rho Between Measures of Council Characteristics ^ S & «? $ i Co

PAGE 256

237 autonomy is associated with governmental problems. Those members with the least experience perceive the greatest community pressure; those more experienced indicate higher levels of autonomy. Younger members report community pressure; older members feature autonomy. The role of council relations in the cognitive structures of the Florida council members can be usefully depicted by playing out the reenforcing associations in the manner used earlier. Figures 6-3a through 6-3g portray these associations in graphical networks. There is notable relationship between conservative predispositions and member-to-member interaction. Conservative predispositions are linked with council viscidity, autonomy, participation and intimacy. On the other hand, there is a consistent relation between liberal beliefs and measures which capture various cleavages, such as stratification, polarization and community pressure. Such associations remain even when the link between beliefs and council relations is examined for the effects of life experiences. For example, the relationship between political conservatism and council polarization remains at -.389 when the first order Kendall tau-c is calculated, controlling for the effects of member age. However, the relationship between member age and polarization drops from -.104 to -.035 when the effects of political conservatism are partialled out. Likewise, the association between political conservatism and council polarization, controlling for education, is -.383; partialling the effects of political conservatism out of the link between council polarization and member education, the coefficient becomes -.062. However, the effects of life experiences on the relation between predispositions and council relations are not repeated when other measures are controlled. Problem sensitivity and group orientation remain important despite statistical controls.

PAGE 257

2 38 (a) [BUSINESS [GOVTQ] (b) [ () CONSPOL k [HUMPROB ] [PEOPLE ][STRATIF] [YREXP] [YOUNG] [SERPROB] (c) [HUMPROB] [(-) CONSPOL] [SERPROB] [(-) CONSPOL] [EDUC]' (d) [CONSREL] [YREXP] <[ GOVTQ] [CONSPOL] [BUSINESS] [GROUPQ]' (e) [HUMPROB][ (-) CONSPOL K ^ [PRESSURE] [CITIZENQ] [YOUNG] [GROWPROB]Figure 6-3. Closed Networks of Bivariate Relationships Centering on Measures of Council Relations for All Members

PAGE 258

2 39 .[AGE], (f) [AUTONOMY ][GOVTQ] [GOVPROB] [CONSTOT] (g) [AGE] [CONSTOT] [VISCIDITY] [govtq ; Figure 6-3. Continued

PAGE 259

240 The Council Context So far in this chapter, I have assumed each member's responses can be taken and analyzed independent of the council on which he serves. Indeed, it well may be that predispositions, life exe P riences , and goals so form an individual's cognitive structure as to render it relatively impervious to the collective choice situation in which he finds himself. However, it is also possible that contextual factors may systematically prejudice the relations among the members or between the council and the citizenry. In order to test this proposition, I used the mean scores of the 61 councils whose characteristics were formed from the responses of their respective members. An array of associations was calculated in a fashion identical to that reported with respect to the member's cognitive structure. If the council context fundamentally affects the relations already presented, the nature of the associations should change. This change should be reflected in the individual associations and, if the changes are dramatic, in the general pattern of association. Of course, there are two difficulties with this approach. First, the possibility of idiosyncratic error is significant. In 11 of the councils, one member determines the patterns attributed to the entire council. Perhaps more important, however, is the nature of the test itself. It looks in only one direction. It cannot substantiate an effect that re-enforces the associations already present. Nonetheless, it can detect distinct effects, and those effects can alter the developing picture of councils and members. For convenience, I will briefly note the significant changes in the individual associations rather than present the complete panoply of association coefficients. Then I will examine the network of associations in

PAGE 260

241 a fashion similar to that used in exploring the cognitive structure of the members. When the scores of council members are aggregated according to their membership in a particular council, certain changes occur in the pattern of association coefficients. City councils with older members are no longer viewed as politically conservative (CONSPOL) , but such councils do tend to be religiously conservative (CONSREL) and to score high on the summary measure of conservatism (CONSTOT) . The council context also diminishes the relationship cited earlier between low levels of education and religious conservatism. However, it strengthens the association between business orientation and little education. When councils are the focus of inquiry most of the relationships between predispositions and group orientations disappear. The major exception is the rather strong link between councils with a sense of religious conservatism and their orientation towards the business community. Councils do not appear to be sensitized to community problems by the groups they consider important, although extending importance to peopleoriented groups does tend to be associated with the perception of human resource and financial problems. And where councils put a great stock in the decision-making influence of governmental groups, the association with serious problems across a range of areas is consistently high. In fact, in such councils only governmental problems are not seen as serious. The greater the average age of the council, the less inclined its members are to cite human resource problems as serious. Well educated councils, on the other hand, are quite sensitive to such issues. Strangely enough, so are those councils whose members have the greatest average number of years of service. This is a reversal of the pattern that was found among individual council members.

PAGE 261

242 It might be expected that those councils which indicate a preference for citizen or group information cues would also be attuned to a range of problem areas. The associations do not substantiate this view. Instead, only governmental information cues are connected to any of the problem perception areas, and this is linked with the perception of governmental problems. When a council reports that it relies heavily on governmental information sources, its members tend to be older, religiously conservative and relatively harmonious. The members can be expected to participate in council activities, to be on familiar terms with one another and to share perspectives on the manner in which the council should pursue its purposes. In situations where the council is divided whether by experience, by community pressures, or by the goals of the members, it is unlikely governmental information sources will be important. In fact, where community pressure is great, it is citizen cues which members rely on. As a general rule, a divided council will be more liberal in its predispositions. Conservative councils (with the exception of the social conservatism measure) exhibit substantial amounts of intimacy and viscidity. However, as seen in Table 6-8 , council autonomy is not generally related to predispositions on the council. Instead, only religiously conservative councils evince feelings of great autonomy. Strangely, the autonomy of the council does not appear related in a systematic way to the problems facing the council. It might be expected that councils which have less serious problems to cope with would also perceive themselves more free from community demands. At any rate, this is the interpretation one could muster from the manner in which serious problems are

PAGE 262

243

PAGE 263

244 associated with the perception of community pressure. Serious human resource problems, growth-related problems and service provision problems are indicated in communities whose councils report the greatest community pressures. Indeed such problems are also linked to council stratification and to a polarization of the council. Thus it is somewhat of a surprise that councils do not link perceptions of autonomy with problems in the community. This relation argues for the distinction made between the measure of community pressure (PRESSURE) and council autonomy (AUTONOMY). The context of the council does affect many of the associations between measures of council properties. However, as the above discursive xndicates, its impact is fragmentary and difficult to discern. Most of the change comes not in the composition of the emerging picture of councils but in its detail. For the most part, the description of councils matches that of council members. Positive relationships remain positive. Negative ones endure. The only notable exception to this pattern is the complete change in the nature of the link between the years of experience a member has on the council (YREXP) and the perception of human resource problems (HUMRPROB) . For all members, the Kendall's tau-c between these two variables was negative; at the council level, the relationship becomes positive. In the final analysis, the greatest single impact of placing members in their council context comes in the reduced number of statistically significant relations which obtain between the many measures. This result is not entirely unexpected; it is the probable effect of aggregation since the variance of the measures is reduced and the number of cases is reduced dramatically. The overall impact of the council context can be depicted as a series of bivariate relationships in a network similar to the ones used to por-

PAGE 264

245 tray the members* cognitive structures. These networks are shown if Figures 6-4a through 6-4g . A comparison with the networks established for the cognitive structures shows some remarkable similarities. For example, the triad of youth, perception of human resource problems, and stratification reappears when members are considered in the council context. Indeed, several of the networks share core associations with those seen in the cognitive structures of the individual members. Only in the constellation of factors associated with the perception of active membership (PARTICPT) does context appear to decisively affect the earlier finds. As seen in Figure 6-4a , participation is involved in no single re-enforcing relationship when considered as an emergent phenomenon. This measure is involved in a much different set of relationships when considered in the council context. But other aspects of council relations are bonded to re-enforcing measures which resemble Figures 6-4b through 6-4g. Intimacy, autonomy, and viscidity are once again correlates of conservative predispositions (see Figures 6-4d , 6-4f , and 6-4g ) . In each case, the importance of governmental cues continues to be re-enforcing. Council divisions can once again be connected to the perceptions of various problems by the council members. And although the nature of the problems contingent to such cleavages changes, the perceptions of problems remain an important aspect of division among the members. The greater the problems perceived to be confronting the council, the more likely the council is to be stratified, polarized and exposed to community pressures. In the networks, I have included several re-enforcing relationships which link various problem areas to the forms of council relations which emphasize member cooperation and interaction— intimacy, autonomy, and viscidity. In each case, the associations are negative. Apparently, in the

PAGE 265

246 (a) (b) (c) (d) (e) [CONSREL][particpt; [PEOPLE]*[ GOVTQ] .[YOUNG]. [(-)GOVTQ] [STRATIF]-[HUMPROB] -[SERPROB] ^"[SIZE] . — X [(-)GOVTQJ-[POLARIZE] — [(-)CONSPOL] [YOUNG]' -[YREXP] [G0VPR03] y K [ INTIMACY]^ -[GOVTQ] 1 [(-)CONCLNO] [CONSPOL] [(-) SERPROB] [conspol; [(-) SERPROB] -[VISCIDITY] [ (-) GROWPROB] [AGE] [GOVTQ] Figure 6-4. Closed Networks of Bivariate Relationships Centering on Measures of Council Relations for All Councils (N = 61)

PAGE 266

24 7 (f) [YOUNG] [conspol; [GROW ROB] S [PRESSURE]< [SERPROB]•[(-)GOVTQ] (g) '[CONSREL] [GOVTQ]^[AUTONOMY] [EDUC] [GOVTINF] Figure 6-4. Continued

PAGE 267

248 face of strong member interaction the perception of community problems is discouraged. A close examination of the networks will reveal two factors not previously introduced into the analysis. The first is the size of the city which the council represents. As used in this analysis, size in an ordinal variable; the cities of Florida are placed in five categores (10,000 -19,000; 19,001-38,000; 38,001-74,000; 74,001-142,000; 142,001 and above). 14 It was introduced into the analysis at this point in order to determine the relative effects of environments of different complexity upon the relationships portrayed in the networks. 15 However, city size enters a reenforcing relationship in only one instance. It is the larger cities which have councils which are polarized, and it is the councils of the larger cities which have the greatest tendency to be liberal. The second new variable is the size of the council itself (CNCLNO) . Councils are small groups, but they vary in size. The number of members might be supposed to have an impact upon council relations. Alternatively, the size of the council may interfere in some systematic way with the amount of information available to the members. As much small group research suggests, the number of members forming a deliberative group could have such impacts. An examination of the figures displaying the re-enforcing networks reveals only one instance in which the number of members sitting on a council has an impact on council relations. Namely, the greater the number of council members making up the council, the lower the reported intimacy among the members. However, this is the only reported impact of size on the council. At least among the Florida councils, the importance of group size is minimal. 17

PAGE 268

249 Council Relations in SES Group F The description of Florida city council members and their cognitive structures developed in this chapter has not been changed markedly by an examination of the effects of council membership. The patterns of association displayed in the emergent aggregation of council responses sustain substantially the basic description of member attitudes and beliefs. But this interpretation must be accepted with considerable circumspection. The source of this caution emanates from the responses used to produce the findings. Clearly, the non-random nature of the individual responses should give us some pause. And the manner in which the emergent councils were constructed places a large onus on each member's response, especially in those councils where only one member responded. But the important aspect of these data limitations is the manner in which it helps condition the type of analysis which can be done. In particular, councils formed from the responses of but a single m ember cannot investigate differences among members. Exchange analysis cautions us to discover the rationale for behavior in the costs accrued to members. The costs -ay be generated in many different ways, but at least one specification of costs identifies their appearance with propensities toward conflict. 18 In this sense, the ability to investigate differences among council members becomes quite important. A complete set of responses becomes quite advantageous. Fortunately, among the cities of SES group F most council members are represented. An examination of the pattern of associations manifest in this set of councils will provide a check on the previous findings in a more complete data set and facilitate the investigation of council differences. The fundamental hypothesis of exchange theory links perceived costs

PAGE 269

250 to individual behavior. Thus the relations that characterize a council are the product of perceived costs. As argued in Chapter two, significant differences among members with respect to predispositions, life experiences, and member goals are likely to be associated with increased decision-making costs, and council relations are likely to be less than harmonious. !9 If a council already experiencing significant cleavages is subject to severe external costs, relations will have an even greater tendency to deteriorate. Such costs are difficult to specify as a general proposition. However, Cobb and Elder suggest the greatest costs are generated by those elements of the community with the greatest ability to increase controversy over a particular issue. 20 Broadly based community groups appear the most likely source of such costs. It may well be, however, that council relations have a much stronger stylistic component than can be detected in the measures I have offered. For example, among the council members there is s strong link between conservative predispositions and the tendency of councils to establish harmonious relations. This remains true of the councils themselves. The theoretical basis of this connection might be found in the predilection of conservatives toward orderly proceedings, an emphasis on procedure. But there is little reason to suspect stylistics are linked exclusively with any form of ideological predisposition. Indeed, several scholars have demonstrated this is not the case. 21 A more likely hypothesis would suggest these elements of the member's cognitive structure appear as an artifact of the statistics used to measure the association and the data upon which they are based. In other words, councils with high scores on the conservatism scales may be harmonious not because conservatism is linked to harmony, but rather that a high score on the conservatism

PAGE 270

251 scale is the aggregate of several components. It has already been noted that the typical Florida council member is conservative. Recognizing this, it is clear that significant differences among the members will serve to lower the mean score of the council member, and hence affect the overall score of the council. While this hypothesis squares with the notions developed in Chapter two, the possibility remains that predispositions, life experiences, or member goals can have independent effects upon the relations of a council. Such a possibility is strongly indicated in the analysis of the 61 councils. Consequently, in examining the patterns of associations among the councils of SES group F, I included not only measures of differences among the members-the standard deviation of scores in a given council-but also the aggregate measures used previously. The inclusion of such measures is not without merit. Highly educated councils tend to be more liberal, more sensitive to people-oriented and media groups, and more aware of human resource problems than councils whose members possess a lower degree of educational achievement. Among the councils of the high SES cities of group F, those whose members have the longest tenure are likely to be most conservative on all measures, but tenure also is associated with a sensitivity to people-oriented groups. These well tenured councils report that groups typically viewed as having little influence in community decision making such as bar associations are important. This orientation is consistent with the reliance of such councils on groups cues for information. What of conservative councils? Aside from being more experienced and less educated, how do such councils perceive their environment? For politically conservative councils, business groups, governmental agencies

PAGE 271

252 and groups typically seen as unimportant such as bar associations are rated as most important for community decision making. Among the cities of group F, a council whose members evidence substantial religious conservatism is likely to view only that cluster of community forces which includes labor unions and the bar association as important. And with the notable exception of such councils, all bodies whose members are predominantly conservative report a lack of serious problems facing their community. In religiously conservative councils, governmental problems once again identified as quite serious, despite the reliance of such councils on governmental information sources. The politically conservative council, on the other hand, is much more likely to rely on citizen or group information cues. As among the 61 councils previously investigated, there appears to be no significant set of factors which act to explain the participation of the members in the activities of the councils in the high SES cities. Examine Table &-9_. The familiarity of the members of the intimacy of a council is associated with a low degree of educational achievement, an absence of important people-oriented groups, a general feeling that council membership had assisted members in their private occupations, and a small difference on predispositions. Councils which evidence substantial amounts of viscidity display a quite similar pattern. They tend to be less educated and not overly affected by strong people-oriented groups in the community. However, such councils are characterized by older members, an absence of serious community related problem perceptions, and quite small differences among the members in beliefs related to religious conservatism. As in the patterns of a member's cognitive structure, the serious nature of governmental problems is likely to be accepted in these councils.

PAGE 272

253 p. 3 o u o IH o H U c 3 O cj o

PAGE 273

254 H l-H Q H C_> CO o

PAGE 274

255 What are the characteristics of councils whose members typically feel themselves autonomous of community pressures? In many ways, they resemble those councils which are most harmonious. They are less educated as a group than most councils in SES group F tend to be. In addition, people-oriented groups are not likely to be important, human resource problems and growth-related problems are not viewed with alarm, and the members of such council are quite likely to share similar predispositions over a broad range of topics. However, the councils which register as most autonomous have members who report their intentions to continue running for re-election as quite low. Moreover, the most autonomous councils are likely to have members who share this preference for limited council service. Not surprisingly, I think, councils whose members need not attend to the local electorate because they do not highly value future council service are those councils whose members feel most free of the community. In polarized councils, continued council service is highly valued by the members. It may be this goal which helps account for the perception of serious problems across a range of issues. Only governmental relations are excluded from concern. In any case, polarized councils are likely to have young, liberal members who recognize the importance of people-oriented groups in their community. These members are quite unlikely to share similar political predispositions. Instead, members serving on polarized councils are likely to differ considerably in their fundamental beliefs. In this sense, the most polarized councils are very much like the most stratified councils in SES group F. However, stratified councils are not significantly linked to liberal predispositions (the association coefficient is in the right direction). Moreover, de-

PAGE 275

256 spite the fact that stratified councils have members who are willing to serve on the council in the future, there is substantial and significant variance in the aspirations of members on such criteria. The experience and power differentials captured in the measure of stratification appear linked to differences in goals as well. Stratified councils attend to peo P le-oriented community groups and human resource problems in much the same fashion as the most polarized councils. A council whose members report the presence of significant community pressures (PRESSURE) typically will be quite young. Nonetheless, there is a significant relationship between the average number of years served on the council and the amount of perceived pressure (tau-c = .481). The greater the number of years of service on the councils in SES group F, the more pressure appears manifest. In addition, such councils assign considerable importance to people-oriented groups within the community in the decision-making process. They rely on citizen information cues. But pressured councils need not be liberal. Instead, the members are likely to differ substantially in their political predispositions. The patterns of association which have been sketched above are depicted in Figures 6^5a through ^Sf. Of particular interest is the importance of shared or conflicting dispositions in each of the networks. For each of the networks describing bivariate relationships as they center on council relations, differences in predispositions are involved. Lest if be thought differences in political predispositions dominate other links, an inspection of the first-order Kendall's tau-c would demonstrate otherwise. When differences in dispositions are controlled, other relationships remain. However, when the relationship between differences in political predispositions and measures such as the impor-

PAGE 276

257 (a) [PEOPLE] [YOUNG] [EDUC] [HUMPROB] / [FUTUREA] [CONSRELA] (b) [(-)GOVTQ] [SERPROB] [CONSTOTA]^ [YOUNG] [HUMPROB] [EDUC] [PEOPLE] (c) [(-)FINPROB] [INTIMACY] [(-)CONSTOTA] [(-)PEOPLE] [(-)EDUC] (d) [PEOPLE][YREXP][PRESSURE]; [(-)GROUPQ]•[CONSTOTA] •[(-)GOVPROBJ Figure 6-5. Closed Networks of Bivariate Relationships Centering on Measures of Council Relations for the Councils of High SES Cities (N = 12)

PAGE 277

258 (e) [(-) FUTURE] [(-)FUTUREA] \ [EDUCA] [CONSPOL];^[AUTONOMY]— [ (-)FINRROB] [(-)CONSPOLA]/ \f(-)EDUC] -^[ (-) PEOPLE] (f) [ (-) EDUCL[ (-) CONSRELA] [(-)PEOPLE] [AGE] [(-)HUMPROB] [GOVPROB] ^[GROUPQ] Figure 6-5. Continued

PAGE 278

259 tance of community organizations, information cues, or problem perceptions is controlled for the effects of council relations, the associations fall to near zero. The link between council relations and other characteristics of the council remain important. While the patterns displayed in the networks may involve different sets of relations, Table 6-10 indicates there is an interrelationship between the different types of council relations. Only member participation in council activities cannot be linked to most of the other types. There are two basic forms of council relations at issue, consensual and conflictual. Councils whose members report high levels of intimacy usually indicate substantial levels of council autonomy and viscidity. Likewise, autonomous councils register highly on the intimacy and viscidity scales. Councils with the highest viscidity scores also rank high on council participation, intimacy, and autonomy. This encapsules a generally consensual pattern which contrasts dramatically with the conflictual aspects of council relations. Polarization, stratification, and felt community pressure are inter-related, conflict-centered patterns of member interaction. Each of these two forms of relations is connected with different features of the council and its members. As a rule, conflictual relations are marked by the importance of people-oriented community groups in the decision-making process and significant differences among the members on basic beliefs. Consensual relations, on the other hand, are associated with a greater similarity of political predispositions, lower educational achievement and the absence of significant people-oriented groups in the decision-making processes of the community. That differences among the members' fundamental beliefs or that the importance of people-oriented groups might help establish the difference

PAGE 279

260 ' Table 6-10 Kendall's Tau-c Between Measures of Council Characteristics for High SES Cities of Group F PARTCIPT _. 329 _.056 .028 -.122 .336 .454 STRATIF POLARIZE INTIMACY PRESSURE AUTONOMY 683 -.277 .692 -.769 -.650 -.536 . 761 -.838 -.563 -.515 .525 .529 -.581 -.706 .559 (Italicized Relationships Significant at . 05 Level) Table 6-11 ^t 1 ^;:/ 311 " 6 ':'" 6611 Measures of Predispositions and Diff xn Predxsposxtxons for Councils of High SES Cities of Gro erences Group F Predispositions

PAGE 280

261 between the two broad types of council relations is understandable within the context of exchange analysis. External costs and decision-making costs may shape the character of member interaction. However, in several of the networks of associations, it is liberalism or conservatism, per se, which is important. It is not entirely clear why the content of political predispositions might affect council relations. Earlier in this chapter, I suggested one type of explanation for this phenomenon. But at that point I suggested a more likely possibility might lie in the mere number of council members in the sample who were conservative. We are now at the stage where this assertion can be examined. The association coefficients between the variation in the beliefs of a council's members and the magnitude of a council's score on measures of political predisposition are indicative of the worth of the assertion. When the coefficients are positive and high, this demonstrates a strong link between the depth of the conservatism and the variation of the members. This offers support for the notion that the data itself may be causing the relation between the content of the predisposition and the council relation type. However, this is not the case, as can be seen in Table 6-11 . Only the social conservatism scale shows any relation to the differences in members' predispositions. Thus it appears that it is the content of the predispositions themselves which predispose council members to relate to one another in one fashion as opposed to another. Summary In this chapter, I have conducted a prolonged data search aimed at depicting, in rough fashion, the cognitive structures of Florida city council members. A member's political predispositions were found to be fundamental to his view of the political events swirling around him.

PAGE 281

262 For example, members with politically liberal predispositions were shown to be sensitive to human resource problems in their community, to perceive people-oriented groups as important in community decision making, and to be receptive to citizen information cues. Political conservatives, on the other hand, were shown to view business-oriented groups as important in community decision making, and to be receptive to group and environmental agency information cues. However, liberalism and conservatism take many different forms. And among Florida council members, religious conservatism proved to be a powerful discriminating agent. Religious conservatism is associated with the perception of governmental problems and a reliance on governmental agency information cues. But this is not a formula for political conservatism. Rather, the religiously conservative member is likely to attribute some importance to people-oriented groups in community decision making. When this line of analysis is extended to include a member's perception of council interrelationships, additional patterns emerge. Specifically, liberals are likely to be sensitive to community pressures; conservatives will report autonomy. For liberals, councils are likely to appear stratified, polarized and pressured. Conservatives are more likely to conceive of their situation as succoring participation by all members, intimacy among members, and viscidity toward council goals. It might be supposed that when council members are placed in their contexts the relationships noted above would disappear. Instead, the context of the council has only a fragmentary impact upon the associations. Intimacy, autonomy and viscidity are once again the correlates of conservative predispositions. Liberals are more likely than conservatives to perceive serious problems in the community. The greater the problems

PAGE 282

263 perceived to be confronting the council and the community, the more likely the council is to be stratified, polarized, and sensitive to community pressures. When councils are taken as the unit of analysis, member predispositions do not play as significant a role in the understanding of council characteristics. Rather, community problems appear to foster such relations. However, several other factors such as the age of the council members and the information sources relied upon also condition council relations. Many of these same relationships reappear among the councils of SES group F. Councils whose members report high degrees of intimacy are also likely to indicate high degrees of relative autonomy and viscidity. Where viscidity is high council participation will also be high. These consensual patterns stand in contraposition to the more conflictual relations of council life. The perception of community pressure, council stratification, and polarization form a distinct pattern of interrelated variables. As a rule, conflictual relations are marked by the importance of people-oriented community groups in the decision-making process and significant differences among council members on measures of political predispositions . The profiles I have drawn in this chapter do not immediately bear upon considerations of policy making. However, it is clear within the context of exchange theory that the types of relations obtaining among the council members can affect the structure of costs and rewards attendant to a policy. In the next chapter, I will make the connection between the concepts developed in this chapter and various measures of council conflict. In particular, I will draw the links between the ways in which policy is made and the council conditions which seem to favor such methods.

PAGE 283

264 Notes 1 Heinz Eulau and Kenneth Prewitt, Labyrinths of Democracy: Adaptations, Linka ges, Representation, and Policies in U rban Politics (IndianapolisBobbs-Merrill, 1973), p. 81. 2 See Theodore Lowi, "American Business, Public Policy. Case Studies, and Political Theory," World Politics , Vol. XVI (July, 1964), pp. 677-715 for an early pronouncement of this view. 3 For one statement of this view, see Sidney Verba and Norman Nie, Participation in Ame r ica: Political Democracy and S ocial Equality (New York: Harper and Row, 1972). ' ~^ „ x _ . 4 , T ° r a su ccessful discussion of these statistics, see William Buchanan, Nominal and Ordinal Bivariate Statistics: The Practitioner's View," American Journal o f Political Science , Vol. XVIII (August, 1974), pp. 625-646. _ 5 Jere Bruner, "What's the Question to that Answer? Measures and Marginals m Crosstabulation, " American Journal of Pol itical Science, Vol. XX (November, 1976), pp. 781-804~ ~~ 6 See, for example, Lester Milbrath, Political Particip ation (Chicago: Rand McNally, 1965). ~ — 7 Verba and Nie, op. cit ., p. 298. 8 Ibid. 9 Richard Fenno, Congressmen in Committees (Boston: Little, Brown, _. . For a discussion and bibliography relating to personality and conflict, see Kenneth W. Terhune, "The Effects of Personality in Cooperation and J™" 1 Ct ' Thc st ^cture of Conflict , ed. Paul Swingle (New York: Academic, 1970), pp. 193-234. TTV , 1:L John Emmet Hughes, "The Political Summer," Saturday Review, Vol. LIX (June 26, 1976), p. 12. 12 Ibid . 13 Eulau and Prewitt, op. cit ., p. 81. 14 The 78 cities used in the analysis were placed in these categories by a log-linear transformation. 15 This effect figured prominently in Eulau and Prewitt 's analysis of city councils. They use size as a variable and a context. Eulau and Prewitt op. cit. , p. 73. '

PAGE 284

265 16 Daniel Katz and Robert Kahn, The Socia l Psychology of Organizations (New York: Wiley, 1966). ~~ ~" ' 17 Eulau and Prewitt, op. cit ., pp. 185-188. 18 Ian Budge, "Consensus Hypotheses and Conflict of Interest; An Attempt at Theory Integration," British Journal of Political Science, Vol. Ill (January, 1973), pp. 73-98. " 19 Budge, op. cit ., p. 91. 20 Roger Cobb and Charles Elder, Participation in American Politics; The Dynamics of Agenda-Building (Boston: Allyn and Bacon, 1972), pp. 11221 Robert Putnam, The Beliefs of Politicians; Ideology, Conflict and Democracy m Britain and Italy (New H.vp, '; v.l. UnT^sTt ^ 1"7?), pp . 3', ',5 .

PAGE 285

CHAPTER SEVEN COUNCIL CONFLICT AND DECISION-MAKING MODES Social observers as diverse as Karl Marx and E.E. Schattschnieder have seen in conflict an analytical key to the understanding of political relations. "The degree to which individual goals are incompatible and the consequence of such incompatibility constitute a staple problem of political science."1 Conflict can take many forms, however. It ranges from personal quarrels to interclass conflict. It can also represent different states of incompatibility. For example, some scholars would distinguish between conflict and competition in terms of the degree to which incompatibility permeates all aspects of a dispute. 2 Conflicts can differ in their modes of resolution as well. Some conflicts are intended to be resolved in the electoral arena, others in committees. In truth, governmental institutions are often designed to handle conflict in a certain manner and to insure that specific forms of conflict are resolved in very specific arenas. City councils, in theory, are intended to amalgamate the demands of the citizenry, while charging the community with commitments. The design of many local governments suggests councils are to serve as forums in which community conflicts may be mirrored. However, conflict may be introduced into council proceedings in several ways. Beyond the relationship of council and citizenry, the differences which become manifest among members over matters of policy may issue from the relations of council member to council member. 266

PAGE 286

267 In this chapter, I will investigate several aspects of conflict as it appears on Florida city councils. My inquiry will focus first on the way in which the expectations of members about the nature of council service impacts upon the appearance of council conflict. Second. I will trace the correlates of council conflict as they are portrayed in Figure 7-1 . In particular, I will examine the way in which personal attributes, contextual factors and councils relations are associated with such conflict. Finally, I will extend my discussion of council conflict to include the manner in which councils organize themselves for policy decisions. Here I will focus on the differential use of certain decisionmaking tacts, which I call decision-making modes, to handle the policy issues appearing before each council. This treatment will examine the way in which each mode appears related to the amount of conflict characterizing the council, the attributes of the members, the perceived contextual factors bearing upon the members and the relations among the members . Council Membership and the Prospect of Conflict What does it mean to be elected to a local council? How is the new member to understand his job? How is he to best represent his constituency? These are questions facing every council member. The answers are not straightforward; they turn on differences in each member's orientation toward democratic processes and toward the role of the local council.-' It is just such differences which lie at the base of the distinctions usually drawn between the "free mandate" and the "imperative mandate" forms of representation. As argued by Burke in his well-known speech to the Electors of Bristol, representatives are trustees of the public confidence who are extended a free mandate to act independently of their

PAGE 287

268 [DECISION ORGANIZATION] Figure 7-1. Major Categories and Variables Treated in Chapter Seven

PAGE 288

269 constituents' immediate wishes. This independence is sanctioned by and conditional on the electoral conflict which proceeds the assumption of office. The direct resolution of the conflicting interests among a constituency is manifest in electoral competition. The constituent's right to reject and choose his representative at regular intervals and through consistent procedures insures the articulation of interests. Once in office, the trustee has considerable latitude to pursue his own interests; reelection buffers his independence. For individuals given an "imperative mandate," responsibilities to the electorate are conceived differently. Theory would have them act as delegates of majority interests, responsive to the constituency which elected them. The electoral arena is but the first of many political situations where the interests of this majority clash with other opinions. This conception of representation does not isolate the interests of the masses in the electoral arena. These two forms of representation are different. They are different not only in their conception of the relationship between electors and representatives, but also in their expectations of how an officeholder is to perform, and how policy is made. Trustees are given free reign in office not merely because they are winners of a constituent mandate. They are also presumed to be highly competent, knowledgeable men capable of sitting with similarly endowed individuals and making policy in the best interests of the entire community. Delegates are also winners in the electoral -arena, but it is their constancy to constituent interests in the legislative arena which distinguishes them. Their competence is defined by their repsonsiveness. If representatives with different orientations do indeed exist, it might be expected these differences would have a significant impact upon

PAGE 289

2 70 the policies they initiate and support. For years political scientists studying national and state legislatures have attempted to discover the orientations of legislators, and to exploit these findings in explanations of legislative actions. 5 Recently scholars have turned their energies toward city councils and council members hoping to extend the domain of such explanations. 6 Eulau and Prewitt in Labyrinths of Democracy present an intriguing argument using such concepts. However, their argument is seriously circumscribed. In making the transition from their research on national and state legislatures, Eulau and Prewitt neglect the historically different role assigned city councils and the different expectations of member behavior. 7 In concentrating on representation, they slight the consequences of representational expectations because they bring to their •analysis a definite conception of what a legislature should be. It is a conception at odds with many of the theories of municipal governance. There are several different types of legislatures. Nelson Polsby has drawn the distinction between legislative forms on the basis of the activities performed by the body and its openness to public influence. His typology is presented in Table 7-1. Table 7-1 Typology of Legislative Forms When Political System Is: CLOSED OPEN 1. No Legislature: Junta 3. No Specialized LeUNSPECIALIZED or Clique Makes Laws gislature: Town £ Meeting Makes Laws Bovernment Activity Is: SPECIALIZED 2. Corporate Board of Di4. Parliamentary Arerectors; Rubber Stamp nas ; Transformative Legislatures Legislatures Source: Nelson Polsby, "Legislatures," in Handbook of Political Science Vol.5, David Greenstein and Nelson Polsby (Eds.), p. 263.

PAGE 290

271 Both the U.S. Congress and the British Parliament, according to Polsby, are relatively open and specialized. 8 However, Polsby suggests the U.S. Congress is significantly different from Parliament. Polsby calls the U.S. Congress a transformative legislature. Its legislative outcomes are determined by the internal structure, the subcultural norms, the perceptions and regulation of interests, and the predispositions of the legislators. 9 In Polsby 's view, the Parliament of Great Britain represents the other end of a continuum of legislative types which are both open and specialized and which he calls a parliamentary arena. The outputs of such bodies are best dealt with in terms of the social backgrounds of the legislators, the organization of parliamentary parties, extraparliamentary party politics and the internal debate. Theories of representation, whether free mandate or imperative mandate, are primarily intended to deal with the legislators of open and specialized legislatures. Yet city councils in the United States have not always been either open or specialized. Nor have councils always been conceived as transformative legislatures in the manner suggested by Eulau and Prewitt's concern with accountability. 11 Early American cities often had no specialized legislative bodies. Moreover, if the writings of scores of elite theorists have any validity, numerous cities throughout the United States have councils to which no ordinary citizen can have meaningful recourse. Such councils are little more than corporate boards. The corporate aspect of American cities is substantiated in more than circumstance. In law, they are considered municipal corporations. Traditionally, the control of cities over their own destiny has been severely circumscribed. "They are, so to phrase it, the mere tenants at the will of the legislature. "12

PAGE 291

2 72 The corporate aspect of local councils appears often in the reform literature of this century. During the Progressive era, the profligate ways and corrupt manner (and perhaps the threat of lower class takeover in the cities) of urban politics generated a rising tide of reaction to the patterns of municipal governance. New institutions and administrative arrangements were proposed to rationalize and democratize city government. The principal tools which the reformers picked to maximize this "representation of the city as a whole" were the commission, and later the manager form of government, and nonpartisan election at-large. City manager government, it was argued, produced a no-nonsense, efficient and business-like regime, where decisions could be implemented by professional administrators rather than by victors in the battle over spoils. *-3 Reform may have been aimed at corruption, but it also changed the nature of the local legislatures. The rhetoric of reform evinces images of a corporate board, autonomous from the pressures of community life, and free to pursue "good" policies for the entire community. The board would be democratically elected, but the reforms would free the members from "undue" community pressures and special interests. While the reform image of city councils resembles not so much a modern legislature as a board of trustees, the councils are open in Polsby's sense. The citizenry has access and, for limited purposes, can exercise influence over the deliberations. However, such councils are hardly transformative, rather they are conceived as parliamentary arenas. But they are small curious parliaments. There are to be no political parties. The pernicious distortion of community purpose brewed in political wrangles are to be eliminated if possible. Good policy, reform theorists argue, comes through reasoned debate. It is in such debate the parliamentary aspect of reform council is to be found and in the attributes of

PAGE 292

273 the mem bers themselves. The African reform theorists became literal, if not actual, disciples of the British reformers who for years had written: fo/wMnh ° 1 f Municl P al Government... will promote the object ofZtlt T 1 \° U8ht t0 be desi ^d...the maintenance of public order, the pure administration of justice or *fc e harmony and happiness of the societies to which it L to be 2nt of 'm " i tS / ireCt tende -y ^ to commit the management of Municipal affairs to the hands of those who from the possession of property have the strongest interest in good fnTin^ ' and fr ° m the ^ lif "ation of high characteT and intelligence, are most likely to conciliate the respect and confidence of their fellow citizens. U respect Reform theorists would give city government over to those who might best govern. As envisioned by reform theory, a city council member is a trustee. The council itself is a board, at best a parliamentary arena, where rancorous and raffish conflicts have little part. Politics is anathema to this conception of urban governance. And it is just this conception of local government with all its trappings which has come to dominate urban America. Nationally, 47 percent of all cities use the council-manager form of government. 15 m Florida, the proportion is even higher. Fully 89 percent of the 78 cities studied here have at least one of the reform characteristics (,ee Table 3^ of Chapter three for a list of characteristics). Anecdotal evidence appears to link reform structures and the reform ideology. Most of the councils in the San Francisco Bay region investigated by Eulau and Prewitt used the council-manager form. Sixty percent of all the council members surveyed could be categorized as trustees and only 18 percent saw themselves as delegates. Other studies yield similar results. -^ memIn short, while the purely administrative, independent council ber has always been an elaborate fiction, it is a fiction rounded-out and given substance by the expectation of the members and the poverty of a

PAGE 293

274 typical council's scope of action. Transformative councils exist, but the very expectations of city council members mitigate against their pervasiveness. City council members, the reform ideology has it, should be part-time legislators acting out of a spirit of community good. This conception is re-enforced with low pay, poor staff support, poor facilities, and an inattentive public. 17 Such conditions help shape each member's understanding of what being a council member means. Such circumstances, together with the traditional and conservative notion of representation, provide guidance for each member's actions. There may well be several different orientations a member can assume, but "regardless of what they may say, legislators do not have this wide a choice. "18 m such a manner the rules of the council game are established. These rules act to subdue rivalries into an organization. Open and generalized conflict among the council members is not part of the traditional view of their performance. This traditional view can be subverted by the immediate circumstance and the objective interests of the members themselves. A member's conception of the manner in which the council should perform may well affect his own behavior, but this conception is tied into the "common understanding" which is, in turn, affected by the changing nature of the world in which the council member lives. The incredible growth of all levels of government in the United States and the devolution of power to communities in what James Sundquist has called the "politics of deference" has altered the possibilities in many local governments. 19 Traditional reform structures and ideas have been challenged by new responsibilities and the emerging influence of councils and local government throughout the country. Even in reform

PAGE 294

2 75 governments, the council is involved in determining new policies. And with this new circumstance, a different understanding of representation and council behavior is evolving among council members and the public. It emphasizes value inclusiveness over mere competence. Traditional attitudes, which viewed conflict as anomalous or the product of idiosyncratic actions, are being challenged by ones which accept the political role of council members. Exchange Theory and Council Conflict The exchange perspective developed earlier suggests conflicts are most likely to occur when severe costs predicate no other course of action. In particular, council splits were associated with clashes of goals, predispositions and the external costs accrued by the actions of highly salient groups or individuals in the community. In the exchange perspective, the conflict or disagreement that characterizes a council is the result of a complex nexus which finds the member at the center. His background, JJLfe experiences, goals and ideological predispositions are fundamental to his response. However, these factors can be mediated by the manner in which his perceptions subjectively work to assign importance to various factors in his environment. Putnam captures this idea within the context of exchange theory by referencing shared interests: One important determinant of whether participants in a game will engage in cooperative problem-solving or conflictful bargaining is their image of the extent of shared or conflicting interests in the game. If the players believe that mutually beneficial solutions are possible, they will search cooperatively for them, even if the solutions are initially unknown. On the other hand, if the players believe that the discovery of new points beyond the present "outcome line" is highly unlikely, there is little incentive to engage in problem-solving. Instead their attention will be directed at achieving maximum individual gains in the bargaining process. Costs may determine such responses, but costs themselves grow out of the

PAGE 295

276 normative constitution of an individual; it is a construction of immense complexity. Nevertheless, expectations about costs can be formed. Eulau, for example, has argued council decisions are likely to be unanimous even when members act rationally. My frankly speculative discussion of the effects of representational modes on behavior suggests still another reason for unanimity. The council-manager form of government attempts to construct a structural propensity toward amicable agreement. 21 But the council-manager form of government has enjoyed an impact extending far beyond its structural manifestations. Reformist arguments have helped define how council members think of themselves and the council on which they sit. Taken together, the reform litany and the conservative representational mode are likely to act as a powerful agent acting toward consensus. And this encouragement extends beyond the decision-making situation per se. The search for amicable agreement is predicated on low levels of hostility among the members. Reform theorists would eliminate political wrangling and replace it with informed debate. Thus, in reform councils, members are expected to display civility, if not intimacy, toward their colleagues. Patterns of council relations should be related to members' understanding of the rules-of-the-game, if this argument is correct. Unfortunately, this argument cannot be tested directly. However, the relationship can be hinted at by referencing a measure of council performance not used previously (see Appendix A) . Each member was asked to evaluate the performance of his council on seven items relating to the council's ability to pursue its purposes and

PAGE 296

277 to exercise administrative and policy initiatives (PERFORM) . Evaluations of performance should be linked to ideological predispositions and to perceptions of council conflict. If conservatives share an acceptance of reformist arguments and the traditional representational role, they would view conflict as detrimental to performance. Liberals, with a less administrative conception of government, would view conflict more positively. In fact, liberals should be much less willing to judge performance in terms associated with the amount of conflict in which the council engages. These expectations are borne out in the data on Florida council members. Among all council members and among the councils taken as a group, there is a significant negative relationship between the evaluation of council performance (PERFORM) and the perceived amount of council conflict (CONFLICT) . There is also a positive relation between predispositions and performance among the responses of the members and among the councils. In SES group F, however, such relationships do not hold. Neither predispositions nor conflict are linked to performance. This same pattern appears when the relationships between council relations and evaluations of performance are examined. At the individual level and among all councils, most of the associations are statistically significant. Among the cities of SES group F another pattern emerges. Here the links are sporadic and accord only in their direction with the two basic patterns of relations discerned in the last chapter. Consensual relations such as participation (PARTICPT) and viscidity (VISCIDTY) are positively associated with measures of performance (PERFORM) ; conf lictual relations such as stratification (STRATIF) are negative (see Table 7—2 ) . Thus, if the connection between relations and predispositions is

PAGE 297

278 well drawn, then it can be supposed the reasons the councils of SES group F are different lies in their liability to conflictual relations. Table 7-2 Kendall's Tau-c Between Measure of Council Performance and Council Characteristics (Relations) PERFORM All Members All Council Group F PARTICPT .485 .560 .462 STRATIF -.250 POLARIZE -. 276 INTIMACY . 199 PRESSURE -.402 AUTONOMY . 271 VISCIDITY . 322 (N = 169) (N == 61) (N = 12) (Italicized Relationships Significant at .05 Level) This discussion of the manner of association between the expectations of council members about the nature of their job and their behavior in that job has been intended to suggest yet another factor in the exchange analysis of councils. Clearly, expectations must be considered in attempting to understand the processes whereby a council becomes involved in political rows. The conflicts which city councils evidence are the product of a complex set of factors. Predispositions, differences among the council members and their relations help determine the degree of conflict which characterizes a council. But the members' orientation cannot be extracted from the context of the council. For example, predispositions typically condition a member's involvement in conflictual -.364

PAGE 298

2 79 situations. But in high SES cities, even conservatives are willing to engage in political conflict. The link between predispositions and an orientation to conflict is not a purely stylistic feature, but rather the outgrowth of the historical context of the council. The rules of the game that prevail in the councils at large are not those which persist among SES group F councils. Nonetheless exchange theory maintains conflictual political situations remain important. They still gauge situations of high cost. Council Conflicts The political rows in which a council member is involved are diverse, and often are difficult to understand. For sake of convenience then, I asked each member to rate the frequency of disagreements among the council members over matters of public policy, and to gauge the frequency of such disagreements with the city manager or mayor. These responses formed the basis of the CONFLICT measure, a measure of the total amount of conflicts related to policy in which the council was engaged. Notice that the measure includes disagreements with the city manager. The CONFLICT scale does not relate solely to member-member relations. It is intended as a summary measure. Figure 7-2 displays the manner in which each of the SES groups are rated on this summary measure when the mean conflict for the council in the group is calculated. The most striking feature of the figure is the manner in which the high and low SES groups appear at opposite ends of the scale. High SES cities have councils whose members report much greater levels of conflict than their counterparts in the low SES cities. Both groups are more than one standard deviation apart in their respective directions, from the mean of all councils. Only one city council in

PAGE 299

Z-Scores of Council Means by SES Groups on CONFLICT 280 1.5-i 1.0__ A .C E D -1.0 -1.5HIGH CONFLICT 10.5 10.0 A C 9.0 _E D 8.0 7.5" LOW CONFLICT Z-Scores Figure 7-2. Conflict in the SES Groups Conflict Scale (Range 3-15) CONFLICT Scores by SES Group

PAGE 300

281 group B (Belle Glade) reports higher conflict levels than the lowest council in group F (Gainesville) . Figure 7-2 also demonstrates a curious symmetry between SES groups and conflict levels. Groups F and B were at opposite ends of the incomestage in the life cycle factors of the discriminate functions which were used to construct the SES groups. Groups C and D were poles of the sizeethnicity dimension (see Chapter seven) . The larger cities of group C display higher conflictual politics in their councils than do their smaller opposites. The two remaining groups of cities whose centroids are also distinguished on the income-state in the life cycle function appear quite different on the conflict scale. The older retirement councils of group E are less conflictual than those in group A. Given these relationships, it is easy to imagine conflictual relations in the Florida councils grow out of the predispositions (or orientations toward being a council member) of the councils. In a previous chapter it was shown large cities and high SES cities (groups F and C) were the most liberal on the summary measure, while low SES cities (group B) were the most conservative. However, an examination of the nest of empirical relationships between CONFLICT and other variables shown in Table 7^ only partially substantiates this view. Conservative predispositions are associated with conflict, but only religious conservatism is statistically significant. However, as might be expected from the findings on council relations presented in the previous chapter both the consensual and conflictual patterns demonstrate statistically significant relationships with measures of total conflict. Not surprisingly, conflictual relations are positively associated with conflict, while consensual relations are negatively associated. And these two patterns are re-enforced by

PAGE 301

282 Table 7-3 Kendall's Tau-c Between Selected Measures Of Council Life and Total Council Conflict For Ail Councils (N = 61)

PAGE 302

283 religious conservatism. City councils whose members evidence the most concern for matters of discipline and punitive punishment — high religious conservatism — are most likely to form consensual relations and to report low levels of conflict in the political dealings of the council. These same councils also indicate an absence of conflictual relations. As can be seen in Figure 7-3 , the re-enforcing associations are much the same as those discovered earlier in the investigation of council relations. Once again an attention to governmental information sources is shown to be associated with those councils which are most consensual. The most conflictual councils report governmental information sources to be unimportant. This is particularly important to note with regard to the CONFLICT measure which includes disagreements between the council and the city manager or mayor. It seems entirely reasonable that as the council begins to have disagreements with the city manager they would rely less on him as a source of information about affairs in their city. However, it is less reasonable to expect disagreements among members to affect this relationship. Such disagreements might be expected to spill over into relations with the manager but the association between conflictual political situations and rejection of the manager as an information source should be greater than that between conflictual relations and rejection of the manager. Indeed this is just what is found. Conflict is very strongly and negatively associated with governmental cues, while the highest association between relations and a reliance on governmental information cues is substantially weaker. In an extremely conflictual political situation, relations between members and the manager break down. Governmental bodies cease to be a likely source of information. These same relationships are not demonstrated among the councils

PAGE 303

284 [STRATIF] [POLARIZE] [PRESSURE] [CONFLICT]. [(-)CONSREL]« [(-)GOVTQ] N [(-)AGE] (-) [PARTICPT] [INTIMACY] [AUTONOMY] [VISCIDITY] CONFLICTUAL RELATIONS CONSENSUAL RELATIONS (a) For All Councils (N = 61) [(-)CITIZENQA] [PRESSURE]^ [CONSRELA]' [CONFLICT] [STRATIF] (b) For Councils of High SES Cities of Group F Figure 7-3. Closed Network of Bivariate Relationships Centering on Measures of Council Relations and Total Council Conflict

PAGE 304

285 of SES group F. In fact, the re-enforcing relationship network involving total council conflict (CONFLICT) is quite simple compared to that among the councils at large. The generalized negative relationship between consensual relations on the council and conflictual relations has disappeared. The relationships are no longer statistically significant. Rather only two of the conflictual relations are associated with conflictual political situations. The councils under the most pressure and those which are most stratified are the ones which report the highest levels of conflictual politics. As can be seen in Figure 7-3 , these two factors are also related to differences in the religious conservative beliefs of the members. The greater the differences among the members on religiously conservative predispositions, the greater the stratification and pressure reported by the council. Such differences are also related to conflictual political situations as well. At first glance, these relationships might appear inexplicable. However, if the properties of the councils in group F are remembered, they become more understandable. The clue here is the negative relationship between differences in a reliance upon citizen information sources (cues) and conflictual political situations. The less the differences among members in their reliance upon citizen cues, the more conflictual the political relations. This relationship can only be understood if it is remembered SES group F councils rely upon citizen cues much more than other SES groups. On the CITIZENQ scale, the mean for the group F cities is 5.65; the mean for all councils is 4.11. The group F cities exceed the mean by more than a single standard deviation. Thus differences among the council members are most likely to result when some members do not accord citizen cues importance, and a shared concern for citizen cues is

PAGE 305

286 likely to result when all members believe citizen information cues are important. As seen if Figure 7-3 , when all members are basically attuned to citizen cues as important sources of information, community pressure is believed to be higher. Similar reasoning can be used to explain the manner in which great differences in religious conservatism are linked to indications of community pressure. Members of the group F councils are more liberal than those of any other group; the mean for group F is 6.6, while among councils as a whole it is 8.6. When members display small differences on this predisposition, it is probable they are liberals, and it is religious liberals who are most likely to report the presence of community pressure. Councils identifying substantial amounts of community pressure are likely to also report conflictual political situations. This association squares with the earlier discussion of group F councils as arenas where the politics of deference had arrived. The evidence marshalled thus far suggests the councils in SES group F, the high SES cities, are likely to be transformative arenas, to use Polsby's term. Nonetheless, the councils differ among themselves in the amount of conflict present. Figure 7-4 arrays each of the councils in group F along the CONFLICT dimension. The severe conflict noted earlier among the members of the Boca Raton council is dramatically illustrated here. Boca Raton and Melbourne rate the highest on the CONFLICT scale. From observation, however, I can report the conflictual political situations are somewhat different. In Melbourne during the fall and winter of 1975-1975, the council was divided. Environmental forces were cast against more growth-oriented individuals. The Melbourne city manager was the focus of much of the hostility. Charges flew back and forth between

PAGE 306

287 HIGH CONFLICT 1.5 -i 1.0 0.0 -1.0 LOW CONFLICT -1.5 J Melbourne Boca Raton Bartow Rockledge Ft. Waltcn Beach Titusville Hialeah Tallahassee Cape Coral Cocoa Beach Cocoa Gainesville Figure 7-4. Z-Scores of Council Means in SES Group F on CONFLICT

PAGE 307

288 the mayor , the council, the city attorney and the manager. The city which had been formed out of the merger of Eau Galli and Melbourne in 1969 was divided by the charges. The council members, several of whom had served on the councils of the two cities before their consolidation, divided along traditional lines. The manager who had previously served as the manager of Eau Galli was at adds with a new mayor and the emerging forces within the old city of Melbourne. The conflicts in the Boca Raton council also spilled over into the manager's office. Some council members accused the "city hall gang" of supporting one faction on the council. However, the divisions on the Boca Raton council were much more keenly centered on policy disputes between the members. The pattern of personalized divisions which marked the Melbourne council did not characterize the Boca Raton council to the same extent. Decision-Making Modes An exchange theoretical perspective argues conflict is important in determining the types of policies which a council promulgates. But the importance of conflict lies in its association with the costs experienced by the members in dealing with policy. Such costs are not limited to the immediate circumstance of a particular issue. In the definition of policy I have adopted, policy is an outgrowth of consistent and regular patterns of member bahavior. The relations members form are a part of these patterns. But the relations themselves need not be part of the manner in which members attempt to resolve the issues confronting them in the institutional arena. It is not difficult to imagine a council whose members may be quite polarized with respect ot the goals they hold for the council and the community and yet quite adroit at producing political so-

PAGE 308

289 lutions to pressing problems. It is true conflicts can transcend a given decision and spill over on all a council undertakes. But it is similarly true councils can evolve methods of dealing with the relations that prevail among the members. Eulau and Prewitt go so far as to suggest compensating factors will in general act to smooth a council's performance. 23 There is a learned aspect to the benefits and costs accrued in council decision making. Over time members learn how their behaviors will affect their colleagues; they learn how certain issues will likely affect each other member. Occasionally, councils split into persistent factions, and as with the Boca Raton body, the vote on any given issue can be predicated safely. More often, however, the voting patterns of councils are unclear. In any council, the sheer number and diversity of issues, together with the limited amount of time available for their study, act to promote unanimity in many voting situations. But even unanimity can be organized in many different ways. In L^yrinths of Democracy, Eulau and Prewitt offer a fascinating and extremely clever presentation of the manner in which councils form decision structures. Three decisional structures were identified. In bipolar structures clearly identified factions persist. Nonpolar structures have no set voting pattern, while in unipolar structures the members vote together. 24 Empirically, Eulau and Prewitt discover such structures on the basis of several questions asked of each member. (I have used their questions in building some of my own scales). Analytically, each of these structures is related to three behavioral dimensions used to characterize council relations. For example, bipolar structures score high on conflict, low on permissiveness, and moderate on fragmentation. The ingenuity of this characterization is considerable. Eulau and Prewitt

PAGE 309

290 are able to deal with both the behavioral components of decision structures and the decision structures themselves by merely permutating the arrangements among the structures on each of the behaviors and forcing different ordinal measures. 25 As clever as this approach to decision structures is, it has distinct drawbacks. The most important is the implicit assumption that councils evidence distinct structures. They assume councils can be characterized by a single decision-making pattern. And in fact, among the 82 councils with which Eulau and Prewitt were concerned, they present evidence which provides some justification for the characterization of a council and the various ordering*. 2 * Among Florida councils, the patterns are not as clear cut as among those in the San Francisco Bay Region. The policy disagreement scales which correspond to those used by Eulau and Prewitt show considerable overlap within councils. Table 7^ arrays the number of councils reporting high scores on each of the scales. Thirty-one councils report a higher than average incidence of generalized council splits (SPLITGEN is equivalent to bipolar councils); 2k councils indicate a higher than average score on idiosyncratic council splits (SPLITIDO is Eulau and Prewitt's unipolar); and A3 councils identify themselves as higher than average on particularized council splits (SPLITPAR is nonpolar) . These figures suggest the absence of the type of exlcusionary decisional structures found by Eulau and Prewitt. Table 7^5 which shows the scores of the councils of SES group F on these scales offers even more dramatic proof of this fact. Only four of the 12 councils can be placed squarely in one of the decisional structures. The remaining eight display mixed patterns, defying a single characterization. Thus, on empirical grounds alone, the typology of decisional structures offered by Eulau and Prewitt is ill-suited for a description of Florida city councils. 27

PAGE 310

291 Table 7-4 Number of Councils Above and Below the Sample Mean On Policy Disagreement (MODES) Scales (N = 60) Above Mean (X > X) Below Mean (X < X) SPLITGEN

PAGE 311

292 Eulau and Prewitt's conception of decision structures has further ramifications. As they state: Implicit is the question (of whether structures can be so classified) was, of course, the assumption that councilmen line up with each other more or less regularly regardless of the content of specific issues. If one were to ask this question of legislators or judges on other levels of government, it would be absurd and ridiculous. But at the local level and in the case of small groups it is neither absurd nor ridiculous. 28 This conception of decision structures forecloses the possibility council members may be triggered by policy content in much the same fashion as legislators at other governmental levels. Research by other urban scholars would suggest such foreclosure may be premature. 29 There is more substantial reason for not adopting the scheme presented by Eulau and Prewitt; theoretically, it is at odds with the notions developed in the exchange analysis presented earlier. It rejects the possibility council politics may turn on issues themselves, and suggests interpersonal relations may dominate the decision-making processes of the council. While interpersonal relations may play a crucial role in such decision-making processes, exchange analysis argues a specific issue or set of issues may generate costs which contravene typical relationships. In a sense, then, the term "decisional structure" is misleading. It denotes a single patterned set of interrelationships which comes into play in any decision-making situation. In Florida city councils, members identify different patterns of response to policy situations which co-exist in the same council. The decision patterns which typify a single council are not so much structures as modes which the members access in different circumstances. Within a single council, there might be several different decision-making modes and these modes should not be thought of as primar-

PAGE 312

293 ily involving disagreements among the members. The exchange perspective suggests unanimity is likely to be a pervasive phenomenon among city councils. But unanimity is merely the outcome of a vote. There are several different ways in which it can be reached. Each of these "ways" can be thought of as a different decisionmaking mode. Different modes can be used to extend and enlarge the agreement on the issues which come before the council. Decision modes can be placed in two broad categories: conflictual and consensual. Each city council member was asked to first comment on reasons behind the unanimity on his particular council. Four distinct agreement modes emerged from the responses. The first was labeled "AVOIDCON." It involves items which suggest a form of anticipated logrolling; bargaining takes place at a very subtle level. Each member sees himself as part of a group which will continue to operate for quite a while. 0vert conflicts are to be avoided as a hedge against the need for future cooperation. A second mode, called "INFORMAL," grows out of an informal approach to council politics. Council conflict is to be avoided, but not in the sense of the previous measure. There is no sense of members trading off on issues. Rather, differences are something which should not be aired in the council chamber unless absolutely necessary. It is part of the rules of the game. A third mode of council agreement, labeled "ALLONE," turns on the perceptions of the members that they are in basic agreement on the issues. They have no need to differ. Finally, many members suggest consensus is a goal in itself. This mode resembles amicable agreement; I have designated it "CONSENT." Each member attempts to work toward involving each other member in the resolution of every issue. Each of these modes can be adopted for any single issue facing any

PAGE 313

294 of the councils. However, circumstances can be expected to condition the use of the modes. For example, it seems improbable that basic agreement (ALLONE) could emerge with great frequency where the members were at odds over their ideological predispositions. Further, informal decision modes can be expected to occur most frequently where the members are intimate or share a consensus over the goals of the council (VISCIDITY). Similarly, serious problems facing the council would seem to mitigate the use of such a mode. When might a form of amicable agreement be expected to emerge? Again it might be expected the council members would have to share close relations of some form. Surely a severely polarized council should not be expected to work in such a fashion. Nor would bargaining appear to be the correlative of conflictual council relations. Logrolling or bargaining is a conscious tactic or decision mode commonly associated with transformative legislative bodies. If, as I have suggested previously, the accident of conservative dominance in Florida city council and the reform ideology have acted to re-enforce one another, then predispositions themselves should be linked to agreement modes. In particular, liberals should resort to logrolling tactics; conservatives will access the other three agreement modes. Of these three modes, the ALLONE measure appears the least likely to be associated with predispositions. Per force, basic agreement or consensus is indicative of a phenomenon which relates to the differences among the members and should be independent of their orientations. Amicable agreement and informality, on the other hand, appear to be likely candidates for conditioning by these very orientations. For example, the reform ideology views council governance as similar to the behavior of corporate board decisions. Such a view would hold that a united front should accompany each decision.

PAGE 314

295 The conflictual modes which were measured among the members' responses were developed from response codes used by Eulau and Prewitt to corroborate the presence of decision structures (see Appendix A). Each of the three scales were formed from items which attempted to account for the manner in which the councils divided over issues. The items in each scale grouped together in much the same fashion as those among the San Francisco council members. The firstmode, called Generalized Splits (SPLITGEN) , was used by Eulau and Prewitt to indicate bipolar councils. It is associated with conditions of a generalized pattern of differences among the members when making decisions. These differences include ones over policy, background, ambitions, and supporters. A second mode involved what I call Idiosyncratic Splits (SPLITIDO) . This is a disagreemment mode in which members report the divisions on the council either do not occur or occur as momentary and fleeting personal actions. The splits are difficult to explain. This mode is distinct from splits which are associated with differences of opinion on particular issues (SPLITPAR) . Here the council members link council division to the nature of the issue confronting the council. Most council members were likely to suggest that just such a mode often existed in their council. Each of the disagreement modes can be expected to emerge from a different set of conditions. Generalized splits might well flow out of the differences between the members. The fractious effects of interpersonal differences should play a role in producing a generalized split among the members. It is reasonable to assume members who are intimate, or who share views on the nature of council goals are unlikely to divide in a general fashion over policy. The nature of decision costs predicates continuing effects. Thus council polarization, stratification, and community

PAGE 315

296 pressure can be expected to produce generalized splits. If idiosyncratic splits are really idiosyncratic, then no pattern of association should be evident between this mode and measures of council relations. However, idiosyncratic splits (SPLITIDO) also may be taken as the measure of the absence of the use of disagreement modes in the decision-making process. This interpretation implies associations with a conservative predisposition, an absence of serious problems, older council members, and greater years of experience on the council. SPLITPAR measures a decision mode in which members do have differences with one another. Moreover, these splits can be defined in terms of the issues which come before the council. Unlike either generalized splits or idiosyncratic splits (absence of real conflict), such council divisions require the relations among members be such that splits over specific issues will not poison the relations of the members. In other words, the members must already have relatively harmonious relations in order to be able to disagree over an issue without generating the types of costs which are liable to exacerbate relations on the council. Hence, particularized splits should be encouraged in councils where conservative predispositions are surrogates for that complex of beliefs associated with reform ideology and traditional forms of representation. The Correlates of Decision-Making Modes One way of viewing the prevalence of the various decision-making modes among the councils of Florida is seen in Table 7-6 . The pattern of referencing the modes is dramatically illustrated when each of the SES groups and their councils is rated on the seven scales and standardized to a common range. Council members report most frequently that agreement is based on attempts to build toward consensus (CONSENT) . However, these

PAGE 316

29 7 councils also reflect unanimity among members; overall scores on the consensual mode (ALLONE) are high. Significantly, it is the low SES cities, group B, which score highest on these two agreement modes, and the high SES cities lowest. But even in group F city councils, these modes are identified more frequently than either the informal (INFORMAL) or bargaining (AVOIDCON) modes. Table 7-6 Council Means on Decision-Making Mode Scales For SES Groups: Normalized in Range From 1 to 10 Councils of SES Group B C D E Agreement Modes

PAGE 317

298 cils of the functionally large cities of group C. Indeed, it is councils in the largest cities which also report the most frequent use of the idiosyncratic decision mode. Not surprisingly, the councils of the low SES cities in group B evidence the least use of the generalized disagreement mode. At this level of analysis, it appears the low occurence of generalized splits is occasioned by the greater frequency of general agreement among the members. In order to better explicate the factors which promote the tendency of councils to reference certain decision-making modes, I turned to the associations between the modes and the member's personal attributes, their perception of contextual factors and council relations. These associations are presented in Table 7-7 . The character of these associations is striking. As a rule the disagreement modes are not statistically related (significantly) with any cluster of variables outside of council relations. The link between each of the personal attribute variables and idiosyncratic splits is the main exception. Overall, the associations between agreement modes and the several clusters of variables is not much better. Again council relations are associated with several of the agreement modes. But for agreement modes, predispositions do play a role. Two of the modes associate poorly with the many variables I have used to capture the council situation. The consensual (ALLONE) and idiosyncratic split (SPLITIDO) modes are involved in only three statistically significant relationships apiece. The consensual mode is unlikely to be referenced in religiously liberal, highly stratified councils where members indicate their willingness to serve on the council for several more terms. Idiosyncratic splits are described most often in councils where

PAGE 318

299 a, 01 oTJ CO i-i -cr o
PAGE 319

300 the members are older, less educated and relatively inexperienced with council life. The remaining disagreement modes provide an interesting contrast both to the idiosyncratic modes and to one another. First, neither is associated with life experiences or member goals. And only one — generalized splits (SPLITGEN) — bears any relation to member predispositions. Nor are either related to many of the members' perceptions of contextual factors. Rather it is council relations which act to condition their use. Generally speaking, conflictual relations promote generalized splits while consensual relations bear on particularized splits. But this is phrasing the relationship too simply. Certainly the assymetry is not perfect. However, in the councils where viscidity (VISCIDITY) is high and community pressure is low, particularized splits (SPLITPAR) are encouraged, while when these conditions are reversed generalized splits (SPLITGEN) appear frequently. This observation is re-enforced by one of the few other statistically significant relationships in which these modes are involved. Namely, to the degree citizen information sources (CITIZENQ) are relied upon, generalized splits (SPLITGEN) occur. Where such a reliance is not reported, particularized splits are referenced. Of the four agreement modes, three can be described in terms of the variables employed here. The fourth, consensus (ALLONE) was dealt with above. Both the bargaining (AVOIDCON) and amicable agreement (CONSENT) modes display a complex pattern of interrelationships with the many variables. In many respects they are quite different. Bargaining occurs most frequently in liberal councils; amicable agreement marks conservative councils. Bargaining is referenced most where conflictual relations obtain; amicable agreement appears in the midst of consensual relations.

PAGE 320

301 Perhaps the most interesting pattern of relationships between an agreement mode and the several variables describing council life involves informal (INFORMAL) decision-making. The informal mode, it should be remembered, is linked to the view that disputes among the members should not be aired publicly, that a sense of council order should prevail and arise out of cooperative Interaction of members. It is surprising then that this mode should be associated with polarization (POLARIZE) among the members. However, this frankly unexpected relationship is explicable in light of the other associations it bears and the manner in which polarization acts to condition the use of certain modes. Informal decision making (INFORMAL) is frequently cited in religiously conservative councils, but not in politically conservative ones. It is referenced where members see themselves as relatively autonomous from the community. We can speculate then that this mode might be used in councils whose members and their perceptions augur for consensual relations, but which, for some reason, have become polarized. Informal decision making represents an alternative to bargaining or a generalized split. The use of decision-making modes in the councils is highly complex. Before attempting to categorize how such modes develop and what their correlated are, it will be useful to examine the modes and their use in the councils of the high SES cities. Decision-Making Modes in Group F Previously, it has been noted that the character of associations which obtain among the variables describing the political situation of the councils in SES group F is somewhat different from that of the councils as a whole. It might be expected then that the propensity of the group F councils to use decision-making modes is also different. But be-

PAGE 321

302 fore examining the correlates of the various modes within this group, it is instructive to understand how the cities stand in relation to one another in their use of the different modes. Table 7-8 contains the Z-scores for each of the councils in group F along each of the decision-making modes. These statistics suggest a complex pattern of referencing used by councils, even councils which share a relatively similar socioeconomic environment. Certainly the exclusiveness Eulau and Prewitt found is not evident. Only councils such as Boca Raton, Melbourne, and Cocoa even approach their standards. Moreover, the scores indicate the relationship between modes and total council conflict (CONFLICT) may not be simple. For example, Cocoa Beach is among the lowest cities in the group on the CONFLICT scale, yet it ranks high on the generalized split mode (SPLITGEN) . There is an additional element to this complexity. The councils of the SES group F cities report the highest total council conflict (CONFLICT) of any SES group. However, these councils are not ranked as the most frequent users of any of the disagreement modes. They rate as the least likely group of cities to use the amicable agreement (CONSENT) and consensus (ALLONE) modes. These patterns indicate that the distinctive effects of conservative predispositions can be expected to exist in the high SES cities. Although predispositions should involve agreements primarily, their importance in the group F councils tempers our expectations about their connection with the disagreement modes. The high SES councils are significantly more liberal on the CONSREL scale than the other SES groups. In effect, examining the patterns of association among the group F councxls acts as a control on high religious conservatism. The re-enforcing effects of conservative predispositions and the reform litany should not prevail among the councils of group F.

PAGE 322

303 o to CD u o CJ CO I N erf < 00 u0 in CO m en CO in CO m CO in CO m m o m r-co m co m

PAGE 323

304 Kendall's tau-c was calculated in order to gauge the strength of the association between each of the decision-making modes and the several variables which have been used to characterize each council. In keeping with the format developed earlier, these relationships have been summarized in Figure 7-5. Again I remind the reader that these networks are not path diagrams, merely the pattern of re-enforcement among statistically significant variables. One of the most noticeable aspects of the composition of the empirical relationships involving the generalized splits (SPLITGEN) is the absence of a re-enforcing association with either predispositions or differences in ideological predispositions among the members. While differences in political conservatism encourage the use of this mode (Kendall's tau,c = .356), this association is not statistically significant. As with the councils at large, it is the nature of the council relations which are associated with the tendency to adopt this mode. Again conflictual relations encourage its reference, consensual ones discourage it. Among the councils at large, idiosyncratic splits (SPLITIDO) give the appearance of being just that. Only the life experiences of the members are associated with this mode. The situation is different among the councils of group F. All forms of conservatism appear linked with this mode. In fact, conservative councils reference both this mode and particularized splits (SPLITPAR) as rationales for the manner in which they divide over public policy. Conservative councils, it seems, are much less likely to adopt generalized patterns of division than liberal councils, they are more geared to particular moments and particular pieces of legislation than are liberals. This relationship stands up even when other factors impinging upon the decision-making mode are controlled. An in-

PAGE 324

305 (a) [(-) INTIMACY ]-r [FINPROB] /[(-) AUTONOMY] ;[SPLITGEN], [PRESSURE][(-) VISCIDITY][POLARIZE] [EDUC] (b) [CONSPOL]: ,[(-)SERPROB]' [SPLITIDO] 2 •[CITIZENQ] I [YREXP] [(-)GROWPROB] (c) [(-)PEOPLEA] = r ,[SMALLINF]-[SPLITPAR]' [GROUPQ] [CONSTOT] I (d) [GROUPQ]> [CONSENT]. [(-)STRATIF] '[AGE] (e; [GOVPROB]. [ALLON'E].[(-)CONSTOTA]' [(-)STPATIF] Figure 7-5. Closed Networks of Bivariate Relationships Centering on Decision-Making Modes for Councils of the High SES Cities in Group F

PAGE 325

306 (f) [EDUC] [(-)CITTZENQ] AVOIDCON" [(-)CONSPOLA] [(-)BUSINESSA] (-)CONSPOL] (g) [(-)HUMPROB] [GOVTQA] [AGE] [ AGE ] [ INFORMAL ] [VISCIDITY][ CONSTOT ][ INFORMAL ]« -[ YREXP A ] •[(-)EDUC] [YREXP] Figure 7-5. Continued

PAGE 326

307 spection of the first-order partials reveals predispositions to be of continuing importance in the tendency for councils to adopt such modes of disagreement . Council relations were important for particularized splits among all the councils. In group F councils, council relations are not linked to scores on the SPLITPAR scale. Thorough-going conservatism appears to be the key to understanding this mode's use in group F. The substantial association coefficient (tau-c = .566) between the reference of this mode and conservatism also gives meaning to the association of this mode, group information cues, and differences among the members in their view of people-oriented groups. It is conservatives who are most likely to view group cues emanating from the chamber of commerce as important, and it is among conservatives that differences over people-oriented groups is least pronounced. The association between conservatism in a council and the amicable agreement mode (CONSENT) exists among the councils as a whole. This relation disappears in group F. However, the negative association with stratification (STRATIF) remains. This relation also reappears in the association between the consensus mode (ALLONE) and stratification. However, in the case of consensus predispositions do play a role, but it is the differences in the predispositions of the members which is crucial. The fewer the differences among the members, the greater the probability the council will indicate the consensual mode is used. Logrolling or bargaining (AVOIDCON) is linked with small differences among the members' predispositions. However, this mode is also cited where liberalism (CONSPOL) is high. Thus, it would appear bargaining is referenced in conditions where there is little difference among liberal members.

PAGE 327

308 Council relations do not play an important role in establishing the conditions for bargaining among the council members of group F. This is quite unlike the pattern among all the councils. But such relations are key in establishing the informal decision-making mode (INFORMAL). Council viscidity is strongly linked to informal practices used in reaching unanimity. A glance at Figure 7-5 will reveal that bargaining and informal decision modes are likely to be used by councils with dissimilar characteristics. In this way, the patterns of these two modes resemble those seen in Table 7-6 . But among the councils of group F the variables involved with the modes differ from those seen earlier. The correlates of the decision-making modes among the councils in the sample and among those of SES group F bear striking similarities and differences. Those variables associated with specific agreement modes act in somewhat the same fashion in both groups of councils. The consensual (ALLONE) and amicable agreement (CONSENT) modes appear less in stratified councils; bargaining (AVOIDCON) occurs in liberal councils. Informal (INFORMAL) procedures characterize conservative councils. Differences between the two groups of councils appear when the disagreement modes are considered. Generalized splits (SPLITGEN) are associated with predispositions and council relations in the sample. In group F, relations dominate. Idiosyncratic splits are associated only with life experiences among the sample councils; several other factors emerge in group F. Particularized splits (SPLITPAR) surface where relations are consensual in the sample of councils; they appear to have no connection in group F. Overall, member predispositions and the differences among members on predispositions play a much more important role in the associations in the SES group F councils than among the councils as a whole.

PAGE 328

309 Decision-Making Modes and Conflict From the discussion above, it is easy to imagine that the decision modes must complement one another in some fashion. For example, it might be argued that since bargaining and informal practices appear to involve contrasting predispositions among the councils of the group F cities that they would not be referenced in the same council. Perhaps! Table 7-9 should illuminate this expectation. There is no significant relationship between any of the various agreement modes in either the councils as a whole or in the SES group F councils. Only the relationship between the prevalence of logrolling and amicable agreement in the group F cities bears close scrutiny. Here the relationship is fairly strong (Kendall's tau-c = .403), but it is statistically insignificant at the .05 level. The suspected relation between logrolling (AVOIDCON) and informal decision modes (INFORMAL) is in the right direction (negative), but again is not statistically significant in the group F cities. Similarly, an inspection of the association between various disagreement in both groups of cities shows no statistically significant relationship. A council's referencing of a specific disagreement mode can complement the agreement modes which it uses. However, Table 7-9 reveals the complementarity of modes is affected by the councils considered. For example, there is a moderate and statistically significant relationship between generalized council splits and anticipated logrolling in the Florida city councils. But in the high SES group F cities, the direction of the relationship is reversed although not statistically significant (Tau-c = -.340). There are four other reversals in the direction of the relationships among the decision-making modes. And while none of these reversals

PAGE 329

310 H

PAGE 330

311 occurs where the relationships are statistically significant in each direction, such a phenomenon should not be left unexplained. One plausible interpretation of such reversals would turn on the amount of conflict each council is involved in. Group F councils experienced significantly higher levels of conflict than councils in the five other SES groupings. Such conflict levels might be expected to have ramifications, for the decision-making modes adopted by the councils. In councils where generalized splits are prevalent, the nature of the conflict might make the logrolling which usually occurs in such councils impossible. However tempting the explanation, it does not seem to be born out in the data. It is difficult to understand how such conflicts could at once mitigate against logrolling and yet be conducive to amicable agreement. It seems reasonable some other factor is contributing to the reversal. Perhaps the change is due to the different ideological predispositions of the members. Group F is, after all, the most liberal SES grouping on both the summary scale and religious conservatism (CONSREL) . However, an inspection of the first order partials, controlling for predispositions, suggests neither political nor the summary index of conservatism helps to explain the phenomenon. Religious conservatism, so often a variable with strange effects, also fails. In fact, no single set of variables seems to adequately account for all the reversals. However, two sets do a fair job. The first is the differences in the ideological predispositions of the members. Controlling for either political or religious conservatism tends to lower or reverse the sign of associations between generalized splits and anticipated logrolling. In a similar fashion controlling for the problem perceptions of the members, especially

PAGE 331

312 growth-related and financial problems, can act to reverse the directions of other relationships. For example, when controlling for financial problems in the group F cities, the association between generalized splits and amicable agreement returns to that of the councils as a whole. Of course, such an explanation is not entirely satisfactory. But a more detailed analysis of this phenomenon would require a larger number of cases if the analysis is to proceed much beyond first order partial associations. The reversals which characterize several of the associations can be made to seem more important than they really are. When the N of the sample is small, in this case 12, the chance for random error is large. Thus it becomes imperative to rely on levels of significance. However, a , casual reliance on such coefficients can often be oversimplified and unsatisfactory. It is important to look at the cases involved. SPLITGEN and AVOIDCON are negatively associated in the councils of SES group F. Another way of considering this relationship is to identify the cities which form this pattern. One convenient way of doing this is to consider each council's scores on the various decLs ion-making modes in relation to the average of all councils. Each council can then be identified as either above or below the average. With such information, simply and easily understood two-by-two tables can be constructed which portray the nature of interaction between the various modes. Table 7-10 displays the number of cities in group F above and below the mean on the Table 7-10 Generalized Disagreement and Bargaining in SES Group F SPLITGEN Low High High 3 2 AVOIDCON Low 3 4

PAGE 332

313 SPLITGEN and AVOIDCON scales. Here, the nature of the negative relationship is graphically apparent. In group F cities, seven of the cities lie on the negative diagonal. An assymetry exists between the modes. The high anticipated logrolling cities of group F are Bartow, Gainesville, Rockledge, Tallahassee, and Titusville. Those with generalized splits are Boca Raton, Cocoa Beach, Hialeah, Melbourne, Tallahassee, and Titusville. The overlap or consociation of these two modes occurs only in Tallahassee and Titusville. Strangely enough these two cities are also involved in relation between SPLITGEN and CONSENT (which in the group F cities is positive) . The cities which most frequently reference the amicable agreement mode in group F are Cocoa Beach, Gainesville, Rockledge, Tallahassee, and Titusville. Thus Cocoa Beach, Tallahassee, and Titusville form the overlap of cities high on both these modes. But the positive direction of this relation comes from those cities which use neither generalized splits nor amicable agreement, as Table 7-11 shows. Table 7-11 Amicable Agreement and Generalized Disagreement in SES group F CONSENT Low High High 2 3 SPLITGEN Low 4 3 These four cities are Bartow, Cocoa, Ft. Walton Beach, and Cape Coral. The complementarity of modes among the cities of SES group F is fairly complex. Only two of the 12 cities have identical patterns: Tallahassee and Titusville. These cities have higher than average scores on SPLITGEN, SPLITPAR, AVOIDCON, CONSENT, and lower than average scores on SPLITIDO, INFORMAL, and ALLONE. The Gainesville council comes closest

PAGE 333

31'4 to matching this pattern and differs from these two only in being low in SPLITGEN and high on SPLITIDO. The patterns of the remaining cities vary considerably. Four cities evidence only one high disagreement mode. Boca Raton and Melbourne both report high scores on only generalized splits; Cocoa identifies highly on idiosyncratic splits, and Bartow references SPLITPAR highly. Rockledge is low on all the disagreement modes. Four cities report high use of agreement modes. Cocoa, Ft. Walton Beach, and Melbourne suggest frequent use of the informal mode, and Cape Coral reports ALLONE as the only agreement mode in frequent use. Table 7^12 gives a crude picture of the use of the various modes by the councils in the high SES cities. It depicts the number of modes, by disagreement or agreement type, which each council references highly. Table 7-12 Number of Decision-Making Modes Referenced Highly By Each City Council in SES Group F Bartow Boca Raton Cape Coral Cocoa Cocoa Beach Et. Walton Beach Gainesville Hialeah Melbourne Rockledge Tallahassee Titusville Number

PAGE 334

315 processes assume. Only one city, Boca Raton, identifies the use of a single decision-making mode. The mean number of modes used by councils in group F is three. And in only three cases are the decision-making modes purely conflictual or consensual — Rockledge, Hialeah, and Boca Raton. Thus the presence of reversals in the direction of the associations is quite difficult to explain with reference to the generalized features of the councils. While the nature of the problems and the differences among the members succeed in accounting for some of the changes, the complexity of the adoptation of specific modes is considerable. In fact, such complexity is likely to elude the crude measures used in this study and involve features which might be captured only in detailed and extensive observation. Fortunately, most of the associations in the two groups remain in the same direction. And the statistical significance of the associations allows one to gain some insight into the nature of the decision-making modes within each group of cities. Anticipated logrolling typically occurs in those cities where generalized splits are cited as frequent reasons for council division over matters of policy. Idiosyncratic splits are not associated significantly with any of the agreement modes. This is not surprising unless idiosyncratic splits are associated with unipolar council types. In such councils, it might be expected idiosyncratic splits (SPLITIDO) would be linked to either the consensual (ALLONE) or informal (INFORMAL) modes. Florida cities display no such pattern when taken as a group. The consensual (ALLONE) mode is not significantly related to any of the disagreement modes, but the informal (INFORMAL) mode is linked to councils which report splits among members on particular is-

PAGE 335

316 sues. The greater the use of this mode, the less informal practices are used. If particularized splits seem to discourage informal decisionmaking practices, they are associated with attempts to work towards unanimity or amicable agreement. Among the high SES cities, the pattern of empirical relationships is different. Informal attempts to reach unanimity are least likely to be found in those councils which are typified by generalized splits. Instead the informal mode (INFORMAL) is associated with idiosyncratic council splits. It is as if in the high SES cities informal practices are not referenced unless the disagreement levels are quite low. Indeed, this interpretation is borne out by the direction, if not the significance, of the association of the informal mode with both the disagreement modes and the general measure of conflict. When members of the councils of group F suggest "there is simply no disagreement on many issues" — the members are all one — they are also likely to report the splits that divide the councils come on particular issues facing the council. This represents a reversal in the sign of the relation which obtained among the city councils as a whole. Council relations do not appear to lie behind particularized splits in the high SES cities. This seems to account for the association of this disagreement mode with amicable agreement among the several councils and with ALLONE in the group F cities. Because each city council does reference more than one decisionmaking mode, the pattern of association between the modes varies considerably between the cities as a whole and group F. Modes supplement one another in response to a complex set of conditions which obtain in each council. Each council must decide whether "to hold with the hound or run with the hare." It must trade off the possibilities of making de-

PAGE 336

317 cisions in one way to the degree that it frequently adopts other methods The modes are complex surrogates for accrued costs generated in repeated decision-making tasks. In spite of this complexity and the specific needs of each council, the decisionmaking modes of both groups of cities bear a consistent relation to the summary measure of conflict. Figure 7-6 displays levels of association between each decision-making mode and the measure of total council conflict (CONFLICT). For both groups of cities, the relationships between generalized splits and conflict (SPLITGEN and CONFLICT) , and between amicable agreement (CONSENT) and conflict are significant. Generalized splits are referenced in those councils whose members indicate the most pervasive presence of conflict. Amicable agreement, on the other hand, is negatively associated with generalized conflict. These associations are particularly strong among the councils of SES group F, the Kendall's tau-c being greater than 1 .44 in each case. Not only do the councils of the two groups reference fractious dissent and amicable agreement in a similar manner, the ordering of the association levels is almost identical. As displayed in Figure 7-6, the difference in the ordering turns on the relative levels of consensus (ALLONE) and particularized splits (SPLITPAR). These two modes both bear a negligible relationship to conflict. But if their levels of association are considered as weights by which to judge the amount of generalized conflict they tap, then their involvement on a conflict dimension is relatively similar in both groups of cities. Notice that for each group of cities amicable agreement (CONSENT) and generalized split (SPLITGEN) stand at opposite ends of the conflict dimension. For both groups of cities, the amount of conflict in which the council is engaged s

PAGE 337

318 w u H M a

PAGE 338

319 is a fairly good predictor of the types of decision-making modes a city will adopt. In particular, highly conflictual councils will have a tendency to establish generalized disagreement; highly consensual councils, on the other hand, display a propensity to work toward amicable agreement. Summary In this chapter I have attempted to trace the way in which council conflict is encouraged by the relations among the council members and their predispositions toward cooperation or conflict. In addition, I have suggested the amount of conflict in which a council is engaged may affect the manner in which a council faces policy issues. No council faces every issue which comes before it in exactly the same fashion. However, councils are likely to use a given number of modes to resolve issues. Most councils have a tendency to use just a few modes. It is easy to imagine the use of modes should not be independent of the issue which is confronting a council. If a council has established a mix of disagreement and agreement modes by which it resolves most of the issues coming before it, then it would be reasonable to suggest an issue which generates considerable costs for the members might well trigger the use of -a disagreement mode. While I have no data to suggest that this is the case — I cannot link the modes directly to the type of issue under consideration — the exchange perspective suggests this might well be the case. Each issue should actualize the politics of the council; part of these politics is the decisionmaking mode adopted to settle a specific issue. I will attempt to address a portion of the above proposition in the next chapter. Issues should be linked to conflict. Using Lowi's scheme and the objective set of values imputed from Adrian and Williams' scheme it may be possible to establish this link.

PAGE 339

320 Notes 1 Ian Budge, "Consensus Hypotheses and Conflicts of InterestAn Attempt at Theory Integration," British Jou rnal of Poli tical Science Vol. Ill (January, 1973), p. 73. ~* ~~~ _ f c 2 i o C1 ^ t0n J" Flnk ' " S ° me Conce Ptual Difficulties in the Theory 1968)?"* Tl J ° Urnal ° f C ° nflict R ^" 1 "I1^ I Vol. XII (September, *-iIn -, H t nna Pitkin ' The Concept of Representation (Berkeley: University of California, 1967). ' — ' y M Z ,u ^u ek Sobolewski > "Electors and Representatives: A Contribu10 ! ° ft The °^ °5 ^presentation," Representation . Vol. X, Nomos, eds. 95-107 n ° Ch and J ° hn W Cha P ma n (New York: Atherton, 1968)Tpp. ti, t 5 i J ° hn C " Wahlke > Heinz Eula "» William Buchanan and Leroy Ferguson gg-^|^ in Legislative Behav ior (New York: ' 6 See, for example, Ronald Matheny, "Role Conflict: A Quantita2^ I*,* ° f ' he Urban Le gislator," presented at Annual Meeting of the t?7ir an H . al u SClenC " Association (Chicago: August 29-September 2, vll.L n" 1 Vln " ent ,. HaUg6 ' Clty Council °^ ™ Representatives: Burkeans Versus Delegates, presented at Annual Meeting of the American Political Science Assocxation (Chicago: August 29-September 2, 1974). A , f l. See . Heinz Eulau and Kenneth Prewitt, Labyrinths of Democracy : . Adaptations, Lin kag es , Representation, and Policies in Ur ban Politics Undianapolis: Bobbs-Merrill, 1973), pp. 22-25"! " Vnl J r NelS ° n Polsb y' "Legislators," Handbook of Political Science, Nelson Polsby (Reading, Mass.: Addison-Wesley, 1975), p. 263. 9 Ibid., p. 291. 10 Ibid. 11 Eulau and Prewitt, op. cit ., pp. 444-462.

PAGE 340

321 13 Robert Lineberry and Edmund Fowler, "Reformism and Public Policies in American Cities," American Political Science Revie w, Vol. LXI (September, 1967), p. 701. 14 E.P. Hennock, Fit and Proper Persons: Ideal and Reality in Nineteenth-Century Urban Government (Montreal: McGill-Queen' s University, 1973), p. 331. 15 City Councils: Structures and Procedures (Washington, B.C.: Urban Data Service, 1971). 16 See Jerry Yeric, "City Managers, City Councilmen, and Public Budgets," prepared for delivery at the Annual Meeting of the Midwest Political Science Association (Chicago: April 29-May 1, 1976) or Hauge, op. cit . 17 City Councils ; Structures and Procedures , op. cit . 18 E.D. Pennock, "Political Representation: An Overview," Representation , Vol. X, Nomos , eds. J. Roland Pennoch and John W. Chapman (New York: Atherton, 1968), p. 14. 19 James Sundquist, Making Federalism Work: A Study of Program Coordination at the Community Level (Washington, D.C.: Brookings Institution, 1969), p. 247. 20 Robert Putnam, "Politicians and Politics: Themes in British and Italian Elite Political Culture," Diss., Yale, 1970, pp. 191-192. 21 For a discussion of this term, see Jurg Steiner, Amicable Agreement Versus Majority Rule , trans. Asger Braendgaard and Barbara Braendgaard (Chapel Hill: University of North Carolina, 1974), pp. 5-7. 22 Ibid . 23 Eulau and Prewitt, op. cit . , p. 215. 24 Ibid ., pp. 174-178. 25 Ibid ., p. 191. 26 Ibid ., p. 176. 27 Table 7-4 can be misleading itself. It was constructed using the mean value on each variable as the breaking point by which to gauge "high" or "low." Eulau and Prewitt construct their decision structures using two measures. The first and most important came from the responses to the following question: "When the council is in disagreement on an issue, would you say there is more or less the same line-up of votes here in (city)? I mean do some members seem to vote together on controversial issues? If yes: With whom do you usually vote on controversial matters? Now, what about the others? Are they united or split? If split: Who would you say votes most often together when the others are split?" (see

PAGE 341

322 Eulau and Prewitt, op. cit ., p. 173). From these responses they generated sociometric measures of the disagreement in each council on controversial issues. These sociometric findings were then corroborated by a question which asked each council member: "How do you account for the fact that the council divides as it does?" (see Eulau and Prewitt, op_^ cit., p. 175). It was the response codes to this open-ended question which I used to generate the policy disagreement scales (see Appendix) . Thus when I resort to means to gauge overlap I am really oversimplifying the presentation of "decision structures." Decision structures should not be referenced to the group of councils but rather to the absolute magnitude of each council's response on a given scale. The difficulty with this procedure with respect to the data on Florida city councils^stems from the nature of the data itself. First, I have no "objective measure of decisional structures, nor do I have a direct corroborating technique. More importantly, however, the scales which I have constructed to measure policy disagreements can be normalized to the same domain only with the possibility of considerable distortion. And it is such a normalization which facilitates comparison of absolute magnitudes. This distortion is a result of the aggregation which each scale assumed. Different numbers of items are involved in each scale. SPLITGEN involves four items; SPLITPAR only one. Since the response codes on |; heS ;? "ems were dichotomized, this severely restricts the range responses ror bFLliPAR. It increases the possibility that a bias exists for councils to be considered high on SPLITPAR. It is a bias which is moderated by the aggregation by council, but nevertheless it is bias which must be considered. As will be seen later, a similar bias exists for CONSENT. Despite these difficulties, it is possible to compare councils on a normalized schedule of responses. If SPLITPAR is considered then most Florida councils would be labeled Unipolar to use Eulau and Prewitt' s classification scheme. However, the overlap with other categorizations is still more prevalent among the Florida city councils than among those in the San Francisco Bay Region. Both SPLITGEN and SPLITIDO often range high m the same council where SPLITPAR dominates. This phenomenon lead me to re-examine the nature of "decision structure," and to question the manner m which Eulau and Prewitt made their assignments to various decision structures. Eulau and Prewitt reference their sociometric question to controversial issues which confront the council. Thus their decisional structures are really very specific types of decisional structures. They relate only to the most controversial issues appearing before the council. Ihe question the authors ask ultimately becomes, do the normal relations of the council affect the very special circumstances of a controversial issue decision. Moreover they ask this question explicitly assuming the nature of the issue will have little impact upon the decision itself These two assumptions, taken together, appear extremely unreasonable to me. 28 Eulau and Prewitt, op. cit ., p. 177 29 Kenneth R. Greene, "Dimensions of Issue Voting in a Local Legislature," prepared for delivery at Annual Meeting of American Political bcience Association (Chicago: August 29-September 2, 1974).

PAGE 342

CHAPTER EIGHT DECISION-MAKING MODES AND POLICY The prerogatives of local councils are many. Councils are charged with enacting ordinances, adopting budgets, and regulating building, housing and zoning. Yet the treatment given them in the literature of political science has largely overlooked their role in the determination of municipal policy. To counter this neglect, it was supposed councils are crucibles in which local policies are fired, if not molded. Indeed, the analytical concern for decision-making modes grew out of this supposition. The internal organization of decision, it was thought, should be affected by the policy choice under consideration. In this chapter, I will first argue that local councils do have policy choices to make and that these choices can bear on the way in which policy decisions are organized. Council relations alone will not determine the mode used. Next I will investigate the relationship between decision modes and policy. Following the notions developed in the previous chapter, namely that different modes are associated with councils' situations marked by varying amounts of conflict, the manner in which specific modes are linked to conflict over issues will be examined. It is possible decision modes are associated with issues which themselves generate different amounts of disagreement among the members. This means that certain portions of a council's decision repertoire may be accessed to manage different types of controversy. As in the previous chapter, I am concerned with council conflict as 323

PAGE 343

324 it manifests itself in the disagreements among the members. However, in this chapter I will detail the amount of disagreement among members over a number of general policy areas. In addition, I will examine the importance to members of policy areas in which disagreement is said to exist. Finally, I will treat the relationships depicted in Figure 8-1 . Among the councils of the high SES cities, group F, I will investigate the relationship between the various decision-making modes and the disagreement over a range of policy areas. [DECISION ORGANIZATION] TOTAL ^COUNCIL ISSUE CONFLICT^ 1 ISSUE CONFLICT CLUSTER [POLICY ISSUE CONFLICT] Figure 8-1. Major Categories and Variables Treated in Chapter Eight

PAGE 344

325 The Importance of Issues Typically, city councils have but a few members. Eulau and Prewitt argue this fundamental observation conditions all further commentary on the political realities of local legislation. 1 As a small group, a council must develop a "means of coordinating, integrating, regulating, and •J directing the work of its members. 1 "So it is then that Eulau and Prewitt come to accept the notion, given wide currency in small group research, that councils may be profitably considered as systems of social interaction. Indeed, the great worth of their investigations issues from just such a consideration. Unfortunately, the investiture of this proposition is not entirely estimable. In particular, it has ramifications for the manner in which the authors treat a city council's determinations of public policy. "Councilman," Eulau and Prewitt recognize, "are not participants in a small-group laboratory experiment that purposefully isolated the group from its external environment." 3 Nonetheless their exposition relies heavily on concepts employed in small-group research, concepts whose theoretical foundations are laid in the interaction perspective. 4 Their approach is particularly well suited to grapple with problems relating to communication, interaction, role differentiation, leadership, and coalition formation. But perhaps because this perspective attaches such importance to social organization and structured networks of statuses and expectations, it neglects a refined consideration of the way in which the . tasks of the group impact upon manner in which the group handles those tasks. This neglect is not immediately apparent. Consider the following passage:

PAGE 345

326 The overwhelming reality facing any legislative group is that, sooner or later, it must make decisions which, for all practical purposes, involve a choice between only two alternatives. It can only say "yea" or "nay" to a proposition before it. This very simple circumstance has a far-reaching impact. While the council is characterized by a variety of other properties and processes, its most distinctive social process is voting. Clearly, Eulau and Prewitt recognize the voting act as the central task of the council. But, their differentiation of the voting act is circumscribed by the perspective they adopt in considering the council's tasks. "Voting," note Eulau and Prewitt, "is a repetitive process." Moreover, in local legislatures, it is a repetition geared to a narrow set of issues. They argue the "issues with which a city council deals are considerably less diverse than those facing state or federal legislative bodies or courts." Under such circumstances, Eulau and Prewitt suggest, the importance of face-to-face interactions increases; the nature of the task or the matters under consideration become less important. The importance of council interactions becomes pervasive. The council's most pervasive structural feature is its decision structure — the relatively stable pattern that emerges from the voting situation when councilmen take stands either with or against each other. Indeed it is the agreed-on decision rules in terms of which voting is conducted that make the council a collective unit of action. Eulau and Prewitt focus on the council's decision structures, and in doing so, they attend to the assumptions of interaction theory. But such attention serves to diminish the conceptual import of voting as a task, and helps divert concern from the specific aspects of the task — choosing among specific issue alternatives. The exchange theoretical perspective I have adopted takes the act of choice as its conceptual cornerstone. It does not reject the possi-

PAGE 346

327 bility that face-to-face interactions help mold a council's performance of its tasks. Rather it suggests a particular issue may condition the manner in which established patterns of interaction come to bear in a choice situation. It is this animus which stands behind the notion of decision modes. Decision modes represent a repertoire of behaviors employed by councils in addressing the issues which appear before them. The particular mix of modes adopted by a single council can be considered to result from the character of the relations among the council members and the predispositions of the members. Yet the decision mode which is used to resolve a specific issue is likely to be conditioned by the reactions of the council and the content of the issue. It is only when council relationships have become extremely ossified, as in the case of Boca Raton discussed in the last chapter, that the issue under consideration will have no substantive appeal. As seen in the last chapter, most city councils can be considered to evidence a broad repertoire of decision modes (see discussion of Group F councils). Yet this alone does not indicate that issues are important for council politics. Eulau and Prewitt were unable to ascribe a single structure to 29 of the 82 councils they investigated. And this realization did not force them into a detailed examination of the impact of policy issues on council politics. The exchange perspective is predicated on the notion that issues are important. It offers the opportunity to consider the documented importance of council interrelationships while emphasizing the manner in which choices before the council can impact on such relations. In highlighting the complex character of issue perception and recognition, the exchange perspective stands in contraposition to the argument of Eulau and Prewitt which features the constrained set of choices considered by local councils,

PAGE 347

328 If the exchange perspective is to be usefully brought to bear on the relation between council politics and policy, it must first be shown that assumptions with which Eulau and Prewitt underpin their argument do not necessarily hold. After demonstrating the weakness of their position, my own argument can proceed. Eulau and Prewitt maintain, first, that the diversity of issues facing local councils is less than that before state or federal legislative bodies, and second, that this lack of diversity encourages the formation of stable decision-making structures. Yet aside from the intuitive appeal to be found in the fact that state and federal bodies have purview over a more diverse array of matters, they offer no serious consideration of how diversity within a set of issues might be determined nor how the diversity of issues might influence members' attention to issues. The problems associated with this neglect can be easily illustrated. Consider Table 8-1 . It lists in summary fashion the responses of 173 Florida council members to a question asking each to name the most important matters of policy to come before the council in the last year (see Appendix A for questionnaire). As a first approximation, an examination of the members' responses can be used to annotate the concern over the diversity of issues appearing before Florida councils. Twenty-four areas were mentioned as quite imoportant by at least one member on three or more councils out of the 61 represented in the sample. Twenty-seven different areas were mentioned by at least three council members; 61 areas were mentioned in total. Do such numbers portend diversity or the lack thereof? Without a more precise understanding of just what Eulau and Prewitt mean by issue diversity, it is impossible to say. Nonetheless a basic perspective

PAGE 348

Table 8-1 Number of Council Members and Councils Citing Policy Issues as Important 329 Number of Number of Members Making Councils Issue Mention Citing Issue (N = 487)a ( N = 61 ) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. Planning and Area Growth Municipal Employees Public Utilities Financial Administration of the City Natural Resources Rate Increases Preservation of Open Spaces and Density Control Garbage and Sewer Annexation Municipal Procedures Crime and Police Funding Zoning Property Taxes Administrative Problems Urban Redevelopment and Housing Parks and Recreation Form of City Government Comprehensive Ordinance Reviews Municipal Auditoriums and Convention Facilities Fire and Public Safety Streets and Highways Revenue Sharing Money Disputes Downtown Development Projects Local-County Relations Education Mass Transit Bond Issues Bidding Procedures and Franchises 86

PAGE 349

330 on the sense of their argument can be purchased, if only in oblique fashion, using this same set of data. Perfect heterogeneity among issue responses would have yielded 487 areas. But such heterogeneity is contextually constrained by the co«n set of distances shared by .embers in each of the 61 councils. And hence examination of the range of issues appears to have some inherent limitations. There is another way of considering diversity of issue importance, however; that is, by positing the nature of perfect homogeneity within councils. Under such conditions, each council member would share a common evaluation of important issues with his colleagues. The greater the number of issue mentions per council, the less agreement over the importance of issues is. A quick glance at Table 8-1 suggests, on the average, no council has more than two members who cite the importance of the same issue despite the fact that each council as a whole faced the same set of issues. One interpretation of • this observation would suggest merely that members disagree over the relative importance of issues. But it may also be taken to indicate the diversity of important issues is such that agreement among members is unlikely, m either event, such speculation does suggest members attend differentially to issues, and perhaps, it can be argued, it is the attention of the members which merits conceptual import rather than the mere diversity of issues. Diversity, per se , raay Qot provide ^ ^ ^ ^^^.^ & ^ ber's orientation toward issues. But what does? The exchange theoretical view begins by considering the degree to which the subjective interest of the council members are involved in the issues. But it recognizes that issues, as matters in dispute or to be decided, are never considered in_vacuo. They percolate into the political arena with their

PAGE 350

331 backdrop already primed. For council members, the meaning of an issue is framed not only by its relation to a member's set of interests but by its interpretation within a set of well established rules. These rules reflect not only the relations among the members but the configuration of decisions on issues which has emerged through time. This argument follows the seminal formulation provided by Lowi: (1) The types of relationships to be found among people are determined by their expectations— what they hope to achieve or get from relating to others. (2) In politics, expectations are determined by governmental outputs or policies. (3) Therefore, a political relationship is determined by the type of policy at stake, so that for every type of policy, there is likely to be a distinctive type of political relationship. 9 There is a clear and persistent interdependency between politics and policies over time for members of the same council. It is an interdependency mirrored in Eulau's definition of policy as "a strictly theoretical construct that is inferred from the patterns of relevant choice behavior by political actors and the consequences of choice behavior." 10 It is important to recognize that there is a difference between issue choice and policy. The former helps inform the latter. But policy is inferred from patterns of choice; no single issue choice determines the character of policy. Both Eulau and Lowi underline this point. They posit regularities in choice behavior, and argue such regularities can be discovered. Of course, this means decision makers must be able to perceive and classify the meaning of issues in some consistent fashion. On this note, Lowi observes, "It is on the basis of established expectations and a history of earlier government decisions of the same type that single issues are fought out." 11 For Lowi, issues themselves are ephemeral; the specifics of each choice situation are likely to be different. None-

PAGE 351

332 theless they have attributes which trigger predictable responses on the part of decision makers. Each issue can be conceived as part of a larger bundle of current concerns and past decisions termed a policy. And the number of policy bundles is severely restricted. The argument follows Clausen: The proposition that a small number of policy concepts are widely shared is not nearly as unreal as it may" first appear It is through the use of policy concepts of high generabxlity that political discourse is possible. A "policy concept provides a summary of the detailed statement ot a policy proposal in shorthand terms understood by others. ^ There is a seductive simplicity to Clausen's proposition. But does it provide the key for understanding a member's orientation toward issues? In many ways it does. Certainly it is appropriate to suggest such policy bundles might couch the expectations alluded to by Lowi and serve to order choice behavior among council members. Such a conception squares _ neatly with the exchange perspective; it suggests a time-ordered perception and recognition of decision costs on the part of each member. It addresses the posited interdependency of politics and policy. Notice, however, this formulation makes severe conceptual and empirical demands upon the researcher. First, it is not at all clear that policy bundles provide the basis for meaningful political discourse in every choice situation. As the earlier discussion of decision modes demonstrated, the dynamics of council interaction may, on occasion, transcend the issues at hand and provide the most appropriate key to understanding a particular council. Within the exchange perspective, this realization heightens the importance of strategic interaction. The implication is that decisionmaking modes, as adaptations to interactions among members, condition the reaction to policy bundles in a pervasive manner. In fact, the modes

PAGE 352

as structural manifestations of mutually acknowledged expectations can be anticipated to be referenced in a consistent fashion for the resolution of choice situations involving issues within a specific policy bundle. But the point remains that a single mode may be referenced without regard for the specifics of a choice situation. Conceptually, this raises the following question: What is the relationship between policy bundles and decision modes? Or more specifically, how do issues relate to decision modes? One way of approaching the conceptual specification of this link begins with the effects of an issue as they impact upon the council circumstance. Of course, these effects are likely to be quite varied. However, the manifold aspects of a policy bundle or a specific issue can be simplified in brutal fashion to consider only the conflicts they engender. Working within the exchange perspective, this pares considerably the diverse nature of the costs accrued to each member in a choice situation. It is likely to understate the full range considerations confronting each member. Nonetheless, such a paring seems reasonable if, as argued, conflict provides an indicative shorthand for high levels of cost. Conflict offers a simplistic but useful way of linking the content of an issue or policy bundle to the political behaviors of the members. Issues can be expected to generate or foster conflict differentially. Certain types of issues, certain policy bundles should be associated with disagreement among the members more than others. However, if the animus standing behind the notion of decision modes is on target, then it must be expected the repertoire of modes used within a council will bear a consistent relation to the conflict associated with the issues. A related phenomenon was illustrated in the previous chapter. The modes were shown ' 'VJWj^liJJ* ' '^ ' " '^tf*^*^^*^*'! » i! ^> «t ^JW i ),»T );»il ^-4ia B l »«^»^WJ^

PAGE 353

334 to have an ordered relation to the total amount of conflict in which the council was involved. It might he expected the various decision-making modes will bear a similar relation to the amount of council conflict over all issues. In addition, it can be expected decision modes will be linked to issues which themselves produce different amounts of conflict. If the discussion surrounding total council conflict presented in the previous chapter is a guide, then this relationship is likely to be quite complex. There appears to be no one simple relation between components of the decision repertoire and the types of policies they are used to manage. Yet conflict among the members may be used to indicate, however indirectly, the link between modes and policy types in the fashion described in chapter two. The conceptual simplification that allows a substitution of conflict for costs is not entirely estimable. One particular problem involves the use of conflict as a means of distinguishing between the various decisionmaking modes. As was seen in the last chapter, conflict is most useful in differentiating generalized splits from all other modes. This differentiation is less complete among the consensual modes. While conflict over issues is likely to narrow the conceptual dimensions of the relationship between conflict and modes, explicit hypotheses on the form of the relationship are tenuous. In councils where there is a great deal of conflict over policy, the generalized and particularized split modes might be supposed to exist. Extremely low conflict over policy should be associated with the more interaction dominated modes— the informal (INFORMAL) and consensual (ALLONE) modes. The remaining modes should bear no relation to conflict over the issues. Because the link between decision-making modes and policy is so dif-

PAGE 354

335 ficult to draw precisely, I wi H resort to an argument with several d±f _ ferent steps. Each step is related to aspects of the connection between the two as I have discussed the*. Taken as a whole the argument should be suggestive of the nature of the relationship. In particular, I will examine : 1the amount of disagreement over a number of gene-al policy areas; 2. the importance to members of policy areas in which disagreement is perceived; 3. the associational link between conflict over a range of policy areas and the decision-making used. This argument is directed toward establishing the relationship between the total amount of policy conflict in a council and the modes which are used. It does not and cannot make the specific connection between a specific mode and policy types. Rather it bears on the proposition that there is likely to be a systematic link between the conflict which policies generate and the political efforts used to resolve them. It establishes the minimal conditions for the possibility that issues might condition decision-making modes-the importance of a policy and disagreement over its resolution. However, it should be remembered both the policies and the political efforts of each council have an historical aspect which I can capture only indirectly. Hence my argument is necessarily circumspect. Conflict, Issues, and Decision Modes The proceedings of a typical council move slowly; deliberations are seldom punctuated by sustained debate after the fashion of national legislatures. Rather there is a welter of the mundane, the technical, the everyday matters associated with the governance of a city, however small. S^lWMfw

PAGE 355

336 But the tempo and tenor of council meetings may act to disguise the existence of significant disagreements among the members over particular issues or matters of policy. Indeed the frequent reference to the particularized disagreement decision-making mode by the Florida councils is testimony to this proposition. How is it then that councils can be conceived to be so harmonious? No doubt, the illusion of harmony is fashioned, in part, by the considerable strain on each member to quiet public dissension for reasons already discussed in chapter two. Then too, there is the typically vulnerable position of the council among the many forces and personalities dominating local politics. But it is also evident that issues often surface in the council arena in a manner dissimilar to that of other legislatures. In city manager forms of government, the council is presented with detailed proposals by the manager; in mayor dominated forms, the executive office may submit proposals. In either event, however, the council itself will work through much of the lawmaking process. Typically even the existence of subcommittees will not displace the consideration of minute detail by the entire council. At times, the council will engage in the formulation, instigation, publication, interest-aggregation, mobilization, and modification of a particular piece of legislation." Thus much of a council's time and energy is spent on matters of considerable specificity. Such deliberation may frame the issue at hand only poorly. Oftentimes it is possible to forget altogether what the thrust of the proceedings is; much of the council's efforts is spent "inside" a particular issue or matter of policy. 14 Because council members are involved in an extended decision-making process of many stages, the workings of any one council meeting are quite

PAGE 356

537 deceptive. The form of the proceedings belies concern for general issues and public policy. Each issue or policy may be accompanied by a series of decisions. A focus emphasizing the detail of each meeting may lose the perspective on the importance of such meetings. Theodore Lowi has noted perceptively, "When one assumes that policy making is policy making is decision making, and therefore does not enter into a priori analysis of the character of the choices being made, one almost inevitably becomes incrementalist and manipulative." 15 Hence, if the researcher is to investigate policy matters, his analysis must transcend the mere form of council decisions. Somehow it must examine issues and policy matters. Researchers investigating legislative policy at the national or state level most frequently turn to member votes as indicative of the relationship between policy and politics. And while exclusive concern with analysis of congressional roll calls has been roundly criticized, the intention behind the analysis is well founded. 16 Voting on the floor of national or regional legislatures provides the opportunity for many members to have a voice on particular matters of policy. In many instances, important preliminary decisions on legislation are made in specialized policy committees. The roll call provides each member with the opportunity to publicly register his opinion on a final draft of the legislation, regardless of his role in its formulation. Precisely because city council members tend to be involved in the extended decision-making process, a singular concern for specific votes within the council would likely be misleading and liable to understate the member's involvement with and concern for issues. Councils are not national legislatures; the familiar landmarks of analysis in the workings of larger legislatures cannot be transferred uncritically to the

PAGE 357

338 local level. Thus, rather than attending to the manner in which council members have voted on council roll calls in order to determine their feelings about issues or to gauge the amount of disagreement over issues among the members, I asked council members about issues and policies as they, the council members, viewed them. 1 attempted to engage a member's thoughts on the entire decision-making process with respect to general policy or funding areas by envoking well-known labels for each such area. In particular, I wished to know in which areas, generally speaking, the council members were most divided among themselves. 17 Each council member was asked to evaluate 23 general policy funding areas. These areas represent expenditure reporting categories used by councils throughout the state of Florida. 18 They also outline a diverse range of policy matters likely to be considered in roost Florida cities. 19 Table 8-2 offers a glimpse of the amount of disagreement council members perceive among themselves in a series of general policy areas. The typical Florida council member perceives considerable division in about seven of the general areas. Two areas of council deliberation were named as among the most divisive of areas by over half the members. In fact, a substantial majority of all members indicate matters of zoning and of planning and area growth as areas of considerable disagreement. Five other areas are named by at least 30 percent of all council members. Only one area— hospital (construction and operation)— is seen as divisive by as few as 10 percent of the members. It is difficult to know in just how many areas or with what frequency, the perception of council disagreement might be taken as definitive grounds for accepting the notion that issues do indeed coalesce different types of politics. Certainly any criterion would be arbitrary. Nonetheless,

PAGE 358

339 Table 8-2 Policy or Funding Area Ranked By Proportion of Members Indicating Council Disagreement Rank General Policy or Funding Area Proportion 1 Planning and Area Growth .707 2 Zoning .620 3 Municipal Employees . 396 4 Form of City Government .384 5 Natural Resources .348 6 Financial Administration of the City .335 20 21.5 Public Health Care 300 293 7 Parks and Recreation 8 Public Utilities 9 Garbage and Sewage .256 10 Industrial and Economic Development .244 11.5 Public Welfare .226 11.5 Urban Redevelopment and Housing .226 13.5 Crime and Police Funding .220 13.5 Streets and Highways .220 15 Property Taxes .196 16 Public Safety .183 17.5 Airports and Water Terminals .128 17.5 Municipal Auditoriums .128 19 Racial and Ethnic Relations .122 Courts and Their Operation .116 21.5 Mass Transit or Bus Systems .110 110 23 Hospitals .061

PAGE 359

340 the results reported in Table 8-2 suggest members suppose there is substantial disagreement over policies in several areas. Of course, it is not especially surprising that zoning matters would be seen as so divisive. In most council's, zoning variances occupy a significant proportion of each meeting's time, and such variances are likely to involve rather specific interests in the community. However, among Florida council members, zoning is less important as a divisive force than deliberations involving planning and area growth. This finding re-enforces the documented concern over such issues known to exist among the voters of Florida in 1974. 20 But more important than either of these findings is the evidence which demonstrates the presence of additional areas of considerable disagreement. Fully 25 percent of all members indicate there is substantial council division in more than half the general policy areas in question. Of course, the results reported in Table 8^2 reflect only the presence of council disagreement in a significant number of areas. The numbers cannot be used to infer members believe disagreement occurs in matters of importance. In fact, I have no data which addresses this question directly. However, the open-ended question used to gather the figures presented in Table 8-1 does bear on the proposition. It asked each member to name the most important matters of policy to come before the council m the last year. These responses can be compared to those in Table 8-2. Presumably, if disagreement occurs over matters of some import, then those same matters should be mentioned frequently in the open-ended format. Table 8-3 presents a comparison of the two sets of rankings. The correspondence between the two sets of figures is relatively close in the upper ranks. Planning and area growth leads both lists;

PAGE 360

341 Table 8-3 A Comparison of Policy Issues According to Their Number of Ment And Their Propensity to Generate Council Disagreement ions Rank in

PAGE 361

342 municipal employees, natural resources, the financial administration of the city, public utilities, and matters of garbage and sewage rank high on both measures. Of course, there are a few anomalies. Members mention matters of zoning as important with considerably less frequency than they cite it as a matter of disagreement. Similarly, questions concerning the form of city government or parks and recreation do not appear as important to members as they are divisive. Nonetheless, members typically appear to attribute importance to those general policies which are also most divisive. Generally speaking, it can be concluded councils do disagree over a broad range of issues and these issues have some importance for the members. But, these are minimal conditions for asserting there is a link between matters of policy and the existence of different decision-making modes. It remains to be seen whether there is an ordered associational relationship between conflict over the issues and the various decisionmaking modes. Such an analysis can proceed on two different levels. Throughout much of my examination of councils, I have first illustrated relationships as they exist among all the Florida councils for which I have information. Such relationships were always somewhat suspect given the real possibility for bias issuing from an under-reporting in a number of councils. Nonetheless, the relationships were reported because there was little reason to doubt they were not reflecting true relationships. In the case of decision-making modes and their connection with issue conflict, the situation is quite different. The connection shown between conflict and decision modes in the previous chapter suggests conflict may not differentiate the modes equally well. Thus it is probably best to restrict the analysis to those councils in the high SES cities— group F.

PAGE 362

34 3 Before presenting the analysis, the amount of disagreement over issues in the group F councils can be placed in perspective by examining the average number of general policy or funding areas in which disagreements have been noted by members in each of the councils. Table 8-4 reports these results for each city listed by SES grouping. Notice that the group F cities rank behind the cities of group D in the number of disagreement areas. This ranking differs from that reported in the previous chapter on the summary measure of conflict. The high SES cities were shown to be the most conflictual group when relations with the city manager were also assessed. Of course, the high issue conflict of group D is due, in part, to the large amount of disagreement registered in the Miami Beach council. Thus, it might be supposed Miami Beach is an anomaly. In the event that Miami Beach were eliminated from consideration, the high SES cities would again rate as the most conflictual. Notwithstanding t such conjecture, it is illuminating to mark the changes in council rank on the two measures. For example, the city councils of group E appear much less conflictual in comparison to other groups when only disagreements over general policy are considered. Similarly, the councils in group A evidence less conflict with respect to other councils when only general policy disagreement is measured. For the group D cities,' on the other hand, disagreement over policies gives them the appearance of being relatively more divisive than did their rank on the summary measure of conflict. In short, to the degree relative rankings on the two indicators serve as an adequate judge, it appears the summary measure used in the previous chapter is indicative of phenomena different than mere issue disagreement. Indeed this view is collaborated by the lack of association between the two measures (Kendall's tau-c = -.056). Thus it can be ex-

PAGE 363

344 Council Name Table 8-4 Average Number of Issue Disagreements Per Member For Each Council in SES Groups Issue Disagreements (Average/Member) GROUP A Bradenton Clearwater Deland Dunedin Gulfport Leesburg New Smyrna Beach Pembroke Pines Sarasota Tamarac Vero Beach COUNCIL AVERAGE GROUP B Belle Glade Ft. Pierce Homestead Plant CitySan ford Winter Haven COUNCIL AVERAGE 3.50 4

PAGE 364

345 pected that an associational analysis examining the relationship between decision modes and issue conflict will bear a slightly different form from that between overall conflict and the modes. Fortunately however, the form of this relationship should not be perturbed greatly merely because the focus of analysis has been narrowed to the cities in group F. That is, the high SES cities of group F register highly on both indicators of conflict. Only three of the decision-making modes associate significantly with the average disagreement per member taken over all general policy areas. Figure 8-2 arrays this trio graphically along a single issue disagreement dimension. As was hypothesized earlier, the decision-making modes most marked by a reliance on interpersonal interaction for their resolution of choice situations are negatively associated with conflict over the issues. The informal (INFORMAL) and consensual (ALLONE) modes are associated negatively and significantly with policy conflict. In councils where the number of general policy disagreements per member is high, it is likely the generalized split decision-making mode is referenced frequently. At least, that is the sense conveyed by the positive and statistically significant association between policy conflict and this mode (SPLITGEN) . But the other hypothesized positive relation did not materialize. While the particularized split mode (SPLITPAR) was associated positively with issue conflict, the association was not significant. This same result typified the remaining relationships. Neither logrolling (AVOIDCON) nor amicable agreement (CONSENT) bore significant relation with issue conflict, and, as supposed, idiosyncratic splits were not related. These findings appear to sustain the view that some decision-making modes are more permeable to issues than others. In particular, the modes

PAGE 365

346 l.O-i 6 .42_ SPLITGEN* 0.0 -.2 _ 4_ -.6_ -.8_ -1.0 J Council Division on All Issues SPL1TPAR CONSENT SPLITIDO AVOIDCON -ALLONE* INFORMAL^'Significant at .05 Figure 8-2. Correlation of Decision-Making Modes With Average Disagreement Over Issues in Group F (Kendall's Tau-C)

PAGE 366

34 7 dominated by personal relations appear less susceptible to use in a controversial choice situation. This notion is buttressed by an examination of the way in which council characteristics relate to issue conflict. Of the several characteristics, only one is related significantly to issue conflict. Council intimacy is very strongly and negatively related to issue conflict (Kendall's tau-c = -.704). Where council members are not on such friendly terms, issues are more likely to make a difference. Generalized splits are related to issue disagreement, and this relation exists even when council relations are controlled. For example, it might be supposed that policy disagreement would prevail where councils are polarized. To an extent, this is true, but the association is only moderately positive and not significant. However, when the council's polarization is controlled for, the relation between generalized split and issue disagreement remains strong (partial Kendall's tau-c = .562). For those decision modes not strongly linked to policy disagreement, the impact of issues is less certain. An examination of the many partial association coefficients reveals no severe shifts in relationships. In any case, the effects of issues on such modes are difficult to gauge. However, the manner in which overall conflict concerning policy engages differentially the several modes suggests yet another problem. Namely, the gross manner in which issue disagreement has been measured may distort any understanding of the relation between modes and issues. Thus before concluding the specification between decision modes and issues has been amply drawn, it is first necessary to disaggregate the issue conflict measure. Issue Conflict Clusters and Decision-Making Modes In order to decide how decision modes are linked to issues which are fwwwwww

PAGE 367

348 Table 8-5 Clusters of Council Issue Disagreements Cluster 1: Diverse Matter s Financial Administration of the City Industrial and Economic Development Public Welfare Crime and Police Funding Public Safety Airports and Water Terminals Courts and Their Operation Mass Transit and Bus Systems AVERAGE CORRELATION WITH ALL ITEMS IN CLUSTER = .230 Clus ter 2: Internal City Matters Municipal Employees Form of City Government Parks and Recreation AVERAGE CORRELATION WITH ALL ITEMS IN CLUSTER = .327 Cluster 3 : Development Matters Natural Resources Urban Redevelopment and Housing Public Health Care AVERAGE CORRELATION WITH ALL ITEMS IN CLUSTER = .273 Cluster 4: Growth Matters Planning and Area Growth Zoning Practices and Procedures AVERAGE CORRELATION WITH ALL ITEMS IN CLUSTER = . 262 Cluster 5: Prominent Services Public Utilities Garbage and Sewage Property Taxes AVERAGE CORRELATION WITH ALL ITEMS IN CLUSTER = .245 Cluster 6: Main tenance Concerns Streets and Highways Municipal Auditoriums AVERAGE CORRELATION WITH ALL ITEMS IN CLUSTER = .061 Unclustered Items Racial and Ethnic Relations Hospitals

PAGE 368

349 associated with different amounts of conflict or disagreement, it is first necessary to deal with the manner in which issues are to be treated. The natural inclination, I think, would be to consider each general policy area individually. This would portray each relationship most intimately, and would act to specify the fashion in which the various modes are permeable to certain issues. Unfortunately this type of analysis would severely strain the possibilities of the data base since over the range of all issues it is only known whether a member perceived disagreement or not in each issue area. Moreover, such an analysis would be needlessly complex since disagreements in several of the areas are likely to be colinear. Thus I have elected to reduce the number of issue disagreement areas. Using a variant of hierarchical cluster analysis, I produced a spatial representation of issue disagreement which grouped the 23 areas into seven clusters (see Appendix E) . These clusters are portrayed in Table 8-5. With the exception of two items, all 23 items were clustered into six groups. Racial and ethnic relations, and hospitals formed the seventh group, unclustered by the CLUSTER program. The tightest grouped cluster is formed by items concerning city employees, parks and recreation, and the form of city government. The least well-formed cluster is composed of streets and highways, and concerns over municipal auditoriums. The reason for clustering the general policy disagreement areas flows from a concern for the manner in which policies, associated with differing amounts of conflict, are dealt with by the city councils. Each cluster contains areas which when taken over the evaluations of all members involve different amounts of conflict. Since the clusters represent different patterns of conflict, I believe there is no particular basis on

PAGE 369

350 "hich to name the dusters. Nonetheless, after the fashion employed in factor analysis, I have attempted to identify the clusters in terms of the general policy areas they included. The descriptors with which I „m refer to each cluster are italicized in the paragraph below. As illustrated in Table 8-6, Cluster 4 is the tost conflictoal of the seven groups, it contains the two areas ever which council meters indicate councils disagree most: pl a „„i ng and area ^^ and ^^ Florida councils appear to disagree most over matters of ^rosth; Cluster 4 reflSCtS tMS M " C """» iSternal^it^jsttera of Cluster 2 ranh second, but far behind the disagreement noted over the growth matters of Cluster 4. Cluster 5, which includes ff^iaest^ervices provided by the city together with property ta X es, ran.s third in average conflict a„ong the several clusters. The general policies in cluster 3 are indicative of concern over the develop^ of the comity. tod „ hlle the average olsagreement of the cluster puts it s,„arely in the middle of the ordered dusters, the difference in the average disagreement over the policies included in the cluster evidences the largest standard deviation of any of the clusters. This is due to the presence of the public neaUh care area "hich is often not dealt with by the councils in several cities. Concern over health care is often left to county officials, cluster 1 contains the largest number of policy areas of any cluster; it represents a diverse group of concerns typified mostk„ i-i,., j yp most by the moderate amounts of disagreement seen in each of the general policy areas. The dlvet^coscerns of Cluster 1 ^nifest n ore disagreement than Cluster 6. Cluster 6 includes streets and highways as well as municipal auditoriums, for convenience I will refer to this group as the sai^en^c^ster. The usrelated^tems placed in group seven also happen to form the least conflictoal cluster. Of

PAGE 370

351 course, this is due to the extremely low amount of conflict over hospitals registered by most council members. As unrelated items, however, the empirical worth of this cluster is most problematic. Nonetheless, I will use the cluster in further analysis because it contains so few items and thus may be involved in relations which are interpretable. Table 8-6 Average Disagreement in Each Cluster Per Member Per Issue For All Members of SES Group F CLUSTER Cluster 1: Diverse Matters Cluster 2: Internal City Matter AVERAGE .195 360 STANDARD DEVIATION 077 RANK ,053 Cluster 3: Development Matters Cluster 4: Growth Matters Cluster 5: Prominent Services Cluster 6: Maintenance Concerns Unclustered Items, 7 .228 .659 .248 .174 .092 .119 .069 .048 .065 043 6 7 Just as each council manifests different amounts of disagreement over all the general policy areas under consideration (see Table 8-2), so too each council registers slightly different amounts of disagreement in each of the clusters. Table 8-7 depicts the proportions of disagreement among all members in each council in SES group F over the issues included in each cluster. Notice that despite the general importance of Cluster 4

PAGE 371

352 Table 8-7 Average Proportion of Disagreement Per Issue Cluster Per Member Per Issue For Group F Councils City Cluster Number 3 4 5 Bartow Boca Raton Cocoa Cocoa Beach Cape Coral Ft. Walton Beach Gainesville Hialeah Melbourne Rockledge Tallahassee Titusville .125 .143 .086 .887 .144 .020 .457 .144 .115 .119 .000 .141 •250 .083 .625 .083 •917 .667 1.000 .500 133 .133 .800 .267 •780 .110 .520 .557 .390 .167 .835 .167 .190 .287 .640 .240 •733 .467 .700 .400 •277 .330 .920 .057 .600 .330 1.000 .267 •277 .333 .830 .057 083 .333 .875 .333 111 .333 1.000 .553 .083 .000 .000 .330 .440 .000 .300 .165 .000 .085 .125 000 .000 .125 .000 .330 .000 .000 .200 .170 .000 .000 .000 .170

PAGE 372

35 3 or growth issues, two councils appear to have more conflict in other areas. Cocoa Beach and Gainesville are the exceptions to the general tendency. The pattern of disagrees on other clusters is less uniform. Four councils report their second highest disagrees levels on development matters (Cluster 3), W on internal city matters (Cluster 2), and three on prominent services. Indeed the pattern of disagreement between councils with relatively similar levels of total conflict is quite different. For example, Gainesville and Boca Raton both report high amounts of total conflict, but the distribution of conflict across the clusters is quite dissimilar. Even between contiguous cities such as Cocoa and Rockledge the councils report disagreement over issues in different fashion. In short, the measure of conflict on all issues may act to disguise the manner in which policy conflict relates to the decision modes. In order to better understand the relationship between such conflict and the use _ of certain decision modes, the disaggregated conflict must be examined. The disagreement clusters provide one means of getting a better look at the relationship. In the previous chapter, it was shown that the councils of the high SES cities referenced the various decision-making modes differentially (see Table 7-12). Boca Raton can be considered to employ one mode, while Cocoa Beach references five; and yet both councils rate highly on the total i SSU e disagreement index. Conversely, cities rating low on the total disagreement index such as Tallahassee and Cocoa employ different numbers of modes; four and two, respectively . Apparently, there is no simple relation between the number of modes used by a council and the total amount of issue disagreement. But there is no evidence to indicate that consensual modes are favored over conflictual modes when the total level of dis-

PAGE 373

354 Table 8-8 Kendall's Tau-c Between Decision-Making Modes and Council Divisions On Clusters For Group F Cities >uncil Decision-Making Mode ster 1

PAGE 374

355 agreement is high. At this level of analysis the findings of the previous section stand. The generalized split mode is most likely to be referenced in councils with high issue conflict; consensual modes dominated by a reliance on interaction occur most frequently in low conflict councils. However, the disaggregation of council conflict into clusters of policy areas allows for a refined examination of these findings. In particular, using the known conflict ranking of each of the clusters, the relation between an issue cluster and a decision mode can shed some light on the manner in which the modes are permeated by issues. Further, the relation between a single cluster of issues and the several decision modes can be drawn. Table 8^ arrays the associations between each of the policy disagreement clusters and the several decision-making modes. The results are often quite different from those reported earlier with respect to total issue conflict (see Figure 8-2). Notice that the generalized split mode manifests the most significant association across the issue clusters. With the exception of internal city matters (Cluster 2) and maintenance concerns (Cluster 6), each of the clusters is associated quite strongly with this mode. The greater the disagreement over such matters as growth, the greater the chance the generalized split mode will be referenced. This re-enforces the expectation generated by an examination of the total issue conflict measure, but refines it somewhat. In particular, it dashes the notion that the mere presence of great disagreement over a series of issues acts to increase the possibility of the reference of the generalized split mode. At least among the high SES cities of group F, the referencing of modes appears to be discriminating and complex. Several councils rate internal city matters as among the very highest disagree-

PAGE 375

356 ment areas, and yet no one mode is associated strongly and positively with its use. Table 8-8 suggests great disagreement over such matters only appears to discourage the consensual mode. As complex as the policy-mode interaction appears, it does seem as though some patterns may be extracted from the coefficients presented in Table 8-8_^ First, the idiosyncratic split mode bears no significant relation to any of the council division clusters. This is to be expected if the mode does signal disagreement that has no patterned aspect. It is a bit more surprising that there is no relation between the particularized split mode and any of the council division clusters. However, one explanation of this finding comes from an examination of Table 7-8 in the previous chapter. The particularized split decision mode shows very little variance among the councils of cities in group F. In fact, five cities share the same score on the index of particularized splits. While such a finding may adequately reflect the situation in the high SES cities, the presence of so many ties in the ranking of the decision mode acts to decrease the reliability of the associational statistic. A further inspection of the figures in Table 8-8 reveals the four consensual modes are involved with the disagreement clusters in different ways. However, each of the statistically significant relationships in which these modes are involved is negative. The manifestation of considerable conflict in specific clusters is associated with the diminished tendency for certain modes to be referenced. For example, the development concerns included in Cluster 3 appears to discourage the use of the amicable agreement and logrolling modes. This set of policies is the only one in which the two bargaining modes are discouraged. That is, for most disagreement situations, these two decision-making modes are available and

PAGE 376

35 7 used by various councils. However, no consistent pattern emerges in which the bargaining modes are likely to be referenced for high disagreement areas. In sum, the pattern for referencing these two modes is likely to be quite complex, and may not easily be specified using a single measure of issue costs. It is the consensual modes that rely most heavily on personal interaction which are least likely to be referenced in high conflict situations. Even here the relationship is not entirely straight-forward. Consider the maintenance concerns of Cluster 6. While the association coefficients between this cluster and the informal and consensus modes are negative, in neither case are they statistically significant. Apparently, in the high SES cities of group F, even the presence of considerable disagreement over matters involving local streets and highways will not cause the rejection of interaction based decision modes. It may well be certain policy areas encourage the use of such decision modes because of the benefits each of the council members can derive from allocating the policy throughout the community. Members can disagreee over the expenditures for street repair, but this disagreement need not fracture council relations because street repairs can be apportioned throughout the community regardless of the total expenditure. On the other hand, substantial disagreement over some matters of policy can promote situations in which neither the consensus or informal mode will be referenced. Among the cities of group F, the diverse concerns in council disagreement cluster appear to be of this nature. Similarly, when council disagreement becomes intense over matters of internal city affairs, it is unlikely a council will adopt the consensus decision-making mode. Disagreement over policies such as local treatment and hiring of municipal employees appears to discourage the access

PAGE 377

358 to a n,ode which is re-enforced by council intimacy and which connotes substantial philosophical harmony. m like fashion, when a council divides frequently over racial and ethnic relations (element in Cluster 7), it is unlikely to reference the consensus mode. The difficulty of ferreting the relationships between modes and issues is great. A portion of the problem, no doubt, lies in the use of conflict as a means of simplifying the costs to each member associated with a policy or issue. That is, members may see in the issues included in a policy area more than the conflict which results. Individual expectations and a host of other considerations may displace or divert the conflict. Additionally, the clusters themselves may disguise the manner in which members react to issues. The council disagreements manifested in each of the clusters are not exclusive of one another. Nor are they totally different from the measure of total issue conflict. But the pattern of variation reveals no simple structure (see Table 8-9). clusters 3 and 4 are barely associated with the total issue conflict measure. Moreover, these two clusters bear consistent relation only with one another. On the other hand, Cluster 5 is strongly associated with total council on issues and related to only one other cluster. Such figures make it difficult to argue an association between a decision mode and a cluster can be used to get a bearing on the manner in which the intensity of the conflict in a cluster helps establish its relation to a mode. Among the cities of group F, Cluster 4 evidences the highest amount of disagreement. But this very aspect when taken across all councils in a small sample makes it a poor tool for discriminating the penetration of a mode by issues. It is impossible to argue because the issue cluster with the highest disagreement levels is associ-

PAGE 378

359 Table 8-9 Kendall's Tau-c Between Council Division Clusters And Total Amount of Council Conflict on Issues For Councils of SES Group F Council Division on Issue Cluster Cluster Cluster Cluster Cluster Cluster Cluster 1 2 3 4 5 6 7 Total Council Conflict .811 .6^9 .391 .112 . 519 .297 . 863 on Issues Cluster 1 . 701 .051 -.141 .249 . 631 . 766 Cluster 2 .254 -.031 .207 .305 .452 Cluster 3 .665 .200 -.158 .276 Cluster 4 .081 -.365 -.034 Cluster 5 _. 101 . 571 Cluster 6 .332 (Italicized Relationships Significant at .05 Level)

PAGE 379

ated poorl y oitll , mode> „ hereas ^ ^^^ ciuatcr Bit[] ^ ^ ^^ « level u associate, moderately „ ith . „ de> ^ ^ ^ ^ ^^ is permeable t „ certai „ levels of ^^ ^ ^^ ^ ^^ ^ fUct b y cluster ace too co m ple x for such a states, however useful and xnteresting it might be. The com p] e xity „ f toe members' reaction to decision situations can be illustrated in y er another „a y bv considering Table 8,10. It reveals there is not alwa ys a g „od correspondence between a council division cluster and toe perception of important problems within the comm,„nit y . While —era need not disagree merely over important problem, the coefficients in Table 8-10 suggest a single disagreement cluster ma y appear to involve . series of problems. For example, cluster 3, development concerns, is associated strnnol^ n-st-u u faLrongiy with human resource erowt-J, ar .A c' growth and financial problems ^o each association appeers reasonabie when the elements of the disagreement cluster are considered. SI mil ar l y the division of councils over the growth matters of cluster , would app e ar to b aa r a P l aua ible rel at i n to Perceived community problems of ores growth and service provision. These illustrations portrav Mip -m,,, • ^ * L1 Ct y cne illusive nature nf t-u^ anature or the disagreement clusters A single cluster cuts across a number of cog„itive ly important areas. Hence tha abilit y of the clusters to capture precise relationships is diminished. in a sense, however, a concern for problems such as these is funda-tally misdirected. There wouid be little worth in drawing the specific disagreement to decision-mahing mode li„ k for 12 Florida cities. Rather ^e anal y sis has importance in that it shetches the P ermeatio„ of decision — bV Policies of different sorts. High cost P olicies, it was thought should be associated with different modes than low cost policies. An d MM

PAGE 380

361 Table 8-10 Kendall's Tau-c Between Member Problem Perception and Council Division Clusters For Councils of SES Group F Problem Perceptions Council Division HUMPROB GOVPROB GROWPROB FINPROB SERPROB Cluster 1 -.197 .374 -.116 .023 .045 Cluster 2 .271 .435 -.192 .138 -.135 Cluster 3 .735 -.398 .700 .483 .386 Cluster 4 .329 -.554 .474 .067 . 594 Cluster 5 -.001 -.024 .282 .775 . 464 Cluster 6 -.120 .300 -.047 .052 -.160 Cluster 7 -.107 .103 .139 .474 .006 (Italicized Relationships Significant at .05 Level)

PAGE 381

362 indeed there is see evidence this is the case. When conflict is taken as a surrogate for costs, it is found any policy generating substantial conflict is liable to envoke the generalized split decision mode. But this is hardly an iron-clad rule. Am ong the high SES cities, internal city affairs, regardless how conflictual they become, appear unlikely to engage this mode. In fact, for such matters, there is little reason to suspect any one mode will be referenced consistently. The four consensual modes tend to be referenced when disagrees is low. However even for such situations multiple possibilities are open to each council. Logrolling and arable agreement are seldom referenced when disagreement over development concerns is high. Yet these modes are not consistently fore. sworn in other high disagreement policy areas. in fact, there is slight evidence amicable agreement may be used frequently when prominent services within the community are under consideration. Summary In summary, it can be said policy does impact upon the way in which policies are decided. But the configuration of impact is quite complex. While it is possible to portray the form of policy impact on the several -des, the current data do not allow for precise specification of the relationship. At best, it is possible to depict the conditional relation of specific disagreement clusters with certain decision modes. Some decision-making modes are more amenable to disagreement than others. In particular, the modes dominated by personal relations appear less susceptible to employment in controversial policy situations than the other modes. And while a generalized pattern of council disagreement is associated with the controversy over issues, controversial matters are not excluded from consensual modes which demand members work toward agreement.

PAGE 382

363 The use of specific decision modes in particular issue areas can be specified only dimly. However, it is clear some policies, regardless of the conflict they generate among members, do engage generalized patterns of disagreement. On the other hand, it is typically the case that issues on which there is significant disagreement can be addressed by the generalized split modes as might be expected, and do discourage the use of consensual modes relying on personal interaction. Bargaining modes appear excluded only on issues in which conflict is high and group actions in the community are significant. The complexity evident in the relationship between the decision-making modes and policy issues is considerable. But, I think, it connotes more than the infirmity of the measures. The situation is likely more intricate than I have allowed. In fact, there are a number of factors whose confluence is likely to shape a council's reaction to policy other than a repertoire of decision modes. In the next chapter. I shall discuss a few of these factors.

PAGE 383

364 Notes 1 Heinz Eulau and Kenneth Prewitt, Labyrinths of Democracy; Adapta* Hon s, Linkages, Representation, and Policies in Urban Politics (Indianapolis; jfobbi-Merrill, 1973), p. 80. 2 Ibid . 3 Ibid . , p. 82. 4 Jonathan Turner, The Structure of Sociological Theory (Homewood, 111.: Dorsey, 1974), pp. 160-175. 5 Eulau and Prewitt, op. cit . , p. 83. 6 Ibid ., p. 83. 7 Ibid . , p. 173. 8 Ibid ., p. 83. 9 Theodore Lowi, "American Business, Public Policy, Case Studies, and Political Theory," World Politics , Vol. XVI (July, 1964), p. 688. 10 Heinz Eulau, "Policy Making in American Cities: Comparisons in QuasiLogitudinal, Quasi-Experimental Design," General Learning Press Reprint (New York: General Learning, 1971), p. 2. 11 Lowi, op. cit . , p. 689. 12 Aage Clausen, How Congressmen Decide: A Policy Focus (New York: St. Martin's, 1973), p. 16. 13 This analytical formulation is used in David Price, Who Makes the Laws? Creativity and Power in Senate Committees (Cambridge, Mass.: Schenthan, 1972). 14 See Roger Cobb and Charles Elder, Participation in American Politics : T he Dynamics of Agenda-Building (Boston: Allyn and Bacon, 1973), p. 96. 15 Theodore Lowi, "Decision Making Versus Policy Making: Toward an Antidote for Techology," Public Administration Review , Vol. XXX (MayJune, 1970), P. 318. 16 See Lee F. Anderson, Meredith Watts, Jr. and Allen Wilcox, Legislat ive Roll-Call Analysis (Evanston, 111.: Northwestern, 1966), pp. 10-11. For an extensive bibliography, see Duncan Macrae, Jr., Issues and Parties in Le gislative Voting: Methods of Statistical Analysis (New York: Harper and Row, 1970).

PAGE 384

365 17 The difficulties with such a question are obvious — the question seems to call for some answer. As such it is impossible to judge the meaning of "most." While such a format does not strictly allow for the comparison of members' answers in the manner of a likert scale, the pragmatics of the situation called for some sort of a trade-off. In particular, it was not possible to combine sophisticated gradation with extensive coverage within the confines of the interview schedule. 18 See Florida, Department of Administration, Revenue and Expenditure Data for City, County and Special Districts: 1972 (Tallahassee, Fla.: Department of Administration, 1973). 19 The areas were chosen not only for their correspondence to funding categories but also for their fit with the scheme of policies discussed in Chapter Three. An attempt was made to balance traditional (7) policy areas with amenities (7); and arbitration of conflict (4) with self-regulating (4) policies. One area — public safety — was chosen only for its relation to the funding categories. These policy types will be discussed further later in the chapter. 20 Environment and Florida Voters , work paper, No. 7, prepared by Cambridge Survey Research Institute for Urban and Regional Development Center (Gainesville, Fla.: Urban and Regional Development Center, 1974).

PAGE 385

CHAPTER NINE POLICY ISSUE CONFLICT AND POLICY TYPES Decision-making modes alone are unlikely to determine the sorts of policy made in council chambers. Salisbury and Heinz, in an argument presented in Chapter two, suggest policy results from the costs accrued to members by the circumstance of internal decision making and external activity. In this chapter, I will consider in some detail the propositions advanced in Chapter two. The direction the argument will take is sketched in Figure 9-1. First, I will examine community activity initiated by individual citizens and groups for its effects on policy formulation in the council. Then I will treat the manner in which policy appears linked to the types of costs which surround its deliberation. This will be done in two ways. Initially, I will assume council disagreements are indicative of high cost situations and hence can be used to empirically identify policies over which members perceive high costs. Next, I will posit policies which, according to the criteria established in Chapter two, can be expected to generate high costs and in which disagreements can be expected to arise. These tacts will be used to establish the relation between policy and the political situations attendant to its deliberation. I will attempt to test the propositions advanced in Chapter two for a range of issues and for issues of high salience within a council. Conflict Over Issues Conceptually, decision modes are designed to summarize the manifold 366

PAGE 386

367 POLICY ISSUE CONFLICT PARTICIPATION BY GROUPS AND INDIVIDUAL CITIZENS ON POLICY ISSUES: EXTERNAL COSTS POLITICAL BEHAVIOR OF COUNCIL MEMBERS ON POLICY ISSUES: INTERNAL COSTS POLICY TYPES: 1. Traditional Services 2. Amenities 3. Promoting Economic Growth 4. Arbitrating Between Conflicting Interests Figure 9-1. Major Categories and Variables Treated in Chapter Nine

PAGE 387

368 moments of choice situations. But they represent a focus of inquiry which may slight a consideration of significant forces bearing on local public policy. In particular, the modes fail to provide much insight into the reasons for disagreement among the council members over particular policies. Decision modes are accessed in choice situations; they capture the costs associated with a policy. The modes themselves do not trace the genesis of costs or explain why disagreements arise. There are a number of important models of urban governance, but most do not include the local council explicitly in their formulation. For example, studies of the informal power structure seldom draw the conceptual links between influential individuals and members of the council. Rather it is assumed the transmission of power affects local political doings. The pluralist model operates at a similar level of explanation, but with different assumptions. Basic to one formulation of this view is a notion of the "political system as a mere or less competitive arena in which groups compete for political advantage and in which individuals take cues for political behavior from the reference groups with which they identify." In this view, at least, council members can be linked to external forces in the community through the importance they attach to them. Sidney Verba and Norman Nie adopt a variant of the pluralist vision in their discussion of participation in America. And they are specifically concerned with the way in which the participation of citizens "affects the responsiveness of governmental leaders." They suggest the alternative ways in which a citizen can attempt to influence government can be considered to fall into four analytical modes of activity. Each of the participation modes can be expected to exert different types of influence over leaders, to involve outcomes of varying scope, to engage the participants

PAGE 388

369 in differing amounts of conflict, and to require different amounts of initiative on the part of the participant in order for him to participate. From the perspective of my discussion on conflicts over public policy in local councils, the interesting aspect of Verba and Nie ' s presentation lies in their consideration of the two citizen participation modes most likely to impinge upon the sitting council member — what they initially call cooperative activity and citizen-initiated contacts. Both these modes are conceptualized as activities available to the citizenry for influencing political leaders between elections. They constitute the primary forms of citizen input into the policy making of the council aside from voting and campaign activity. Citizen-initiated contacting need not come in "response to elections whose content and timing are set for the citizen, and in which the substantive issues are controlled by candidates and officials." 4 Rather they are determined by the individual alone: "He determines the timing, target, and substance of the act of participation." 5 Of course, the citizen does not act alone in all instances, only where the flexibility and substance of the participatory act is determined by the private citizen himself. Verba and Nie suggest there is another mode of activity which emphasizes cooperative group efforts. Here the individual joins with "others to influence the actions of government.'" These two modes of activity are hypothesized to exert different amounts of pressure on political leaders. As a rough distinction, Verba and Nie propose cooperative acts are likely to convey low to high amounts of pressure, while citizeninitiated acts should generate less pressure, on average. Each mode will differ in the types of outcomes toward which they are directed according to Verba and Nie. Cooperative activity, it was thought, should be geared to collective outcomes, while citizen-ini-

PAGE 389

370 tiated activity would be more appropriate for particularized outcomes. 7 Although Verba and Nie do not attempt to predict the impact such activities are likely to have on policy-making processes in the public arenas toward which such activities are aimed, they do indicate the modes are likely to embroil the participants in conflicts of varying degree. In particular, they argue cooperative activities tend to entail the possibility of greater conflict than citizen-initiated activities, because the "stakes o are usually higher." Given the detail with which Verba and Nie portray the consequences of participation, it is somewhat surprising that they leave the theoretical foundations of their work so bare. In particular, they, as most groups theorists, fail to specify how it is pressure that is generated. Failing this, their explanation of the differences between citizen-initiated and cooperative activity rests on the intuitive appeal with which they invest the two modes. In fact, there is good reason to accept their propositions if they are first animated by the notions developed in exchange theory. Taken on average, the efforts of individual citizens to influence should produce fewer costs for council members than similar activities involving groups. All things being equal, groups can be expected to marshall more resources and involve more concerns in which a council member will perceive some potential threat to his own goal fulfillment. Moreover, since groups are likely to form around policies which convey benefits in some manner to a number of individuals, they are also likely to cleave the set of interests which make up the community. That is, seldom does everyone profit (or not lose in some fashion) in a choice situation. The cleavages which rend the community may also divide the council; members will attend to the significant costs involved when groups confront the council.

PAGE 390

371 Table 9-1 Proportion of Council Members Indicating Disagreement, Citizen Contacts and Group Actions by General Policy Area General Policy or Funding Area Planning and Area Growth Crime and Police Funding Public Welfare Industrial and Economic Development Municipal Employees Zoning Natural Resources Racial and Ethnic Relations Public Utilities Urban Redevelopment and Housing Public Health Care Streets and Highways Airports and Water Terminals Mass Transit or Bus Systems Garbage and Sewage Property Taxes Courts and Their Operation Parks and Recreation Hospitals Financial Administration of the City Public Safety Municipal Auditoriums Form of City Government Members

PAGE 391

r 372 The tendencies toward unanimity in a council are many. As seen in Chapter two, unanimity enables the council member to diffuse the costs associated with many choice situations. Frequently the rational member will merely "go along." However, council disagreements over matters of policy may be fostered by an attentive and vocal public. Individual citizens can provoke policy disagreements, but, according to the exchange perspective, it is more likely group contacts in a policy area will be related to disagreement in that area. Table 9-1 reports the proportion of all Florida council members sampled who indicated that, within a given general policy area, contacts by individual citizens were "most frequent." For the same set of policy areas, it also lists the proportion of members who gauged each area to be the target of frequent group actions. These 23 areas are the same used in Table 8-2 , and the results of that table are reproduced for convenient comparison. The most noticeable aspect of Table 9-1 is the relative difference in the frequency levels which members ascribe to the contrasting form of external influence. If average rates of frequency can be taken as indicative of this contrast, it will be noted members report citizen-initiated contacts with greater frequency than group actions. The average proportion of members reporting citizen contacts is .350, while that for group actions is .268. Indeed an examination of the 23 areas will reveal only seven areas in which group actions appear more frequent than citizen-initiated contacts. At least among Florida council members, the great wealth of member-constituency interaction occurs between member and individual citizen. The policy areas that dominate the concern of individual citizens

PAGE 392

373 comprise traditional elements in the domain of city governance. The policies most frequently mentioned are as follows (see Table 9-1 ) : 1. Zoning 2. Streets and Highways 3. Garbage amd Sewage 4. Planning and Area Growth 5. Public Utilities 6. Property Taxes 7. Parks and Recreation 8. Crime and Police Funding Only the citizeninitiated contacts in planning and area growth do not fit into the scheme of concerns for which the local citizenry has come to expect redress at the hands of the local council. Certainly the council has long been the place where zoning variances may be had or property tax abatements approved. And members are often responsible for patching roads or hanging street lights; trash pick-up and playground facilities fall under their purview. Thus it is not surprising members find themselves petitioned over such matters. Group actions are most frequent in a slightly different set of policy matters. The eight most frequent areas, as extracted from Table 9-1 , are as follows: 1. Planning and Area Growth 2. Zoning 3. Natural Resources 4. Parks and Recreation 5 . Racial and Ethnic Relations 6. Urban Redevelopment and Housing 7. Streets and Highways 8. Public Utilities Notice that five general policy areas in this list are also among those areas in which citizeninitiated contacts are prominent. However, the additional areas— natural resources, racial and ethnic relations, and urban redevelopment and housing— represent relatively novel items on the local agenda. Moreover, the ranking of those elements common to both

PAGE 393

374 lists shows considerable differences in the manner in which the two types of external influence impact the various policy areas. Before examining this observation further, I will first attempt to demonstrate that, despite the more common use of citizen-initiated contacts, it is group actions which are related to disagreements in policy areas. In order to show the hypothesized relation between disagreements in a policy and external types of influence, I first calculated the phi coefficient, a special case of the Pearson product-moment correlation coefficient used for dichotomous data. Table 9-2 reports the coefficients calculated using all the council members in the sample. The results are extraordinarily clear. In only two general policy areas, do the coefficients of the relationship between policy disagreement and citizen-initiated contacts exceed that manifested between disagreement and group actions. One of these two — municipal employees — is not statistically significant. Public utilities, the other area, provides an interesting anomaly. It was 1974 that the dramatic effects of the OPEC oil embargo first impacted on the oil prices charged to generating stations throughout Florida. Electrical bills skyrocketed. In councils throughout the state, members were deluged by complaints from individuals over the costs of running air conditioners in their homes. Yet in most communities the significant dissatisfaction with the increased electrical charges had not resulted in the formation of groups of citizens. Group actions did not become a prevalent phenomenon until early in 1975, after the survey had been administered. Across the 23 general policy areas, there are only three areas in which the statistical relation between disagreement and group action is not significant. These are public welfare, municipal employees, and gar-

PAGE 394

375 Table 9-2 Phi Associations Between Council Disagreement and Types of Contacting By General Policy Area (N = 166) General Policy Area Planning and Area Growth Crime and Police Funding Public Welfare Industrial and Economic Development Municipal Employees Zoning Natural Resources Racial and Ethnic Relations Public Utilities Urban Redevelopment and Housing Public Health Care Streets and Highways Airports and Water Terminals Mass Transit and Bus Systems Garbage and Sewage Property Taxes Courts and Their Operation Parks and Recreation Hospitals Financial Administration of the City Public Safety Municipal Auditoriums Form of City Government AVERAGE ASSOCIATION . 212 . 271 . 152 .119 -.072 .211 071 .120 ,206 .262 151 .136 Citizen

PAGE 395

376 bage and sewage. It is not surprising cooperative activities were not aimed at the local councils in matters involving public welfare, since in most Florida cities welfare activities are handled by the county. The peripheral role of most cities in this regard would be known to groups for whom welfare was a primary concern. The lack of a significant relation between group actions and disagreement on municipal employees is more unexpected. In Table 8-2 , disagreement over policies involving municipal employees was shown to be quite high, and even the proportion of members reporting frequent group action in this area was greater than that for a typical policy area. Among Florida council members, disagreement over municipal employees appears insulated from the actions of the employees themselves. Of course, it may be that this type of issue produces disagreement only when the employees have yet to coalesce into significant groupings. Once the groups have formed, the members may adopt a posture other than that of outright public disagreement. The unexpected lack of a relation over municipal employees can be contrasted to the more expected absence of one between group activities and council conflict over garbage and sewage. Most of the council contacts with the public over garbage and sewage concerns involve individual citizens. Moreover it appears reasonable to suggest that unless garbage and sewage were to become a real problem, groups would not form to petition the council. In this fashion, such matters are contraposed to other traditional services as streets and highways which are of significant concern to groups involved in the economic development of the community. Since the association coefficients detect patterns of relationships, it is easy to overgeneralize the nature of the findings. In some councils, individual contacts may well be the decisive cause of conflict.

PAGE 396

377 However, the coefficients can emphasize the impact of a pattern occuring in an area of little disagreement. For example, municipal auditoriums ranked 17.5 among all areas in the proportion of members reporting high degrees of council conflict, yet the association of conflict in this area and group actions is quite high (.411). likewise disagreement is exceedingly low over public health care, mass transit, and racial and ethnic relations; but the association coefficients are quite high. In each of these areas, where conflict does divide the council, it is accompanied by the efforts of groups to influence policy. When each policy area is considered individually, the great weight of evidence presented in Table 9^2_ supports the proposition that group actions contribute to policy conflict in a more systematic fashion than do citizen-initiated contacts. Illustrative of this notion are the average association coefficients for the two contacting types. The average coefficient between council disagreement and group action is nearly twice that involving citizen-initiated contacts (see Table 9^2). Of course, this observation is conditioned by the mode of analysis. Only individual areas have been examined; many of the coefficients are low. There is yet another way of examining the form of the relationship. It begins with a simple assumption. If group activities are more important in generating council disagreement over policy issues than citizeninitiated contacts, it might also be expected the relationship will hold across all the general areas. That is, as a general rule, groups actions should be related to conflict on policies more systematically than private actions regardless of the issues involved. Using the proportions of council members commenting on disagreement and contacting activities reported in Table 9^1, it is possible to exa-

PAGE 397

378 mine the systematic connection between the two considering each of the general policy areas as observations. This focuses on the pattern discernible in the overall results of the many interviews. A calculation of the product-moment correlation shows disagreement over policies is related to citizen-initiated contacts with a coefficient of .610. The coefficient of the relation involving cooperative activities is .848. As with the series of relationships focusing on individual policy areas, the link between group actions and disagreement is stronger than that involving citizen activities. It should be noted, however, that the presence of group activity also signals citizen activity. The two are highly associated (r = .709). It might be supposed then group actions are re-enforced by citizen-initiated contacts with members. The combination of the two may be necessary if disagreements are to occur. This view is not sustained by a consideration of the partial correlation coefficients. When the relationship between the pattern of council disagreements in the 23 areas and citizeninitiated contacts is examined, with the effects of group actions controlled for, the relationship falls to near zero. On the other hand, controlling for the effects of citizen contacts on the relation between group actions and council conflict has little impact on strength of the relationship. In short, while group actions occur in the same context as citizen-initiated contacts, it is the group actions which impinge heavily upon the council members. Moreover, citizen contacts do not appear to reenforce group actions in any significant manner. Revealed Interest Versus Posited Interests It is hardly surprising group activities appear to have such an impact on local councils. After all, group theories have a long and hal-

PAGE 398

379 lowed tradition in the commentary on American political institutions. And it is not without some merit such conceptualizations have persisted and prospered. But there are difficulties with the group perspective which go beyond those noted in the previous section. One particularly germane problem was noted, in passing, in the previous section. Namely, how is the differential attention to various policy areas by groups to be explained? That is, why is it group actions are more prevalent in some policy areas than others in councils throughout the state? This is a key question for this study. It forces one to move beyond the focus peculiar to much of group theory; it forces a move beyond the single case, the single community, or the single legislative body. While it may be reasonable to assume groups are likely to form and to seek redress in the legislative arena when there is substantial concern over an issue, this proposition must be tempered with a concern for the many contexts in which groups attune themselves to similar policies. Of course, it might be supposed attentive groups already exist in most communities and they register their feelings with local political leaders in a rather predictable fashion. But this explanation avoids the central difficulty as well— the manner in which groups come to be involved over similar sets of policies in local governments across the state. In a literal sense, it is not the groups which remain the same across the many Florida cities. Per force, it must be aspects of the policies themselves. This realization has important consequences for understanding city councils and their impact on municipal policy. Group theory provides a succinct identification of just these aspects. Policies, it is argued, involve interests. These interests envelop groups within the community and galvanize members of the councils

PAGE 399

380 themselves. All things being equal, the greater the conflict of interest, 12 the more conflictual the behavior. Once said, however , the notion that interests are invoked by a policy may seem of little use in approaching the manner in which councils attend to policy. After all, the manner in which interests become embroiled in a choice situation is likely to be quite complex. Interests are seldom unidimensional, nor are they isolated from one another in time. Nonetheless, an analytical focus on interests can provide a conceptual pathway leading to an improved understanding of how choice alternatives become public policy. But the direction of such insights cannot be divorced from an initial consideration of the nature and determination of such interests. Definitional postures prove to have substantial ramifications for the way in which problems of public policy are understood and researched. Hence, before applying the notion of interest to the problem of council attention to public policy, I will first consider how the concept of interest might be used in such a problem. In an incredibly perceptive essay, David Greenstone draws the issues swirling around the definitional problem. He quotes Berkley's objection to the conceptualization of interest groups as stable classes: '"The real question' about contemporary politics is why 'the living acting men and women change their forms of action and cease to do now what they did formerly. " ,13 But if groups are not merely stable classes of individuals working in concert, how are we to identify them? Greenstone argues most group theorists have focused on the activity of groups. "In the study of either a group or policy processes, the first step is to identify the interest, the direction of these 'changed lines' in which some men are 'shunting' the 'conduct' of others." 14 Interest is to be found in the activity of groups. The concerns of groups are revealed in the matters

PAGE 400

381 with which they become involved. This subjective definition of groups turns on reveale d interest. It is contraposed to the notion of objective interest rejected so long ago by Bentley: A man who is wise enough may legitimately predict, if he is addicted to the habit of prediction, that a group activity will ultimately form along lines marked out by some objective condition which he thinks he detects. But the interests that we can take into account must lie a good deal closer to the actual existing masses of men. l-> The interests of groups, according to this view, can be known in the absence of group activity. In fact, it is argued, group interests do not have to square with the short-run behavior of a group. Interests can have behavioral manifestations which are modified or moderated by contextual factors; objective interests may also remain latent. The researcher, of course, is charged with positing the nature of such interests, an adventuresome but delicate task. Each of these epistemological arguments can be extended to cover the interest shown by council members. Analytically, interest is not limited to groups. The concept is widely applied in the literature of political science. 16 The epistemological positions can be applied to individuals, quite simply. For example, if it is assumed individuals act in accord to logical rules in order to achieve or realize their interests, then the behavior of the individuals can be inferred to document interests. Similarly, it can be assumed an individual has a specific interest or set of interests. These interests when played out according to logical rules should result in specific choices. Thus the epistemological form of the argument is invariant between individuals and groups. Naturally, each epistemological argument has certain advantages and disadvantages. Bentley was certainly correct in arguing the subjective

PAGE 401

382 view heaves closer to the empirical situation. But as Greenstone points out, this approach has very little anticipatory capacity; it is primed for ad hoc explanation. The objective view hinges on the a priori . This has its own dangers, the most obvious being the conjecture animating the assumptions about interest. The researcher must be able to determine the interests of the group or individual in some fashion. But there is a further difficulty. If the actions of groups or individuals are not those predicted, then there is no way of telling whether it is the fault of our set of posited interests or of our attribution of logic to the actions.-^ Hence, this is a path fraught with difficulties; it builds hypothesis on assumption in a manner that is potentially non-falsif iable. Nonetheless the posited interest or objective approach is quite attractive. It promises to lend parsimony to an otherwise chaotic field of variables. It narrows the interests to be investigated and supports a taxonomical consideration of their content. Both approaches, revealed and posited interest, can be used to gather insights about public policy. However, these insights must be fastened with a consideration of the relation between interests and public policy. It should be clear the epistemological position relying on revealed interest lends itself to the view interests are just the policy preferences 1 8 of individuals and groups. But numerous objections can and have been raised to this proposition. 19 jr or the purposes of this discussion, this view has one serious fault; it fails to address the reason why groups and individuals or council members themselves might evidence such concern so consistently, in so many councils, over the same issues. There must be something about particular policies which call for individual or group action, and which prompts councils to disagree in the face of considerable strains toward unanimity. 20

PAGE 402

383 Of course, there are several ways of depicting the relationship between policy and interest. One approach, utilitarianism, focuses on the preferences of individuals in a particular way. As argued by people such as Brian Barry, a "policy, law or institution (is) in someone's interest if it increases his opportunity to get what he wants. . . whatever thar is."This perspective has the advantage of allowing for mistakes on an individual's part and sustains the distinction that policy support may flow from self-interest or moral considerations. Typically, however, this view does not incorporate a substantial social component into the way the wants are regarded. Thus other perspectives have been developed which explicitly include social life. For example, some authors have argued for a set of universal human needs. z But such arguments have difficulty in drawing the connection between need and public policy. So our problem remains. Suppose, though, that the social component could be added to the notion of wants. Assume wants may be taken as generalizable, all men have similar wants; then this view could be used to solve the appearance of similar factions in so many places. Only one problem remains. How do you explain why similar wants percolate into different political arenas in the same fashion? This requires wants which are grounded in a social context transcending the differences between different localities. If wants are to form the basis of our understanding of the interest-policy relation, then we must be willing to assume there is something similar about the various political arenas. In truth, political observers are more often then not quite willing to assume different contexts have much in common. Typically, comparative studies will assume similar levels of development in different countries or states. Societal differentiation and capacity are often included in

PAGE 403

384 a study's caveat, ceteri^aribus. Thus it should not be too surprising if I add another assumption to a growing list. Namely, the meaning of policies to council members, citizens, and groups is assumed to be relatively similar throughout the state. In particular, it is the results of policies which must be held to have similar meaning.23 If this assumption is granted, and if wants themselves are taken to be fairly similar among individuals, then the relation between policy and interest can be drawn, if only darkly. In each community throughout Florida, at any point in time, it might be supposed individuals and groups have their wants met in differential fashion. Socioeconomic data, at any rate, would suggest this is the case. However, if Hobbes was right, wants are continually renewing themselves. And to the degree the actions of the local government involve the satisfaction of such wants, each governmental policy can be taken to involve the manner in which the distribution of want fulfillment takes place. Each policy action involves a potential threat to the fulfillment of wants; each choice involves potential costs. But some choices are more costly than others. And if the meaning of policies is widely shared, on the average they should involve similar costs in very different settings. Of course, this leaves a problem. What measures are betoken of such costs? * Previously I argued high conflict situations are indicative of high costs in the choice situation. Disagreement among the members of fers an intimation that the set of policies pursued in a community may be transfigured. The generally shared expectations, established with respect to a policy, act to engage the clash of wants, of interests. The possibility a council will take u P a given policy is likely to engage individuals in the community as well as the members themselves. Group actions, as I

PAGE 404

385 have already suggested, can make choice situations considerably more costly for council members than citizeninitiated contacts. But this proposition does not address the policy areas in which citizen or group actions are likely to occur. In fact, it is difficult to gauge the level of costs associated with policies over which there is frequent external contacts of council members. Still, it can be assumed frequent group actions will be geared toward policies which might have trouble being accepted by the council. While it is certainly possible groups will form around specific policies in isolated instances, it appears more probably frequent group actions will be geared toward policies capable of generating high costs. Citizen-initiated contacts, on the other hand, can be expected to center on less costly issues. Since citizen efforts, on average, will carry less weight with council members than group action and because it is assumed citizens are aware of the meaning policies have in the council; it is supposed citizen efforts will be directed toward those low cost policies in which such efforts may yield positive results. These measures can be used in conjunction with either epistemological approach to interest. In the revealed approach, certain logical rules of behavior are presumed to act within a set of empirically defined policies. An examination of the various measures on these policies reveals in which areas interests, and thus costs, are high. Accordingly, these are the policy areas in which action by the council would appear least likely. Of course, this proposition cannot itself be deduced from the findings. It grows from the consideration developed in Chapter two. In fact, it is just this sort of proposition with which the posited interest approach begins. One form of the posited interest approach was presented in Chapter

PAGE 405

386 two. Collective policies, it was argued, were the result of the interaction of two significant sets of costs. Analytically, these were considered as eternal and decisionmaking costs. 2 * Both sets of costs were supposed to be ah!e to assume different vaiues. Arbitr ari ly . these values were reduced to two levels, high and lew. This distinction represents a departure from the more broadly sketched argument presented ahove with respect to costs involved in group activity, m the argument taken fron, Salisbury and Heinz, and presented in Chapter three, external costs were considered to result from the patterns of group activity impinging upon the decisionmaking arena. Integrated patterns of group iobbying produce high costs; fragmented patterns produce low costs. „y earlier discussion of group activity does not differentiate between these two situations. However, it does argue citiaen-initiated contacts win be associated with low cost policies and thrnst relative!, low costs on council .embers. Conceptually, citiaen-initiated contacts can be taken to be the n»st extreme fort, of fragmented group activity. Analytically, they are groups cleaved in so many ways as to become individual Hence, extending the posited approach, my previous position can be reached. Salisbury and Heinz argue decision-making costs are most likely to be low when the policies under consideration involve matters of value allocation rather than structura! change." Since structure change offers a threat of a general rale change affecting whole classes of firms, industries, groups, and individuals, members often move the deliberation over these high cost policies out of the council chamber and into other deliberative settings. 0, course, the form of policy can differ substantially regardless whether it involves allocation or structural change. According to Salisbury and Heina, allocation policies can be distributive or redis-

PAGE 406

38 7 tributive; structural policies can be regulatory or self-regulative. Each policy type engages a different pattern of costs as it filters through the process of reaction and adaptation of the council members. 26 For example, redistributive policies are likely to envoke the highest costs possible. They represent a visible change in the distribution of benefits and the imposition of costs. 27 Similarly, regulatory policy connotes an enlarged role of government with ramifications for many aspects of the community. On the other hand, distributive policies involve slight costs. "Distributive policies give benefits to groups, case by case." And they represent traditional forms of urban governance. "Most urban policies of the nineteenth century could be labeled as distributive...." 28 Selfregulative policies have a peculiar character. Theoretically, the costs associated with such policies should exceed even those associated with redistributive choices. However, because the costs are such, local councils _ should seldom touch on the matters involved. The policy involves a deference to private rule making. m a sense, there is no public policy involved other than a recognition of private prerogative. Nonetheless, councils do pursue policies which enhance private prerogatives. These policies are like-.y to be re-enforced by the climate of opinion swirling around matters left to self-regulation. As a consequence policies framed to entrench or promote self-regulation are likely to have small decision-making costs associated with them. It appears unlikely, however, that groups and individuals will neglect matters of self-regulation merely because they sustain the climate. As a consequence, the external costs on a cc cil dealing with self-regulation matters may be quite high. To investigate a program of posited interest, it is necessary to specify the matters of policy associated with the different levels of :oun-

PAGE 407

388 cost. Here the policy types labeled by Salisbury and Heinz are of considerable worth. Indeed, Jay Goodman has attested to illustrate how local governmental policies might be grouped within the four broad types of policy. 29 My approach differs fro, his slightly, although the results are somewhat si.ilar. In Chapter two, I depicted the relation between the policy types and the classification of local policies provided by Adrian and Williams. They characterize local policies as providing 1) amities, 2) traditional service, 3) growth or proving growth, and 4) arbitration between conflicting interests. Ite ms one and four symbolize relatively new areas of local concern; two and three fall within the purview of traditional local concerns. In my conceptualization, this distinction, along with the benefits conveyed by the policies, serves to -ke the bridge to the policy types and cost levels formulated by Salisbury and Heinz. And since Adrian and William animate their categories with a series of specific references to local policy, their discussion forms the foundation of my own specification of the policy bundles. Table 9-3 illustrates the policies I have grouped under each of the broader policy types. These policy types for* the analytical base of the posited interest approach. They make the connection to costs that is necessary in order to check if choices are made in accord with our theoretical notions. Empirically, this involves isolating policies by type and monitoring the disagreement and contacting found in each broad type. High disagreement levels and significant group activity should be noted in the high cost Policy types. Citizen-initiated contacts should be noted most frequently in the lower cost areas such as distribution or self-regulation. Both procedures, revealed and posited interest, will be used to in-

PAGE 408

389 Table 9-3 Classification of Policy Issues into Four Types Traditiona l Services : [Allocative/Distributive] Educational Fire and Public Safety Streets and Highways Crime and Police Funding Garbage and Sewer (includes collection, disposal and construction) Public Utilities (includes gas, water and electric) Zoning (includes variances, reclassification schemes and procedures not containing areal growth-related elements) Financial Administration of the City Property Taxes Rate Increase (includes gas, water and electric) Providing and S ecu rin g Life's Amenities : [Allocative/Redistributive] Revenue Sharing Money Distribution Public Welfare Urban Redevelopment and Housing Public Health Care Parks and Recreation Natural Resources (includes environmental concerns) Mass Transit Planning and Area Growth Preservation of Open Spaces and Density Control Community Development Projects Promotions of Economic Growth: [Structural/Self-Regulative] Municipal Auditoriums and Convention Facilities Industrial and Economic Development Airports and Water Terminals Hospitals (includes construction and capital outlays) Downtown Development Projects Tourism Annexation Bond Issues (for traditional service or promotion of economic growth) Arbitrating Among Confl icting Interests : [Structural/Regulative] Form of City Government (concerns over charter) Courts and Their Operation Racial and Ethnic Relations Comprehensive Ordinance Reviews Municipal Employees Municipal Procedures (includes non-budget related items) Bidding Procedures and Regulation of Franchises Local-Federal Relations Local-State Relations Local-County Relations Administrative Regulations

PAGE 409

390 vestigate the policies pursued by local councils. It should be re-em P hasized at this point, however, the revealed approach can only describe the data. By itself, it offers no categorization. The posited approach begins with a categorization. But if the observations of council behavior do not square with the conceptualization, the fault may lie with more than the initial grouping of policies. It might also be the extended chain of assumptions standing behind the indicators I have chosen has particular difficulties. For example, the posited commonality of perceptions with respect to the policies examined here may not stand scrutiny. Further, it is also possible that council members do not gear their behavior to their interests in a logical fashion. The posited approach cannot distinguish between these many alternatives when discrepancies have been detected. An Examination of Council Policy Council members throughout the state report the frequency of contact with citizens interested in council policy making exceeds that shown by groups. As a rule though, the proportion of members indicating frequent group activity in a policy area is greater than that suggesting council disagreement over policies is high. But this should not be taken to mean either conflict in councils is always low or citizen activity is persistently high. These are merely tendencies which surface when consideration is given to the range of policies handled by councils. In specific policy areas, the pattern of behavior can be quite different. An examination of Table 9-1 illustrates this point nicely. And, indeed, this very proposition was dealt with, if only tangentially, in my discussion of the conflict over issues. It should be no surprise that different policies attract and gener-

PAGE 410

391 ate different reactions. This assumption stands behind the conceptual position of revealed interest. Empirically, however, the proposition offers a challenge to the way in which we take interest to be revealed. Namely, how are we to know when the conflict in a policy area is high? Or when is the proportion of members reporting group activity over a policy large? There are differences in the reports filed on each policy, but there is no convention fixing the point at which they signal great costs. Thus, I have resorted to the accounts of the members themselves. I have assumed there are high costs to be discovered and have taken them to exist in areas where members manifest special notice of disagreement or external contacting. 30 There are several policy areas which score highly on each of the three measures. The positive figures arrayed in Table 9^ are indicative of more than average activity in each of the policy areas, and five areas share positive scores. Planning and area growth, zoning, natural resources, public utilities, and parks and recreation score highly on each of the indicators. Council disagreement is also higher than average on policies involving municipal employees, financial administration of the city, and form of the city government. But these areas fail to overlap with any of the policies rated highly on the remaining two measures. Indeed, these policies, over which disagreement is high, appear to share little with the policies remaining highly rated on the citizen activity index. Crime and police funding, streets and highways, garbage and sewage, property taxes and public safety, all fall within the rubric, adopted under the guise of the posited interest approach, of distributive policies. These areas, in turn, bear small relation to the types of policy on which group actions are frequently reported. Local community groups

PAGE 411

392 Table 9-4 Z-Scores of Proportion of Council Members Indicating Disagreement, Citizen Contacts and Croup Actions by General Policy Area General Policy or Fu nding Area Planning and Area Growth Crime and Police Funding Public Welfare Industrial and Economic Development Municipal Employees Zoning Natural Resources Racial and Ethnic Relations Public Utilities Urban Redevelopment and Housing Public Health Care Streets and Highways Airports and Water Terminals Mass Transit or Bus Systems Garbage and Sewage Property Taxes Courts and Their Operation Parks and Recreation Hospitals Financial Administration of the City Public Safety Municipal Auditoriums Form of City Government Members Divided

PAGE 412

display concern, unregistered by either eouneils or individual cltIzens en otters involving racia! and ethnIc ralatlons> ^ ^ ^^ ' -« and housing. Both thcse pollcies ^ ^ ^ _ ^^ ^ »"ed a, high cost poUc y types-arbitrating conflicting interests and provision of amenities. Before concluding the high cost polIcies ^ ^ ^^.^ one difficult, should be noted . Table ^ ^^^^ a ^^^ ^^ ^^ ^ one pol icy area compares to another. However, because the standard scores are computed with respect to the entire set or data, individual polio, areas ma y he ne g iected. Fo r example, 14 of the 23 areas share similar levels of cost; five are high and nine are low. This portends -a than a fortuitous conjunction. Of coarse, static response sets by the members might account for the association. However, a more generous interpretation would focus on the Knner „ „ hich actIvIties Jo in _ deed associate and come to generate counci! disagreement. In either case, a .ore halanoed perspective on the costs can he gained by comparing the scores in each p„iic y area. This win direct our concern toward differences in the proportion of responses, controlling in a crude fashion for systematic response and association. There are threa sets of differences between the various measures. The ealeulated results are displayed in Tahle 9-5. As before, the results ean he categorized as high or low, positive or negative. In this instance, the level does not signif y the relative proportion of members reporting activity in a P olic y area, but rather the difference in activity, as gauged b y different measures, for the same poiicy. A negatIve value symbolizes the reiative importence of the subtrahend in the operation. A positive value has the opposite connotation. The set of difff i lO M W

PAGE 413

394 Table 9-5 Difference Between Z-Scores For Proportions of Members Indicating General Policy Area and Conner! Divided, or Attracted Group and Citizen Act lvity PifferenceJSetween Z-Scores of PrrmnrM nn D f: General Policy Area Council Divided and Group Actions [Div Grpl Planning and Area Growth Crime and Police Funding Public Welfare Industrial and Economic Development Municipal Employees Zoning Natural Resources Racial and Ethnic Relations Public Utilities Urban Redevelopment and Housing Public Health Care Streets and Highways Airports and Water Terminals Mass Transit or Bus Systems Property Taxes Courts and Their Operation Parks and Recreation Hospitals Financial Administration of the City Public Safety Municipal Auditoriums Form of City Government .315

PAGE 414

395 ferences can be taken to for m a pattern which provides additional information about the nature of the policies. Logically, of course, there are eight possible combinations. 31 However, the transivities which must exist between the sets limit the number of possibilities to six. Each has a distinctive interpretation which will be examined in turn. Before beginning an analysis of each of the patterns, it is useful to recall, once again, citizen-initiated contacts constitute the greatest proportion of all member responses, while group actions register second. Against this backdrop, certain comparisons can be made. Specifically, it can be expected citizen-initiated contacts, as the subtrahend in two of the difference sets, will dominate their own difference sets and yield negative scores, In those instances where this is not the case, council disagreement or group actions can be taken to be particularly intense. Alternatively, it may be imputed citizen activity is yet to form around the specific policy involved. For example, disagreement scores are larger than citizen-initiated contact scores in five areas: planning and area growth, public welfare, municipal employees, financial administration of the city, and form of city government. These areas give the appearance of being high cost policies, and represent, with the exception of public welfare, a subset of areas on which higher than average scores were identified in Table 9-4. Similar conclusions may be advanced with respect to relations entailing group actions. However, here our conclusions must be tempered with a realization the difference in the average proportion of members reporting council disagreement and group actions is quite small. 32 Thus the dominance of either group actions or council disagreement is likely significant of substantial concern. An investigation of the six patterns can be facilitated by recogniz-

PAGE 415

396 ing that the expectations noted above can be used to gauge the costs of policy areas in a more refined manner than the raw standard scores. Policies associated with the highest costs should find the proportion of members citing council disagreement greater than that reporting group actions which, in turn, should outnumber that in citizen-initiated contacts. This preserves, inversely, the transitivity noted among the average proportion of members reporting frequent activity along each measure. For simplicity, this relationship can be written as: Disagreement > Group Action > Citizen Contacts Only four policy areas are embraced by this relation: Planning and Area Growth Public Welfare Municipal Employees Form of City Government Public welfare is the only policy area of the four not falling within the set of high cost areas identified earlier. A close examination of the differences will show this area barely meets the requirements of the relation. The values are within the discriminating power of the set of differences. However, the other three areas appear to qualify as genuine areas of high cost in Florida city councils, but areas to which groups pay moderate attention and individual citizens pay little. Are there policies to which the councils devote considerable attention and concern, but which hold little attraction for individuals or organized groups within the community? If so, council disagreement might be expected to be high in such areas, while group action will be low. The order of the relation can be depicted as the following: Disagreement > Citizen Contacts > Group Action Searching Table 9-5 for this pattern reveals only one policy whose scores

PAGE 416

39 7 are signed in the proper fashion-the financial administration of the city. Although members disagree frequently over financial matters, groups within Florida communities do not direct their efforts at affecting such policy. Even citizens do not register significant interest in the mundane aspects of financial administration. T his is the only policy area treated in such a fashion; financial matters appear to be the exclusive domain of the city council. The singular concern of members for financial administration is quite unusual. Typically, citizens and groups are attracted to a number of policies. Moreover, it is possible and quite likely citizens and groups will have significant interests which are unmatched by manifested disagreement in the council itself. For example, citizens may well attend to a number of policies over which the council has established convenient operating ru l e , In such a case, citizen-initiated activity should be quite high relative to council disagreement. Of course, this may signal presence of intense citizen concern or the undeveloped state of council interest. In either event, such areas are associated with relatively low costs. The relation of the measures can be pictured as: Citizen Contacts > Groups Actions > Disagreement There are eight policy areas in which citizen contacts predominate: Municipal Auditoriums Hospitals Parks and Recreation Airports and Water Terminals Streets and Highways Public Health Care Public Utilities Note that two of these areas are ones in which, according to the raw standard scores of the proportion of members reporting activity on each of the measures, , here is l arge amounts of council disagreement and group

PAGE 417

398 activity. Namely, public utilities and parks and recreation attract attention across the board. Nonetheless such policies attract citizen interest in great measure, and it can be p reS u ra ed citizens act in order to secure rather Mediate redress. Thus these areas, while assuming the guise of high cost policies, may be of less importance than ones where council or group interest has been registered with greater intensity. These eight policies are ones in which citizens have considerable interest. However, community groups also have formed in each of the areas, and have had their concern catalogued by the members. It might be imputed groups sense there is the possibility for action on such matters. But groups do not form around all matters; some appear relatively mune from group efforts. As with financial administration, groups not likely to attend to some policies. Unlike f inancial admi nistration, such policies are foci for citizen-initiated contacts which proportionally outnumber the degree of conflict found among council members. The Pattern of relations featuring the noninvolvement of groups can be writlm3ups are ten as; Citizen Contact > Disagreement > Group Actions Areas characterized by the absence of significant gronp activity can be estimated from the figures in Tabie 9,5. Six policies dispiay this pattern: Crime and Police Funding Industrial and Economic Development Zoning Property Tax Courts and Their Operation Public Safety Only one policy, zoning, was rated highly on any of the individual mea-

PAGE 418

su.es; tt scor ed highly o „ an ^ flowever> ^ en ^ ^ ^^ — «. indicate gro up actions are far Mtstripped ^ jisaMt ^ CMZe " COnta " S " "* ° f th = »«« «-. citizen-initiated ™ tM predominated; gr o„ p actlons „ ere jargely ^^ Logical!,, there should be some policies in which citi2en COTtac£s are relatively m„ and grou P activity is exceedingly hlgh , or: Group Act i on „ Msagreement , cltl2en ^^ An elation of TaMe ^ „ 1U de „o„strate „„ ^^ _ ^^ ^ this particular pattetn. In every Insta „ ce> ^ ^ ^^ ^ hzgh. citizen activities exceed the ievel of conncii disagreement. Ihls W ' tbe eXlSte " M ° f POli ^ *** — considerable importance ior the community exterua! to the conncii b nt littie importance ior the memhers themseives. This sixth io g icai patterns can he depicted as follows: Croup Action > Citizen Contacts > Disagreement Croup actions, which are typicaliy iess fre q „e»t than citizen contacts can he expected to he g reater tban either citizen activity or council disagreements in those poiicy areas which nave not been deveio P ed as matters of contention within the community at iar g e. Crou P actions fort, in response to .received difficulties within the comities, hot they find their animus in the lar g er environment of P olioy actions. Crou P actions -y well occur in those areas in which si g nificant national g ronps have already formed. Three policy areas exhibit extraordinarily large amounts of group activity. City connciis do not appear to have become fully involved in matters concerning urban redevelopment and bousing, raciai and ethnic relations, and natural resources. Part of tbe reason for tbe considerable group activity in two of these areas can be linked. The Housing and

PAGE 419

400 Community Developraent Act Qf l9?4 ^^ ou[ ^ prMise ^^ ^^ Sroups, especi a l ly Blecbs, might receive £ „ nds ^ ^.^ ^ ^ housin 8 an d ^eveiop^. Throug „ out ae state _ counciis ^^ t ^_ selves petitioned by grouns for f„„j y groups for foods whose disbursement would rot be Voided until the spring or 1975 . Cour , cl]s disag „ ^ ^ ^ esced; tbe issue „ as yet to _ bcfore ^ ^^ ^^ ^^ eyidenoe considerable disagreement oyer matters involving natural re _ sources, group actiyity in this policy _ ,. ^^ ^ ^ ^-res witb tbe significant concern noted earlier oyer environmental -tors manifested by citi.ens and groups ^^ ^^ ^^ aoout the environment ranks mucb bigber among Flo ridia„ s tban among t be citxzens of other static T t• states. It rs very unusual to find environmental problems cites -re fluently tban economic ones, and, in Flo rida, environmental issues are listed four times as fluently as economic ones "33 An examination of tbe six p atter„s of transitivities which exist a— «measures of council conflict and external contacting bas demonstrated tbe considerable diversity of tbe 23 lsS ues. By most accounts tmajority of policy areas are not iinbed to substantial costs. Moreover, issues attract interests differentially. Several are attended by individual citUens alone, otber policies bave surfaced oniy among varrous community groups. Flnancial admlnistratio „ ^^ ^ ^ ^ interest of tbe council members. How ere we to mebe sense of sucb a v a ried pat tern of costs, Tbe posited approecb to interest geve bope policres could be grouped in . „ ay „ hich „ ould order ^ expectatlons ^ now policies become l a „. Provisio „ „ f ^^ ^ ^^ ^^ f " tHe arbltra " 0n ° f »»""«"« '-rests, it „ as argued, sbould involve tbe 8 reatest costs. Traditional policies and matters fostering

PAGE 420

401 the promotion of economic growth should be much less costly in most local communities. These propositions can be tested by using the same standard scores reported in Table 9^6. Using the classification of policy into four types presented earlier, it is possible to group the scores so as to display the approximate costs associated with each of the types. Table 9-6 arrays the figures that allow the impact of each issue to be highlighted while aggregating the policies by type. Contrary to expectation, amenities are not associated with the greatest level of disagreement among council members. On average, members are more divided over the set of traditional issues. The difference between the two policy types on this measure is small, but the thrust of the observation stands. However, the bite of this finding is cushioned somewhat by a closer examination of the policies falling within each of the policy types. Two issues have significantly affected the mean score of the amenities type— mass transit and public health care. Neither of these policies are pervasive subjects of council concern throughout the state. Many communities lack the resouces or the need for mass transit; and in most instances public health care is a county concern. On the other hand, it must be noted both policy types share only three areas of more than average council disagreement. Substantial council disagreement over the entire range of policies does not appear. There is a difference between regulatory and self-regulatory policies, however. Matters involving the arbitration of conflicting interests draw considerably more conflict than those meant to foster economic growth. While this observation may be an artifact of the small number of issues included in each type, the finding is in the direction expected.

PAGE 421

402 -—irss* ss^S^-^,£sr' and Classified Policy Type< Policy Area Policy Type _General_jjlicy Areas Z-Score on Members Divided Traditional Policies [Allocative/ Distributive] : Zoning lg21 Financial Administration of the City Public Utilities Garbage and Sewage Crime and Police Funding Streets and Highways Property Taxes AVERAGE SCORE Amenities [Allocative/ Redistributive]: Plar.n.ng and Area Growth Natural Resources Paris and Recreation Public Welfare Urban Redevelopment and Housing M T. -.HUD Mass Transit or Bus Systems -1 109 Public Health Care _i 10 q AVERAGE SCORE _[ Q 2 2 Promotion of Economic Growth [Structural/Self-Regulating] : Industrial and Economic Development Airports and Water Termj Municipal Auditoriums Hospitals AVERAGE SCORE Arbitrating Between Conflicting Interests I structural/Regulating] Municipal Employees Form of City Government Racial and Ethnic Relations -l.*036 2.509 .333 .036 -.406 -.406 Courts and Their Operation AVERAGE SCORE -1 Z-Score on Citizen Contacts 2.969 1.727 .672 1.000 -.497 .085 -.497 -.352 .305 Z-Score on Group Actions 1.897 .255

PAGE 422

403 Our description of the policy types can be supplemented with the findings obtained in the inspection of the pattern of transitivities. Note that the four high cost issues— planning, public welfare, municipal employees and policies involving the form of city government— fall within the two policy types posited to generate substantial costs. These policies involve regulation or amenities. But this is not to suggest the posited types have performed perfectly when the patterns are recalled Clearly, the dominant role of citizens in the concern over parks and recreation would be example enough to destroy this notion. Still, this example suggests that examining the policy type averages alone can be misleading. Individual policies within each of the types must be considered,. This caveat is re-enforced by an inspection of the mean scores for the four policy types on measures of group and citizen activity. The large value obtained when the scores are averaged for citizen-initiated contacts on issues of traditional concern is indicative of the highly fragmented contacting pattern expected for this type of policy. The fragmentation of interest exceeds that demonstrated on amenities in the manner expected. But again the impact of specific issues must be warned against. Mass transit and public health care again play devil with the observations. But in this instance, their impact does not appear nearly as great as the general lack of interest registered by council members in the four areas constituting self-regulation. Remember matters fostering economic growth should involve little cost. In fact, among economic growth issues, low cost appears to mean neglect. Perhaps, because truly self-regulatory issues do not come before the council, the issues included as indicators of a concern for economic growth should not be

PAGE 423

404 expected to encounter the same pattern of fragmented contacting as distributive matters. Self-regulatory or economic growth issues may exist at the boundary of the rules of the game played in most communities; and hence be beyond the pale of normal activity. Whatever the reason, issues of economic growth rate lowly on each aggregate indicator; yet the patterns suggest these are among the areas where citizen attention outweighs group or council interest. Thus, if the patterns are to be believed, the aggregate indicators have again failed to discriminate finely enough. Matters of economic growth involve a fragmented contacting pattern, but the level of activity lies beyond the power of aggregate measures to discern. Similar comments can be made of the relation between group actions and the policy types. Amenity issues manifest the greatest group activity, and clearly outstrip that associated with traditional policies despite the depressing effect of the scores in mass transit and public health care. However, it is probably best to conclude the areas in which group activities predominate are urban redevelopment, housing, racial and ethnic relations and natural resources. These were detected in the pattern of transitivities, and all fall into either amenities or the arbitration of conflicting interests as predicted. A careful inspection of the aggregate indicators for the various poli£; y types reveals the indicators themselves do not appear to differentiate well between structural policies. This may be because the empirical referents standing for economic growth policies are quite weak. Or it may be economic growth, while a legitimate conceptual concern, may not be easily investigated through a focus on council policies. Perhaps the community itself should be the focus. In any event, it is over matters

PAGE 424

405 of economic growth that the poslted ^^ ^^ ^ ^ ^ ^ its only difficulty. As : have specified| the reiation ^^ ^ costs and the pol icy types appears t „ be oversimpiif ied ,„ ^ fourfoid classification given. At least , the classif lcatlo „ ^ ^ ^ seated to hanale observations ^ ^ ^^ ^ ^^ ^ ^ formation. Pot e^ e , three ateas of policy atttact considerable Stoop activity when compared to citizen ot counci! interest in the same ateas. Ky observations on these issues catch toe™ before the, move into the UgfaXative ateoa. In . very real __ ^ ^^ ^ ^ adjusted to such a situation. Ihe classification scheme is constructed to dead with issnes which have been dealt with previously. The p „ sit ed approach tan mahe predictions about the fate of certain iss uea in the oil to demonstrate the adequacy of it assumptions. I" sun,, aggregating policy areas in the fashion of Table _9-6_ can be fundamentally fading. „h ile the approach is ^^ ^^ tMS aPPMl '"» fr °™ "= »*• ° f -ttera to which the researcher M y attend and not to the assumptions of the posited approach. P„ r the posited approach to be adequately tested, issues which have surfaced in the council and been afforded some interest most be investigated, m the ne*t section of this chapter, ! intend to inttoduce and analy.e just such a set of issues. A n_ Examination of Policy Types Each member inciuded in the survey was as k ed to name the most important matters of poHcy to come before his councii ln the last year. More than 40 different policies were mentioned. Tabie w presents in summary fashion the most prominent issues mentioned by council membets

PAGE 425

406 throughout the state. For each of the issues, members were asked to describe factors involved in the consideration of council action on the policy. Four classes of events were investigated. Following the example of Wayne Francis, I focused on the interest shown in the policy by various elements of the community, the conflict swirling around the policy, the resources involved in the proposal, and finally, the expected impact of the issue on the community. 34 This information together with the documented importance of each of the issues provides the means of testing the posited policy types. Consider Table 9-7. It reproduces the questions asked of each council member about the policy matters he identified as important. The first set of questions concerns the member's evaluation of the interest shown by people in the policy. Five types of interest were examined: 1. manager /mayoral 2. media 3. community group 4. council 5. citizen This represents a diverse group of interests which covers a range of potential important community forces. While the set of interests is not as specialized as those found in my earlier investigation of significant community forces, it nonetheless captures many of the influences. The second set of questions asked of all members deals with the presence of conflict over an issue and the site of that conflict. For example, it was asked if members perceived conflict between community groups over the identified issue or if disagreement had been fostered among the council members by the policy. These two particular questions can be used to address the fourfold classification charted by Salisbury and Heinz. In addition, though, I attempted to determine whether council members

PAGE 426

407

PAGE 427

3 3 a a o u

PAGE 428

409 associated conflict over policies with substantial involvement by the community or whether disagreement was perceived to have its genesis in the discontent of those least concerned with the community. These two questions were designed to gauge the attitudes of members toward the conflict which surfaces around policies and to determine the weight assigned to it. As a check on these responses, and in order to obuain a feel for the level of perceived community involvement in the issue, I also asked each member if there was little disagreement on the issue. Policies typically involve more than mere symbolic gestures to the public. They are characterized by the resources that must be marshalled in order for members to consider them in the council and the resources which they entail when allocated by the council. Technical matters and those which require the coordination of several agencies require greater resources than matters handled easily or through one agency. Moreover, if the argument of Cobb and Elder is correct, issues high on technical or agency components depress the amount of conflict swirling around an 35 xssue. When large amounts of money or great numbers of jobs are associated with a policy, the propensity for conflict over an issue increases. Issues that have never been raised before in a community should be associated with greater levels of conflict, and hence with policy types where conflict is more likely. When members are faced with the prospect of making long-term commitments in a given policy area, they are liable to be hesitant before they act. In the face of such policies, members will see future opportunities foreclosed, and thus the potential for disagreement is high among the members. Similarly, such commitments, when they are understood by the general community, are also likely to be the cause of considerable controversy. Realizing resources committed for such a per^ ^w ^ m ^m f^«fm ^ . imw f 9m m

PAGE 429

410 iod will likely interfere with other projects and other interests, these policies may cleave the locality into the many interests which characterize the political community. Each member was given the opportunity to comment on the potential impact of policies cited as important. The greater the impact associated, the more likely conflict is to develop over the issue. However, the impact of an issue should differ with policy type. Structural policies need not have a direct impact upon the community; rather allocation types will turn on such impacts. The high conflict policies will typically involve a change in contemporary procedures, but in many instances such policies are passed upon only to preempt even more dramatic policies. When a member indicates a proposed action is a marked departure from present practices, it can be expected his response will be similar to his comments concerning the precedence of the matter. But they need not be the same. A shift in practices can involve well worn concerns. In the same way, a member's report that the proposed actions might lead to more extensive actions in the future is not necessarily connected to his acknowledgement that the actions involve making a long-term community commitment. Nonetheless, there is likely to appear a relation between the types of policies involved in each instance. The thrust of these many questions is aimed toward cataloging a highly refined view of the interest associated with important community issues. They go beyond the requirements of the posited policy types, but supply the information necessary to test the fourfold classification. Using the specification I have drawn between conflict and costs, these questions augment the simple approach introduced earlier. Of course, once again I will rely upon the proportion of members registering their comments on -. *» '•**'>*'»«.* r '»'".»i i »*wjp^ *;*..-' «»»» l »»««y^

PAGE 430

411 specific statements in order to gauge the members' description of policies. However, I will not resort to standard scores to highlight my interpretation; the raw proportions will be used. 36 Do community policies labeled amenities differ from those falling within the rubric of traditional services? Table 9-T_ can be interpreted to suggest they do. Community groups are much more likely to contend over amenities than traditional services. Council members divide in similar fashion. If conflict does serve to illuminate costs, then those policies thought to be most costly do indeed evidence the most conflict on the measures most appropriate for checking the relationship. But the other measures frame the policies in more detail. Consider the attention groups and individuals in the community evidence in the two types of policies. Community groups and council members are palpably more concerned with amenities than traditional services. But this does not hold true for other interested influences in the community. The city manager, for example, is perceived by council members to be more enmeshed in the provision of traditional services. In fact, this complements the role defined for managers by council members, and dovetails with the International City Management Association's Code of Ethics. Much of a city manager's job is done as an administrative caretaker of well established policies. 37 The manager is not alone in his manifest concern for the provision of traditional services. According to the council members, this is the focus of media interest as well. Publicity accompanies traditional matters to a greater degree than attempts to begin new programs which would shift the priorities of the community. Perhaps it is the coverage of the newspapers and radio stations which accounts for the greater interest in tra-

PAGE 431

412 ditional services than amenities. However, the considerations featured in the posited policy types hint at another explanation. Namely, it is the divisibility of traditional services which enables them to be distributed to the public in such a fashion as to generate interest and foster expectations among the citizenry. Whatever the reason, Table _9-_7 makes it clear traditional policies, for the most part, are handled within conventional rules. They, more than any other policy type, are unlikely to spark disagreement. This, despite the fact such policies tend to involve more money, more jobs, and have greater direct impact on the community than any other type of issue. Instead, distributive policies are likely to have been considered by the council and the community many times in the past. Their impact, although large, is circumscribed; their ramifications limited. Council members do not believe policy making in a traditional matter is likely to lead to more extensive policies. Amenities exhibit a much different pattern. They are a source of considerable disagreement in the community. More frequently than traditional distributive policies, they involve a departure from current practices or matters new to the community. While the policies are said to be technical in nature, perhaps the most dramatic aspect of re-distributive policies is the commitment which they entail. Over 90 percent of all members commenting on policies termed amenities noted the length of the commitment associated with such matters. This feature of the policies together with the expected social significance of their impact promises potent political situations in which conflict is extremely likely and legislation is quite unlikely. Mitigating against the neglect of such policies by local councils is the widely shared conviction that the pro-

PAGE 432

413 vision of amenities, if delayed, would only foster even more drastic measures in the future. Conceptually, the relation between decision-making costs and external costs are quite predictable. Structural issues should be differentiated by the pattern of group activity they attract. A difficulty was noted earlier in the empirical indicators for policies of self-regulation. In theory, such matters do not appear before the city council. Rather, the costs they generate are such that members shun such matters. A council typically establishes the climate within which activities in the private sector are encouraged and nurtured. Thus, if such matters are to be given empirical referents, the focus must be on policies which act in this manner. In particular, matters of self-regulation are given substance in council policies extending or promoting economic growth. There is little difference between matters concerned with the arbitration of conflicting interests and the promotion of economic growth on the two measures most directly related to the fourfold classification of policy types. Council disagreement and conflict among groups in the community register levels of member agreement differing by a mere tenth of a percentage point. Both score very low on the amount of group disagreement reported, but over half the members indicated such matters provoked considerable disagreement within the council. Other differences between the two sets of policies are considerable despite the findings on these measures. For example, matters of economic growth attract much greater interest in all parts of the community than do regulatory matters. This pattern, together with the small amounts of group disagreement over economic growth suggests the potential for high costs, but a widely shared support for such policies. From the city man-

PAGE 433

414 oger to the average clti2 en, there is strong lnterest lp ^^ ^ -«ers when „ P a reJ to regulatory onea . ^ ^ ^ ^^^ ^ members report economic growth eener a tPO k g owcn generates substantial community wide disagreement. Perhaps members perceive thic ^ perceive this disagreement coming not from the clash of organized community erouDs hut f™ • j . , 7 groups, but from individual citizens. They suggest this when they report the belief that substantial disagree „ eiu over the mstter has its orlgins in citlzens ^^ ^ ^ ^ ^ C °'" Ity C ° ntrary * "« *""— • e x tem a l costs do not appMr h^h. Instead, external costs merely _ the ^^ fa ^ ^ Decisic„Mk i ng costs „ regula£Ive Mtters ^^^^ ^ arwt ^ "on of conflicting interests should be high, uhile the patter „ pf ^ nal contacting c a n be ejected to be Rented and thns result in low »sts. This is the pattern discerned on the two « sms being US ed as the primary test of fourfold ossification. In fact, an agination of the severe! measures bearing on c omu „ity lnterest an d demonstrated convict within the community paInts . consistent ^^ ^ ^^ or conflicting intere^tc ,„•«.!„• ,. g terests withrn the community drews fragmented attention from groups and citizens outside the council. As opposed to issues of economic growth, regulatory matter, are unlikely to involve severe! different agencies. Nor do they involve Urge numbers of Jobs or substentinl .mounts of money in the way economic g rowth rssues do. However, matters involving the arbitretion of conflicting fntarests reflect the intervention of local government into otters hitherto beyond its purview. Kelntively large proportions of council tors reported such regnlntory matters bed never been raised in the co„»unity before, or that they represented a marked departure from present practices. Moreover, for most council members, otters involving the mem-

PAGE 434

415 arbitration of conflicting interests are much more likely than economic growth issues to represent policies which are necessary in order to forestall more dramatic policies. In the discussion of the fourfold classification scheme advanced by Salisbury and Heinz, I have considered, in turn, allocative and structural policies. Considered in this fashion, there is a good correspondence between the policy types and the costs associated with the policies. Only self-regulative policies or matters concerned with economic growth do not squarely meet our expectations. Interest in the community is high on these matters, but the pattern appears more fragmented than expected. What happens if the figures in Table 9-7 are examined not two by two but as a set which bears upon the fourfold classification of policy? In this instance, the absolute value of the proportions would have to be compared in order to test the hypotheses. A cursory inspection of Table 9-7 reveals this can be done without great harm to the findings. For example, the decision-making costs associated with the structural policies are lower than those associated with allocative policies when disagreement among the council members is taken as the indicator. Moreover, if we accept the notion re-distributive policies will emerge from councils only rarely because the dec is ionmaking costs associated with them are likely to be high, then our other expectations are borne out. Traditional services carry low decision-making costs; each of the other types have high costs. Summary This chapter has featured an extended discussion of local policy and its treatment in city councils in Florida. But it has yielded no irrevocable findings. Instead, this chapter argues the determination of

PAGE 435

416 public policy is a complex business, elusive of parsimonious formulation. Nonetheless, several tendencies have surfaced during the course of the argument. Over a substantial range of issues, council members are contacted by individual citizens for the redress of problems. However, this form of citizen-initiated contact does not have nearly the consequence in the council chamber that group-initiated contacts do. As a general proposition, group actions appear to spark disagreement among council members on matters of policy. Although the relation between group actions and council disagreement over issues is positive, for any specific issue the link between the two is likely to be quite weak. Among all Florida councils, it is only over such matters as racial and ethnic relations or parks and recreation that the connection is substantial. This, of course, is hardly surprising. Councils and the issues they face differ slightly throughout the state, and hence a systematic link between group actions and disagreement over specific issues is not to be expected. Rather, it is the pervasive influence of group actions on council behavior which should be highlighted. Group actions appear able to create substantial external decision-making costs. Some policies appear to divide councils more than others; some policies galvanize community groups or citizens to action, others do not. Group theory attempts to explain why this is the case by appealing to the analytical notion of interest. In the works of many group theorists, however, this term subsumes manifold aspects of a choice situation. Exchange theory makes use of the concept of interest by focusing on the pattern of costs attendant to each policy. These costs are functions of an individual's interests.

PAGE 436

417 In this chapter, I have argued interests can be viewed from two different perspectives. Each perspective determines a different research agenda. Each countenances a different set of findings. When interests are taken as revealed in the behavioral manifestations of council actions, planning and area growth, public welfare, municipal employees, and the form of city government are identified as policies generating substantial costs for Florida council members. Yet these policies do not delimit the set of policies hypothesized to generate such costs when a set of objective interests are posited for each of the Florida council members. What this analysis reveals is that over a broad range of issues, selected without regard for their salience to council members, the hypothesis offered by Lowi does not hold. Instead, the pattern is much more complex. Lowi's hypothesized relation between politics and policy holds much better when only matters of great importance in each community are commented upon by council members. Here the posited classification of policy bears up well. Amenities or redistributive policies generate disagreement among members and concern in the community. Traditional services are unlikely to attract such attention. Thus for matters involving allocation Lowi's formulation does quite well. Where Lowi's scheme falters is among those sets of policies deemed self-regulation by Salisbury and Heinz. Apparently matters of economic growth, typically left to the private sector, do not evidence the same systematic form of treatment in councils and communities throughout Florida. In summary, the posited classification of policy bears up well under a scrutiny that focuses on a member's own evaluation of important issues. Each of the policy areas corresponds to the posited costs, and is the outcome of the interaction of external and decision-making costs,

PAGE 437

418 in the prescribed fashion. Hence the policy matters included in each of the types form sets of issues which can be used for further research. More specifically, the policy types can now be used to investigate the politics which swirl around the issues in particular councils. In the next chapter, I use the policy types to draw the relation between cities known to have expenditures in certain areas (see Chapter four) and the type of politics which characterize such cities.

PAGE 438

419 Notes 1 Jay S. Goodman, The Dynamics of Urban Government and Politics (New York: Macmillan, 1975), pp. 209-219. 2 John Kirlin and Steve Erie, "The Study of City Governance and Public Policy Making: A Critical Appraisal," Warner Modular Publications Reprint (Andover, Mass.: Warner Modular Publications, 1973), p. 3. 3 Sidney Verba and Norman Nie, Participation in America: Political Democracy and Social Equality (New York: Harper and Row, 1972), pp. xix-xx. 4

PAGE 439

420 14 Ibid . , p. 255. 15 Quoted in ibid . , p. 257. 16 William Connolly, The Terms of Political Discourse (Lexington, Mass.: D.C. Heath, 1974), pp. 45-84. 17 William Riker and Peter Ordeshook, An Introduction to Poljtive Polit ical Theory (Englewood Cliffs, N.J. : Prentice-Hall, 1973), p. 14. 18 Connolly, op. cit ., p. 48. 19 Ibid . 20 See Heinz Eulau, "Logic of Rationality in Unanimous Decision-Making," Rational Decision , Vol. VII, Nomos , ed. Carl Friedrich (New York: Atherton, 1967). 21 Connolly, op. cit . , p. 53. 22 Ibid . , p. 59. 23 There is some evidence that this is the case. See, for example, M. Kent Jennings and Harmon Zeigler, "The Salience of American State Politics," American Political Science Review , Vol. LXIV (June, 1970), pp. 257-285. 24 James Buchanan and Gordon Tullock, The Calculus of Consent: Logical Foundations of Constitutional Democracy (Ann Arbor: University of Michigan, 1962), p. 45. 25 Robert Salisbury and John Heinz, "A Theory of Policy Analysis and Some Preliminary Applications," Policy Analysis in Political Science , ed. Ira Sharkansky (Chicago: Markham, 1970). 26 Goodman, op. cit . , p. 117. 27 Ibid . , p. 120. 28 Ibid . , p. 188. 29 Ibid . , p. 118. 30 For each policy area and across the three measures, I calculated a standard score. The mean and standard deviation were computed using the sixtynine data points. The measures were aggregated in this fashion to allow for propensity of members to respond in a limited number of areas. Thus each policy area on each measure could be gauged against the members' total response. This technique advances the frequency with which citizen-initiated activities are seen to occur since the average proportion of members reporting citizen activity is greater than for the other two measures. On the other hand, council disagreement will be depressed. These disadvantages are outweighed by the common standard which marks each measure in each policy area. Standard comparisons can be made; simple binary operations can be performed.

PAGE 440

421 31 I assume there is no possibility of ties in the data; only two states are allowed. 32 The difference is .011 as compared to .082 between citizen and group contacting. 33 Environment and Florida Voters , work paper. No. 7, prepared by Cambridge Survey Research Institute for Urban and Regional Development Center (Gainesville, Fla.: Urban and Regional Development Center, 1974), p. 3. 34 Wayne W. Francis, Legislative Issues in the Fifty States: A Comparative Analysis (Chicago: Markham, 1967), pp. 12-14. 35 Roger Cobb and Charles Elder, Participation in American Politics : T he Dynamics of Agenda-Building (Boston: Allyn and Bacon, 1972) pp. 112-124. 36 Standard scores were used previously to help overcome the response set associated with the format of the questions on council disagreement, and contacting in the twenty-three areas. The possibility for similar problems among these questions appears remote. Rather than reading down a list of items whose reference was continually changing, these questions have a constant referent. Members who began the question finished it. If an item was overlooked, it was recorded as missing data, and not a negative response as in the first set of items. 37 Arthur W. Bromage, Urban Policy Making: The Council-Manager Partnership (Chicago: Public Administration Service, 1970), p. 9. 38 Council members indicated most amenities cited did in fact pass the council, but only after considerable modification. This softens somewhat the claim decisionmaking costs associated with re-distributive issues are ineluctably high. According to Salisbury and Heinz, re-distributive policies should obtain council approval when decision making costs are low. High costs, it was assumed, should obtain when the council is stymied. Empirically, it looks as though the cost reduction of bargaining can lower the decision making costs to a level that sustains the passage of re-distributive policies. It well may be council decision modes, such as consensus and informality, can be evoked to handle such situations. However, I have no evidence demonstrating this is indeed the case.

PAGE 441

CHAPTER TEN POLITICS AND POLICY There are more than 70 cities in Florida having populations greater than 10,000. They vary in any number of ways, yet many social observers would suggest the differences are not so great as to disguise the similarities. This supposition stands behind my presumption in chapter four to divide the Florida cities into six groups on the basis of their socioeconomic similarities. However, cities that are roughly identical in their social and economic compositions may not have similar policy priorities. As was shown in chapter four, the spending priorities of cities sharing relatively homogeneous environments can differ considerably. This was the case among the high SES cities of Florida, among those cities which formed group F in my earlier classification. Why might cities with similar socioeconomic environments spend their revenues differently? The systems model presented briefly in chapter three puts the onus on political variables when the environment is held constant. And in chapter four, I speculated that differences in political characteristics might account for the differences in the spending patterns of the high SES cities. In this chapter, I will marshall additional evidence to buttress this speculation. My comments will center on the relationship sketched in Figure 10-1 . I will use the policy types developed in the last chapter and the political features of the councils and communities of group F explored throughout this study to examine the relationship. 422

PAGE 442

423 *********************** ***** [ Environment ] ***** *********************** ****** [ Controlled] ***** a. 4. jljju 4. 4. 4. jJa. ju 4. a. 4. 44. 4. -a. jl 4. 4. .v. y\ /C /\ *\ ** /\ rt <>. /v /* A a A rt <^ A *C 75 ^ /v j-\ /\ ** Political Features of Council and Community Classified Policy and Expenditure Types Figure 10-1. Systems Model of Policy Process Micro-Phase Indicators

PAGE 443

424 Policy Types and Expenditure Groups in the High SES Cities In chapter four, three distinct patterns of expenditures were detected among the 12 cities of group F. Their spending priorities can be characterized with respect to one another on the basis of the spending factors which best differentiated the cities (see Chapter four). Five of the cities marshall their funds for matters of traditional concerns such as fire protection, police activities, and refuse collection. Service expenditures act to discriminate the cities of: Bartow Cape Coral Cocoa Beach Ft. Walton Beach Rockledge Public transportation and urban renewal expenditures are likely to characterize the expenditure directions of the councils in the second group of cities. The common aspect of expenditures in the following cities is their relative focus on newer public works : Gainesville Tallahassee The pattern of expenditures in the remaining cities of group F is likely to constitute a mix of the previous patterns. Nonetheless, expenditures in the following cities form a distinctive pattern of their own: Boca Raton Cocoa Hialeah Melbourne Titusville Classification procedures document differences in spending priorities among the high SES cities. But they do so automatically. That is, it is unlikely a set of data can be subjected to their operation and remain undifferentiated. Puddles of variance are pushed together, and the

PAGE 444

425 result is interpreted. But sense must be imputed to the findings even when the statistical evidence is overwhelming. In the instance of spending groups, however, there is little theoretical reason sustaining the empirically defined groups. The three patterns of expenditures may well be the most ephemeral form of construct, one without basis or use. The doubts lingering about the notion that similar socioeconomic cities may pursue quite different policies can be addressed by an appeal to the policy types considered previously. This is possible because of the fortuitous circumstance surrounding the discrimination of the expenditure policies of the group F cities. This group of high socioeconomic cities has been differentiated by factors which bear considerable resemblence to a pair of the policy types, namely, traditional services and amenities. While the fit is not perfect, the correspondence between the factors used to group the cities and those involved in the policy types appears close enough in order to use the policy types as a means of substantiating the proposition. This can be done in a fashion similar to the test involving the policy types themselves. First, the expenditure patterns must be accepted as descriptive of an actual difference in the policies of the cities. Then the policy types and the pattern of costs with which they are associated can be used to verify the distinction. Consider the figures arrayed in Table 10^1. They represent the proportion of members reporting the presence of considerable council disagreement, citizen contacting, and group actions on issues falling within the bounds of the four policy types. The proportions are the average response of members, on each policy type, from the cities noted to have distinctive spending priorities. Service expenditure cities, it can be argued, should be more oriented toward traditional policies than ameni-

PAGE 445

426 Table 10-1 Average Proportion of Members Reporting Council Disagreement, Citizen Contacts, and Group Actions by Policy Type in Each Expenditure Group of High SES Cities (Group F) Expenditure Group A — Service Expenditures Council Disagreement Citizen Contacts Group Actions Traditional Economic Services Amenities Growth Arbitrating Between Conflicting Interests .300

PAGE 446

427 ties. And indeed this is what Table 10-1 shows. Council disagreement over traditional service is greater than over amenities, despite the general, tendency for amenities to spark such disagreement. For councils in cities such as Bartow, it is traditional matters which remain the center of the legislative arena. But this does not mean traditional policies involve high costs in such cities, only that with respect to other issues considered in such contexts traditional issues are matters of contention. If the level of council disagreement in the service expenditure cities is. compared with that in the other expenditure spending groups, it can be seen council disagreement is lower in service expenditure cities. An inspection of the pattern of contacting in the service expenditure cities re-enforces the perception they are indeed concerned with traditional policies. Group actions, to the degree they exist, are higher over such matters than any of the other policies. Once again this flies in the face of the experience of this indicator. Group actions are typically higher over amenities or matters relating to the arbitration of conflicting interests. Similarly, citizeninitiated contacts over traditional matters outstrips such activity over any other policy matter in the service expenditure cities. From all appearances, the classification procedure has worked well in this instance. The remaining expenditure spending groups share similar patterns on the policy types across the three measures. Amenities attract relatively greater amounts of disagreement among the council members than traditional services. Group actions are reported to be linked more frequently with amenities than traditional services. In both groups, citizen-initiated activity is much greater on matters dealing with traditional services. The difference in the two groups lies in the level of

PAGE 447

428 such activities. Cities such as Gainesville and Tallahassee evidence much higher rates of council disagreement and community activity than do cities such as Hialeah and Melbourne. Where the spending priority is directed toward newer public works, amenities are indeed the policies which draw the most attention. However, this pattern of expenditures is differentiated from the more balanced spending pattern by the levels of political activity. This difference extends to other policy matters as well. Matters involving the promotion of economic growth or the arbitration between conflicting interests are more likely to be seen as conflictual or attracting citizen activity in the cities where newer public works have an emphasis. In short, while the classification procedure differentiates the two groups' spending habits, the measures of policy do not suggest a significant difference in the types of policies pursued but in the political activity of the two sets of cities. Thus, one is tempted to speculate, if there is a difference in the policy orientations of the cities of the two groups, it may have its origins in the political character of the cities themselves. Perhaps cities which emphasize amenities do so because the climate of attention to such policies, and to local government in general, is much higher than in other communities. There is an additional aspectof Table 10-1 which might be noted. In each of the expenditure groups, the level of reported activity on matters of economic growth is exceedingly low. Council disagreement is lowest on such matters in cities with service expenditure policies. Citizeninitiated contacts and group actions, however, are more frequent in cities such as Bartow than in cities displaying a more balanced expenditure Profile. Although the levels are low in either instance, it may be that

PAGE 448

429 external contacting of members occurs in the service expenditure cities in order to foster such policies. This would account for the slight amounts of council conflict despite the higher levels of citizen and group activity. In the balanced expenditure cities, group actions are practically nonexistent. Even individual contacting falls below that reported for group actions in the service expenditure cities. Once again, the cause may only be speculated upon. Perhaps issues of economic growth do not occur in these cities. Alternatively, the pattern of community support for such matters may be in the process of change. Community groups and citizens may not have crystallized their interests in what appears to be a changing political climate. This is yet another possibility. That is, as suggested before, the political character of the cities themselves may be such that political activity is depressed. This view finds some support in the generally low state of group activity levels in the balanced expenditure cities as compared to those in the cities focusing on service. Group actions are significant in cities such as Boca Raton only with respect to amenities. Regardless of the status of these speculations, it does appear the three spending groups are well formed. They differ on the three measures of council disagreement and community contacting in the manner expected. However, it is possible to check this finding with an appeal to the member's more detailed report on issues. In particular, by examining each member's description of the politics surrounding a particular policy, a sense of the costs associated with each policy can be purchased. The policy mentions of each member can be grouped according to the policy types constructed previously, and the performance of each spending group can be examined on the indicators chosen.

PAGE 449

4 30 The members were asked to indicate, for each policy mentioned, whether community interest groups were at odds over the policy and if there was considerable disagreement among the councilmen over the issue. Indeed these two questions were shown earlier to be fairly reliable referents of the costs involved in a policy. A cross-tabulation of these two questions provides an excellent means of examining the costs associated with each policy. Of course, these two indicators alone cannot be used to frame the conceptualization of Salisbury and Heinz. Nonetheless, it can be expected high cost policies such as amenities will involve high amounts of council disagreement and significant group conflict. Traditional services, on the other hand, should display low levels on the two indicators. There Is a slight difflculty „ ith . tabuIar prese „ tatlon> ^^ Consider Table 10-2; lt presents . cross _ tabulations „ f ^ ^.^ with council disagreement. Table 10-2 SES Group F Council Members' Characterization of the Political Aspects of Traditional Issu es Council Disagreement LOW HIGH HIGH 10 23 Group Conflict LOW 20 12 This tabulation arrays the manner in which the members of the councils in the high SES cities of group F report the activity surrounding the consideration of policies involving traditional services. Inspection of the table reveals that for those particular policies upon which the members

PAGE 450

431 of group F reported, substantial costs were involved despite the fact traditional services were considered. But suppose the table were not available for inspection. How could it be determined such costs were involved? Clearly, conventional associational statistics do not portray the distribution between the matrix presented and one in which the numbers on the positive diagonal had been changed to five and ninety-two. Rather than presenting a series of tabulations, I used a vector notation to capture the distribution in the matrix. This was done by assuming the figures residing in each cell can be taken as the magnitude of a vector with its origin in the center of the matrix and thrusting out through the center of each cell. Then the vectors were summed; the resultant is descriptive of the distribution. For example, in a matrix where each cell was the same, the resultant would be the null vector. As one cell dominates the array, however, the magnitude and direction of the resultant favor that cell. If, in the table above> aU 65 ^.^ had fallen in the high disagreement, high conflict cell, then the resultant vector would have had a magnitude of 65 and a direction of 45 degrees Vectors assuming large magnitudes and approaching the mid-portions of the four quadrants dominate the matrix. The mid-points of each cell lie at 45, 135, 225, and 315 degrees respectively. 1 Table 10-3 lists the results of the vector analysis. Consider the figures capturing the distribution of the tabulation involving council disagreement and group conflict for the high SES cities. The magnitude of the vector (3.61) and its direction (11.32) are representative of the cross-tabulation illustrated above. The small magnitude of the vector and its narrow angle with the council disagreement axis are indicative of the minor dominance of the high cost cell in this relation. Compare

PAGE 451

432 Table 10-3 Vector Analysis of Cross-Tabulation of Important Issues Involving Council m! Croup Co„fl ict in C^tieTof £0"?™ "* Magnitude Direction of Vector of Vector ^roup_F (Degrees) Traditional Services (N = 65) 3 g. Amenities / w . ,-.,* (N 63) 31 . 95 Arbitrating Between Conflicting Interests (N = 24) 10 . 20 Traditional Servi ices Amenities Spending Group A Spending Group B Spending Group C 11.32 24.88 Economic Growth (N = 6) 3 -,c ' 3 16 26.58 326.31 Spending Group A ( N = 2 5) a n* ; 8 -°6 262.88 Spending Group B ( N = 29) Spending Group C ( N = n) 9-22 33.49 2 24 18.44 (N = 26)

PAGE 452

433 this with the vector formed with respect to members* evaluation of amenities. Its magnitude gives testimony to the highly charged political environment enveloping such policies. However, the direction of the vector (24.88 degrees) is swung away from the diagonal. This indicates the political posturing over amenities is not uniformly contentious. Often council disagreement will be attended by little conflict among the community groups. A similar statement can be made about group conflict over matters of economic growth; occasionally members will detect very little activity on such matters despite their own intense interest. Among the ambers of the high SES cities, there is a marginal tendency for economic matters to be associated with high costs. This observation must be tendered with a caveat, however. Only six such policies were named by the council members of the 12 cities as being among the most important issues to come before the council in 1974. A similar warning need not accompany an examination of matters involving the arbitration of conflicting interests. Here the number of policies is relatively large. The vector of the interaction between group conflict and council disagreement is substantial, but, unlike any of the previous cases, falls in the fourth quadrant of the table. This connotes high levels of council disagreement over regulatory matters in the absence of significant group conflict. Once again, regulatory affairs appear to be sustained by council interest. Community groups shun such matters for the most part. The most surprising aspect of the political processes accompanying .policy making in Florida cities with the most ample socioeconomic endowment is the manner in which traditional services appear to nuture high costs. This flies in the face of the typical relations found among Flor-

PAGE 453

434 id a city councils. Traditional services are normally accompanied by little disagreement and marginal group activity. This riddle is more feigned than real, however. If the cities of group F do manifest distinct patterns of policy priorities, then the politics of the councils over policies may differ as well. The figures of Table 10-3 testify to this proposition. Spending group A, which includes cities such as Cape Coral and Rockledge, are more oriented toward service expenditures than the other cities of the high SES group. And the council members of spending group A report traditional services typically involve little council disagreement or group conflict. The vector analysis shows most cases concerning traditional service policies fall into the low-cost category (262.83 degrees). In the remaining spending groups, the situation is quite different. Here, even traditional services are likely to incur high costs. It is the cities with the most balanced expenditure profile which experience the highest costs on traditional services. The magnitude and direction of the vector summarizing the distribution of costs associated with traditional services shows that cities which emphasize newer public work expenditures have councils which report costs on these issues may differ considerably from case to case. There is only a marginal propensity for high costs. The magnitude of the vector associated with the provision of amenities in the cities with a balanced expenditure profile is quite large. This, taken by itself, indicates numerous instances in which issues over amenities were accompanied with high costs. The direction of the vector, on the other hand, suggests there were many situations in which the political events enveloping such issues were quite different. In particular, group activity often failed to materialize on issues in which council

PAGE 454

435 disagreement was substantial. This forces the swing of the vector away from the diagonal in the high cost cell (24.77 degrees). In cities emphasizing newer public works, this was not the case. No clear pattern emerges between low levels of either council disagreement or group conflict. However, the dominance of high cost situations is clear in high SES cities on matters involving amenities. Only the small number of cases involved (n = 7) lends caution to this observation. Group conflict and council disagreement is most substantial and most likely on amenities in those cities spending in a manner to create newer public works such as urban redevelopment. It should be noted service expenditure councils also disagree over amenities. But in those cities, members are much more likely to register the absence of significant group activity attending to such issues. The classification of cities by spending priorities does appear to make distinctions which capture more pervasive phenomena. Namely, there do appear to be differences in the policies pursued by the councils in the three groups. Cities such as Bartow which tend to spend their revenue on service are, in general, more interested in traditional service policies. Amenities are more likely to characterize council policy in Gainesville and Tallahassee than the other ten cities. Political Features of the High SES Cities and Their Policies There is more to the difference between these cities than the policies they pursue. The politics appear to be different as well. And, in fact, the politics seem to fall into patterns much like the policies themselves. Perhaps it is, as Lowi argued, that policy does influence politics. It might also be that the political climate in a community can have substantial impact on council members and their behavior.

PAGE 455

436 »— 1 ? , u ,, w possible lo tMt the direcuon Qf the ^ lation using the Hat-i t v.., g cne data I have gathered. Nonetheless T r * n ,, ess ' i can draw the asso». — • w tendencies have already ^^ ^^^ ^ ^ where amenities are feafi.r^ featured, citxzen-contacting and Wr ng and group actions are reported with considerable frequency m all rt i , SS ' f ° r each P^icy type the level of activity in these cities exceeds that ril . S that "Ported among counS1 "<— -> °" — eroup activity in — 3 „ ith . balMced expMditure patternexMMtsdepressed ^^ group activity. These levels fall u fall beneath the average for the sample of members at laree r-r*-, ^rge. Cxtxes where traditional services tend to d , . Ll t;t3 cen d to dominate the policy-asking concerns of the local c levels of man " eSt eXC "«»ely low levels of community contaci-ino n tivitv • Gr ° UP aCti °-d citizen-initiated activity are infrequent i n such cities TM • 6S ThlS Vlew is supported not only -en evasions of fc „ ^ _ ^ ^ ^ ^ : traditionai -— — — y _ insulated 11010 concerned citizen an A and comu „i ty groups than ciues devoted ^ "— y paths . c „ y groups Md cuizens are ^^ —a 1 si tuat ions fa V o ri „ 8 the provislo „ „ f ^^ « — * It is Mt just ^ co _ ity ^^ "" di —— * — „ t s p e„o ing prlorItias . ™ «— ~ l,el y t0 __ ,„_ ntati0 " t0 " 3rd tr ——U acco.anfeo by Uttle dIsag _ — — . «U _ ber s. Bue „ here aM „ lties are likelv i-„ aconcern, members are y — — — , i ndead , even among those ^^ «-, a oalancea e„ re ,„,„., „ ^^ ^ ^

PAGE 456

437 stantial. In fact, the disagreement reported among members in both the balanced expenditure and amenities-oriented cities typically exceeds the general levels of disagreement registered by council members throughout the state. Table 10-4 Average Proportion of Members Reporting Council Disagreement, Citizen Contacts, and Group Actions by Policy Type for All Council Members and Those in SES Group F

PAGE 457

438 Table 10-5 In Each Exnp r H 8e ^^ ° f SeleCted Indic ^ Each Expenditure Group of High SES Cities (Group F) Index/Variable CONSPOL CONSREL PEOPLE BUSINESS HUMP ROB GOVPROB GROWPROB FINPROB SERPROB CONSPOL Differences CONSREL Differences Decision Modes Generalized Splits Idiosyncratic Splits Particularized Splits Logrolling Amicable Agreement Consensual Informal Service Expenditures Expenditure Group B Service/Newer Public Works Expend itures 10.40 8.32 9.36 18.38 6.97 5.51 6.47 8.00 3.35 3.17 3.24 2.13 0.32 0.93 5.04 2.38 4.34 3.88 7.45 4.91 10.30 17.79 8.10 4.01 8.79 8.85 4.09 4.49 3.65 2.40 0.27 0.73 4.74 2.20 3.64 3.70 C Newer Public Works Expenditu res 4.68 6.70 11.00 16.63 9.75 4.50 7.60 11.50 3.90 4.19 4.23 2.35 0.20 1.00 5.30 2.68 3.33 3.22

PAGE 458

439 One of the *,,t dramatic features distinguishing the councils of the high SES cities is the political predispositions of their me.bers. Councils accented h y a concert for amenities are substantial!, less conservative than those promoting traditional services. xhe y are less conservative than cities pursuing a balanced expenditure policv. But the counoils stressing amenities do not have .e.hers whose religious funda.entalis. is extremely low. gather the least levels of frontalis, occur in the balanced expenditure cities. Keme.ber religious fundamentalis. is t yP icall y associated with the .ore consensual for.s of decision maUng, „ ith strong friendsMp patterns _ ng ^ ^^ ^ ^ the absence of polariaation among the .e.bers. It is the cities accentuated by their provision of traditional services which displa, the highest score on the religious fundamental™ index. Their conservatism is overwhelming. Moreover, an inspection of the differences a.ong the _ hers aiong the ideological spectre, shows these councils are q uite coherent. Meters do not differ ideological!, to the extent seen in counoils pursuing less traditional policies. Another interesting difference between councils pursuing different policies is the groups within the co M it y to who. the y ascribe influence. Business groups such as the cha.her of commerce and local ban.ers sre evamated most lo„l y . Helfare organizations, and minority and ethnic groups are seen as having the most impact on decisions in cities more concerned with providing amenities. And while business groups are alwa ys seen as more influential r-*,,., nrtuentral than people-oriented groups, councils which pursue a.enit y policies report business groups have relatival, less influence in their communities than do councils in the other spending groups.

PAGE 459

44 o Not surprisingly. cou „ cUs classed by the higMights ^ ^ a "" ItJ P ° UCieS ™ " 1 '° "»" «-=«. *« human resource problMS ™ -« often r eported . C0U „ CI1 _ bers ^ these comunities ^ ^ 1—7 f suggest housing, poverty and educafIon are serious probi ^ than members in eithp-r n f <-u„ of the two other spending groups. TradltIonal ser _ vice cities, on the other hand, are the least Ifb, t east "«ly to have council members pointin g out such difficulties h„ trcultres. However, council members in amenityo-e„ted cities do suggest their reception of human resource prohlems "mpered * a prudent concern over the source of revenue for such promts. Prohlems with the city's tax base and g over„ment finance are ci« "ore frequently amon g council members who observe amenities as imporP-ern of expenditures are the most li ke ly to cite growth related prob— such as environmental concern and area growth. Public „tili tIes are -~ Sreat concern to such councils. In £act , servlce ..^ _ 1— Prevalent M on g those councils whose spending priorities are _ Seared toward service provision. This might he because members feel as — h the problems have been properly addressed. A lternatively, it can * nested this particular index misses many of the concerns included under traditional senrirpc o servxces. Servxce problems are identified as matters involving public nt--M -;«..* utrlrtres and recreation and entertainment. This is considerably more narrow than the set of issues listed under traditional services. Th e traditional service cities do rate bi g hly on the governmental problem index which includes problems of transportation and trafXbe problem concerns of the various conncii groups act to confirm -e again, their classification into the distinct spending groups. But

PAGE 460

441 there is yet another perspective to be purchased o„ this classification; nameiy, the decision-mahing modes osed „st frequently in che councils. Traditional service provision shonld he accompanied by llttle conflict. The modes used to make such policy can h« Poxicy can be expected to be consensual. In Particular, the low conflict levels shonld he indicative of consensual modes hased on interaction and personal relations between the M bers na„el y consensnal and informal decision-.ahing modes. The provision of unities, on the other hand, can be ejected to e e„erate si g nificant disagreement a„ e the council _bers. The generated split n ode or the consensual .odes calling f or bargaining or amicable agreement should accompany amenities. The balanced expenditure pattern cities can be exPentad to have significant levels of council disagreement. As with the -enity-oriented cities, the more generated manner of council disagreement or the bargaining consensual modes should be used. inspection of Table 4-5 reveals several of these expectations are fulfilled, m particular, traditional services predominate in cities "nose councils fluently adopt the interaction based consensual decisionmaking .odes, mformal agreement and unstated consensus over the issues typify the approach of such councils, w.en spllts do occur , . comparison of the city types suggest the reasons may be idiosyncratic. Disagreement in cities featuring sanities are more than lixely to occur over Particular issues, in „ost instances, these councils resort to agreement -das which call upon members to worh out solution through arable agreement or logroiling. Balanced expenditure cities score higher on the generated spUt m ode than either of the other two groups. However, there is so„e evidence to suggest the cities in this group vary considerably «. the modes they adopt. For example , ^ ^^ ^ ^ ^ ^^_

PAGE 461

442 ~tel, the same score for this group of citIes „ for the group ^^ sizing service expenditures. In any event, it is difficult to reconcile the use of generalized splits and Infernal decision-making modes. In sum, each of the different spending groups evidences not only different relations with the community b „t also different politics within the councils themselves. The more traditional service policies are promulgated hy councils whose members share common and conservative attitudes. These members are unlikely to engage in frequent interpersonal conflict. Rather consensual deicison making, based on perceptions of friendship or a rejection of conflictual politics, u the rule. Amenities, on the other hand, are nurtured by councils where conservatism is less evident. Here people-oriented groups can rival business concerns in influence. But the council members do not systematically disagree. Disagreements tend to occur on particular issues; agreement is achieved not given. Members work toward solutions. Councils such as Hialeab pursue spending policies which hold the middle ground between amenities and traditional services. And on most measures of council characteristics, they rate between the poles identified for the other two groups. Membars are neither conservative nor liberal; business groups are influential, hut do not totally dominate t^ber perceptions. But disagreement a-ong the members is high; generalized splits occur more fluently than in other types of councils. Perhaps because they have relatively balanced expenditure patterns, the propensity to reallocate resources is ever present. Members may always see the possibility to change; political rows capture the tension. Summary Cities sharing relatively similar socioeconomic characteristics

PAGE 462

443 -7 Pursue dl « erent policy prlor±tles _ & ^ ^^ £ ^ ^_ firmed the analysis of city expenditures presented in chapter four. A „ equation of the „ high SES citIes , )as ^ ^ ^^ ^^ differ in a way which complements their ranee of „„i • Liieir range of policy concerns and the most important issues to surface in their communities The politics of socioeconomically similar cities differ and they have different effects. A mon 8 the hi g h SES cities in F l or ida, those which pursue relatively more traditional, service-oriented policies have councils whose memhers share co„o„ conservative attitudes and who are unlixely to engage in f recent interpersonal conflict. Cities whose spending pollcIes feature . ^^ ^^ ^ _^ ^ ^^ with more liberal members; people-oriented groups are „„„ &i.uups are more important in — t, decision MK i„ 8; and there is a tendency for members to wor k "ward agreement on policy matters through bargaining or extensive disc-ion. toong these clties _ the poi±tics ^ ^^ councii ^ ^ ^ heavily on the policles Kh . ch . t promulgate ^ mm*

PAGE 463

444 Notes vector ^jSStt ^".SS^-^ctta, of the resnltant ~£ pitr:!^s&~ *-«--•peczally amenable for two-by-two t.W.. ? g c °llnea rThis fort, is esnegative. Only the fourfold ",° 8 " eS uhere "™e of the entries are = A x + B 2 + c^ + D/ The magnitude of the results,< • vectors and using the nvth glVe " by merel y adding the m i g the Pythagorean theorem to obtain a result^ N 2 |A 2 + c 3 + |B 2 +d 4 |2 orientation,, the resnltant „ as obtain. osin g the oosine la „: + r2 ~ [B n + D I 2 A l + Co R = inlhfre^se'We^f rc^r"" " e »«", ana I of the co „ resnltant in J?A 3 «= .This angle can then be ,l! ""f" 631 axes > «saltantl„s^^VS"/^ »^"tteTbe^ ^^^ tahe the ^uS'oSii ais^ ^'^ -ventio" •*-*««

PAGE 464

CHAPTER ELEVEN SALVAGING THE POLITICAL SYSTEM Consider the city councih In „any co„„unities, it Is ^ ^ ^ get of electoral activity, often, the local council is the only for™, however frail, counity groups have for their claims and q ueries. Councils bedevil city onagers, they frustrate mayors, they bell™ and gesture toward local bureaucracies. Yet, for the attention councils have received in the discussion of urban politics, they nnght wei l be thistledown amidst a strong wind. In this study, I have attempted to piece together an image of city councils which casts them in a more resolute role. This does not mean councils determine municipal policy to the exclusion of other powerful agents within the community. Nor 3 r P t-w, „~ •-,,-. nicy. wor Q re they necessarily first among equals. These contentions are beside the point of my inquiry. I have attempted to give councils a political role within the city and council members a. part in making policy. This chapter will examine the argument I have made in this study. It will tender several speculative propositions which can be drawn from the work. And it will suggest how the study of public policy can be tied into the normative concerns which have traditionally animated the study of politics. The Cit y Council in Policy Making "Policy," says Heinz Eulau, "is a strictly theoretical construct that is inferred from the patterns of relevant choice behavior by poli445

PAGE 465

tical actors and the consciences of choice behavior."! In „ uch of the recent policy research, budgetary allocations have been used to captnre Patterns of choice behavior. Researchers have argued that environments circumstances such as the level of income associated with a municipality's residents or its density fundamentally condition city expenditures. Policy, according to this view, is "caused" by environmental stimuli. I„ fact, when a linear correlational mndel is applied to demographic statistics and expenditures, politica! processes generally appear unimportant. What happens to this view when choice is moved to frontstage of one's conceptual apparatus? Prima^acie, the form of this argument is called into question since demographic factors must be considered in terms of an individual's or a group's choices. More importantly, the perceptions, predispositions, and aspirations of political actors become important for understanding how policy choices are made. City councils are crucibles of choice in many municipalities. This does not mean the set of alternatives open to council members is large. Nor does it imply that each alternative is e q „ally endowed. Rather, my conception of city councils begins with the assumption that, at a minimum, Public policy must he passed through the council. Environmental challenges and significant external political actors exert their influence through council members Thuo =,-, -,•„, ihus, an xnvestigation of councils provides a means of resolving the vprt-nr „* c ng vector of forces impinging upon the policy formation of a municipality. Research undertaken within the premises of the demographic model indicates environmental factors may fundamentally condition a city's public policies. However, when I grouped Plorida cities with populations greater than 10,000 into six clusters each sharing relatively similar

PAGE 466

447 socioeconomic conditions, I found expenditure patterns within several of the clusters were themselves different. Cities displayed different expenditure profiles despite the similarity in their environments. This leaves open the possibility that political forces act in a crucial way to determine public policy and suggests the following proposition: Proposition 1 : Socioeconomic constraints provide sufficient conditions for determining the expenditure priorities of a municipality. This view suggests the possibility that councils may act in response to a number of distinctly political considerations when allocating monies. Such politics, I reasoned, should be especially prominent in councils representing highly endowed constituencies. In high SES cities, the complexity of social interaction, the diversity of social forces, and the affluence of the citizens should foster political involvement which might perceptibly register in the council. Gauging the difference in the political situations of the many councils throughout Florida is a delicate task. I resorted to the opinions and beliefs of the members themselves. An analysis of their remarks forms the basis of this study. Of course, it is always difficult to judge the reliability of survey results. However, the attributes of the Florida council members in this sample and their descriptions of council life do not vary significantly from council members throughout the United States. This fact, together with the statistical properties displayed by the various measures developed in the study, lend credibility to the thrust of the members' comments. If environmental factors do act to condition public policy by acting through actors in political arenas, then it might be supposed that different socioeconomic environments or different clusters of cities might

PAGE 467

448 have councils whose meters „ ould dlffer „„ yarlous ^^ ^ ^.^ activity and perceptions. This is not t yP ical ly the case. Across a hroad spectrum of variables, there is no systematic difference among the clusters. For example, political predispositions do not appear related to the socioeconomic charart-pr nf t-T.„ ^ cnaracter of the cities in which the councils are located. The character of these finding s„ 8g ests the follows speculative proposition: i£2P££ition_2: Socioeconomic constraints do not provide the B££^sarv_ conditions for the abearance of polxtxcal attitudes and perceptions on city councils or among council members. There is a caveat to this proposition. The character of the cities is related to such differences on five specific measures: the degree of religious conservatism among the meters, the perceived influence of businessoriented groups and groups such as unions and har associations, the perceptions of human resource-related prohlems, and, finally, the amount of polarization evident among the council members. «hile socioeconomic factors do not appear to produce differences in the remarks of council members , systematic factors do emerge in the analysis of council ambers' remarks. The cognitive structures of members display a surprising consistency in which political predispositions play a key role. These predispositions, and especially religious predispositions, appear to distinguish the types of relations that obtain among counci! members. Participation by council members, intimacy among members, council viscidity, and perceptions of council autonomy from the comity are associated with conservative predispositions. Liberal members politically and religiously, are „„re likely to report conflictual relations typified by internal polarization, stratification, and external community pressure. These relationships appear a„„„g the members when consid-

PAGE 468

449 ered individually, and taken in the context of their council membership. In addition, other variables fit into these patterns. Notably, liberal ^bers who are likely to report conflictual relations are also likely to be attuned to problems within the community, to be young and to neglect information cues emanating from local government sources. As a whole, the character of the findings suggests the following proposition: P roposition 3 : The personal attributes of members, especially their predispositions, and their perceptions of significant contextual factors are important determinants of intra-council relations. This proposition can be supplemented with the evidence gathered among the councils of the high SES cities. Here member predispositions are important not only because of their orientation but also because of the ways in which members differ on such predispositions. The greater the differences, the more likely conflictual relations are to exist among council members. As might be expected, when councils are examined for the total amount of conflict which typifies their proceedings, relations among members are important correlates. Conflictual relations nurture conflict with the manager and among members over policy; consensual relations discourage conflict. Significantly though, religious predispositions appear to moderate the amount of council conflict. The more fundamentalist the beliefs of the members, the less likely such councils are to engage in overt conflict. Among high SES councils, the form of these relationships changes slightly. Conflictual relations still act to promote conflict, but consensual relations do not insure harmony. And again it is the differences among members on predispositions which foster conflict. When the differences in religious conservatism among the members is great, total conflict is likely to be high. Religious conservatism appears to

PAGE 469

450 s^on a set of assumptions about the nature of _ ^ interpersonai _ lationships which foreswear conflict as a useful way of conducting business. Such observations condition the following speculation: ^P-ition_4: In general, council relations determine a councxl's propensity to engage in conflictributp" Ua f 10nS H ° WeVer ' the P ersona] attributes of tne members, and especially their predispositions help determine the intensity or the conflict. y Councils are s» U groups; each me „ ber ^ ^ ^ councii ^ ^ tter. The relations which members establish with one another emerge out of this circumstance and can act to condition the manner in which decisions are made on issues coming before the councii. Some authors have ch.aracteri.ed the manner in „ hich Issues are dec£ded ^ ^^ ^_ m g to the prevaiiing relationship when the votes are counted on a series Of issues.^ But mv own observations or council behavior suggest councils -re usefully might be considered to exhibit a repertoire of decision relationships which I have termed decision-mating modes. , investigated seven different modes which fall into two broad types: 1) disagreement modes which include generalized, particularized „„d -a, particularized, and idiosyncratic splits, and 2) agreement modes which include K„~ include bargaining, amicable agreement, consensual, and informal agreement. As a rule the interpersonal relations of the members prejudice the use of specific decision-making modes. But each mode can appear in a council regardless of council relations. m fact, most councils will use several modes to resolve issue choice situations. A „„„g the councils in the high SB, cities, the li„ k bet „een council relations and decisionmaking modes is more specific. Consensual relations discourage general-

PAGE 470

451 ^ SHitS; ""««"«« -s t o limit the use of a . lcsble agreement an, consensual agreement. SIgniflcantly> bargalning remains a possibiiity regardless of the interpersonal relations among th e mejlbers . Ihese finQ _ ings suggest the following speculative proposition: Proposition 5: City councils will develop a repertoire of decisionmaking modes by which to handle poJ tr\ SltU V nS Cou ° cil Nation* will help shape the mix of methods a council adopts, but the specific forma a council adopts will evolve out of the kinds of issues consico^cil? tHe hlst ° riMl Circumstance of the "hat appears to be the case, at least among the high SES cities, is that decision-^king modes exist in different mixes among different councils And while extreme forms of interpersonal conflict encourage generalized splits among members, all other modes can co-exist regardless of council relations. Decision-making .odes conceptually capture the orientations members evidence toward deciding matters which come before the council. But every »«« will not he resolved with the same decision mode. Nor are particular modes likely to be used for any given issue. Some decision modes are .ore payable to issues than others. There seem to be some constraints on the use of no dea. In particular, modes dominated by personal relations such as consensual and informal agreement are unlikely to be nsed in choice situations which are controversial. On the other hand the generalized disagreement mode often appears in controversial circumstances. However, this relationship is considerably more complex than ^ght first be allowed. The content of the issue, and not merely its controversial aspect, appears to condition the mode which is referenced. For example, council disagreement over municipal employees and the form

PAGE 471

452 of city government tends to be quite high, yet no particular mode is referenced to handle such issues. At least among the cities most highly endowed socioeconomically, the referenrino n f a^a j, me ieierencmg of decision modes appears to be quite discriminating and complex. Councils develop their decision-making repertory over a period of time. In a certain sense, the modes are related to the relations which obtain among the members. But more importantly, decision-making modes tend to be used differentially by different councils, in a fashion which eludes systematic statistical inquiry. Part of the reason why modes evidence this behavior may lie in the forces impinging upon the council. Community forces can themselves stimulate controversy in the council on a specific issue. This can prompt a change in the mode used to resolve the conflict which is independent of policy content. Group actions within the community appear strongly linked to council disagreement over issues. This finding suggests the following proposition: Proposition 6 : Group actions are more important triggers of council disagreement than citizen initiatives. Yet it is also clear that group actions neither occur with the same frequency nor have the same ramifications across all issues. Again the pattern is quite complex. The impact of issues and their meaning within the council can be quite complex and subtle. How then are we to understand the ways in which citizens and groups become involved in council decisions or the disagreements which develop among council members? Following Lowi, I grouped issues into policy types thinking that such types might capture the interrelated aspects of council decision-making processes and community lobbying. Unfortunately, when issues are considered across a broad range

PAGE 472

453 of areas and with regard for nothing other than their content, this classification scheme proves unsatisfacotry . The conflict associated with structural policies such as regulation and self-regulation are not consistent with the hypothetical expectations. Allocatory policies do much better but, as a rule, Lowi's classification scheme (as modified for use here) does not appear geared to handle the actions of councils throughout Florida. However, there is a sense in which the scheme does make sense. When important issues are considered, and only imp ortant issues, then the relationships are as expected. Apparently, routine policies are subject to the impact of factors not salient for the consideration of important issues. The costs they provoke in the councils and promote from the community fluctuate more than those associated with important issues. This finding prompts the following observation: Pr oposition 7 : For important issues, policies can be expected to result in a predictable fashion from the pattern of council decision-making costs and externally generated costs. Grouping policies together into broad clusters on the basis of their content and expected impact may oversimplify matters in many cases. It is all the more remarkable, I think, that such classification succeeds in capturing significant differences between cities on these policies. There were twelve high SES cities which were examined in some detail. Earlier, I noted that although these cities shared relatively similar socioeconomic characteristics, their expenditure profiles were different. When policy areas over which the councils have had some experience are considered, differences appear among the councils which bear considerable similarity to the divisions on expenditure patterns. Cities with expenditure priorities on traditional services are likely to be rela-

PAGE 473

oun454 tively more attentive to traditional policies than are citie s whose c cils emphasize newer public works expenditures. The policy differences observed between cities whese socioeconomic environments are relatively similar may have many different sources. However, systematic policy differences appear related to systematic differences in the political cbaracteristics of tbe councils. For example, councils wbose profiles featured service expenditures (and constantly emphasized traditional policies, tend to be politically conservative; newer public works expenditures (and amenity policies) occur in liberal councils. Tbe more traditional service policies are likely to be promulgated by councils wbose members are unlikely to engage in fra q „e„t interpersonal conflict. Rather consensual decision making, based on perceptions of friendship, is tbe rule. Amenities, on the other hand, are nurtured in councils where particularized disagreement over issues is common, but where amicable agreement, the willingness to work toward agreement, is also prevalent. Notably, councils emphasizing traditional service policies have members who indicate business-oriented groups are influential in comity decision making, where amenities are pursued, people-oriented groups and organizations are relatively more important then else—». In sum, these observations condition the following proposition: £2P°5itl°n_8 : There is a strong link between the political teatures of a council and community and the policies pursued in that community. Liberal people-oriented councils marked by moderate amounts of issue disagreement will have a heater propensity to promulgate amenity-type policies than conservative, business-oriented the""™™ " here C ° nSenSUal declsI en -*tag is Do councils and their policies shape public policy, I think they can. But this study has not separated out the many factors bearing on policy. Rather, it suggests council politics appear related to policy.

PAGE 474

455 ^-i2Eli£itions_^or_Pol^ "The more we know and the .ore reliably we know it, the more complex will democratic governance appear to be. "3 So wr±te ^ and ^.^ ^ the epilogue of their mammot h treatment of city councils. They argue in the J^r^^o^^cr^ that governance shQuld be treat£d metaphori _ cally as a labyrinth with many entrances and exists. They write: It is not helpful to reduce the maze of democratic governance to a simple formula. There are no simplistic'solutxons to the problems arising out of the need for adapting to or controlling the internal and external environments for establishing linkages between those who inhabit the ' vernmental labyrinth regularly (the active stratum and the governors) and those who visit only occasionally (the electorate) or not at all; or for settling differences disaST^tVS C ° nfliCtS thr ° U8h " s P°-ive behavior on part'oTthe goveST" "* "**"<*" ^^ °" the As written, this statement is difficult to refute. As an epilogue to their substantial work, it belies the sophistication exhibited throughout their study. For of course, "the ways of democratic governance are labyrinthine. "5 Social reality is incredibly complex and no one should pretend otherwise. But the nature and inter-relatedness of the "labyrinth" is likely to be a function of the questions asked about its workings, the levels at which these questions are addressed, and the assumptions underwriting the answexs we accept. As I have tried to indicate in this study, the questions one poses are of fundamental importance to the answers one receives. Demographic analysis attempts to identify the determinants of public policy. However, these determinants are conceived of in the broadest possible fashion. It asks only that the linkages between environment, system and policy be drawn. 6 It does not inquire as to the processual methods used, and

PAGE 475

»y "hich these li„ kag e s are estaMished . As . co „ sequence> the parsiMy and darit y of a denoIlstratlo „ employi „ g ^ demograpMc perspectIve ^ consioerehie. u nee. co mpr ehe nd hot . fe „ variables; ^ ^ ^ ^ dfi.. reiotionshios ts Ukely to be _ u _ pron ^^ vantagepo±at the world is hardly labyrinthine. If one asks the same question—what are thP l-i„i u L aie the links between environ-t. syste., and policy ._ and also evidences a concem for a ^ ^ -** the pr oces s „ orks , the „ the level ol analysis ^^^^ NMessariiy the reseerch progra „ must be attmed to ^^^ aspects ^ ^^ poii£icai system. Hth thls chapge in level and ^ ^^ brMd ^^ ^ ^ ple*it y „ f the research lncreases greaa ^ Euiau ^ ^^ ^ 1^ the reseerch program r have traced in tMs study; shoBBhou £m ^^ c-tal*, c an proceed. I£ becomes nacessary ^ attend ^ ^ ^ ^ cre asing _ ber of variables ln order to ad ^ uateiy ^^ ^ ^^ s.toetion. Parslnony and elegance are foregone ^ ^ ^ ^ ^^ the presses a t „„ rt . The delIfflit „ ^ ^ ^^ ^ ^ In this st udy> , have had to deal ^ proMens of considerabie co» pl exit y . Both the que stio„s , attempted to ,„_ ^ ^ ^ ^ «* I tho„ g ht an adequate explanatlon must ^ cast dictatad ^ ^^ of this study . 0ftentimes my attMpts ^ capture tba macbinations of c» y council, and thelr lnvolv _ t in pgblic process bave bean axtaadad Md inV ° 1Ved Y " ' M — "** ~«ato the n et ap ho r Eulau and —« 3u gg es t . It shapes a vlsIon of politics uhicb undoas £bair ^ culean efforts. So whet is to he d „„ e? Mrst , it _« be ^^ ^ as ^ UveX of analysIs apprehands nore of the poUtiaai ^^ ^^ «~U»g, are hard to Mnc ^ Moreover _ sucb ^^^^ ^ ^ ^ ^

PAGE 476

457 be had are not UW, to render themselves in convenient packages available for easy digestion. In short, complexity most be lived with if certain options are to be addressed. Second, and perhaps mere importantly, it is necessary to come to grips with the assnmptions behind a research program. They animate net only the direction taken in the research bnt the criteria by which ade q »ate explanations are established. Further, they condition the types of q nestiens which can be ashed within a research program. in this study I have attempted to make explicit the theoretical foundations of the research. I advanced an exchange theoretical fragment as a taeans for understanding the workings of local councils. But this focus seems unlikely by itself to illuminate the workings of the. municipal political system. If for no other reas account±ng ^ ^ ^ ^ cording benefits for a large number of actors is incredibly difficult. But the exchange perspective also is best suited for individua! situations and micro analysis. Per force, it limits our view. While the exchange perspective illuminates the workings of city councils as they bear en public policy, it suffers from three defects *l* suggest it has limited utility in future policy research. First, its emphasis en individual level analysis causes it to largely ignore the broad features of the urban political system in the United States and the sweeping ramifications of this entire system of local policy. Second, it competes the analysis of relationships among political elements within a given municipality through its reliance on the undifferentiated notion of costs. Third, it understates the meaning and complexity of choice situations to political actors. Again the analvtical concept of cost is grist for many different views, support for none.

PAGE 477

458 «-» la, however, one salutary aspect to the exchMge perspect ^ f ~ — y on choice: . fulcru „ „„ uhich ^ ^^ the ^^^^ — *, ol a pol ic y and . m _ of reconciUng ae importMce ^ ^_ cental facts in demograpHc studies ^^ ^^ ^ ^^^^^ ^ ^ study has considered. Throughout the 1950's and well into the late 1960's deb , e ±i>Du s, a debate rated M the literature on American urban politics over th. FuxxLj.cs over the question of "Who «*« *>e pluralists. tod „ hUe the ^^^^^ prevaned ^ ^ ^ *— . W and ln thc „ orks of scQres of schoiars> the i±ttm ^ ^ ;""" " ^ P0Sl "° nS — " -P— tapproaches ase , One was methodological, the other substantive. 01 the .ethodological disputes, T „ m „ ^ ^ ^ ^ «*«i„tensit y ol the debate and the severit y ol the attach on various methodological ..errors'. o„ ly presaged the underl y i„g substantive Positions ol the two schoois. Both the elite theorists a„ d the pluralists locused on the banner In uhich ^ ^^ ___ ^ ^ ^ s P eetives investigate, the possibilit y ol a lunctioning d e.ocrac y an d -th attribute. co „si d erable importance to such P rocesses in shaping public polic y . Th e d e P th ol reeling which Mrked these questlons ^^ dlre " l0n ° f r — h *°* -*U. Hunter's CcHunUv^er^trucf* «~ •» ** Polio, decisions to eiite controi in a tentative "• BUt " Uh "" «~ «— * Po"o y concern ol the worhs bui ld *. on his eiiorts d i„i„i shed ln favor of ^^ ^ ^ ^^ of influence in communities. 8 B y the late sixties, the debate between the piuralists an d the elitxsts had reached an imna^P. 9 m.-i . , -passe. whxle the discussion continuedj ^.^ ^

PAGE 478

459 directions were not forthcoming. Moreover, the events of the period acted to make the debate seem beside the point. Political processes paled in importance before the dramatic social changes and extensive policy excursions of the federal government. For the moment, it did not seems as important who controlled policy as what the policy was and what its effect might be. James Q. Wilson summed up the changing research when he wondered, "what difference does it make who governs?"-^ The emerging policy focus took many different forms within the discipline of political science. But the approach which most seemed to fascinate scholars and capture the attention of journal editors, I have labelled the demographic perspective. Research carried on under this rubric focused squarely on policy as measured in the expenditures of municipalities. It attempted to determine whether expenditures were related, in fact, to political processes within cities or whether environmental factors might account for the variation in community spending patterns. As developed by scholars such as Thomas Dye, this perspective demonstrated the relative importance of environmental factors. However, such a proposition did not go unchallenged. As in the community power controversy, a debate over the methodological soundness of the findings ensued. Using different measures, researchers such as Fenton argued political processes did play a crucial role in the determination of public policy. But the findings of Dye and others held up remarkably well under the onslaught of methodological criticism. This prompted some commentators to concentrate their attention upon the substantive contribution of the approach. Shefter summarized the thoughts of many scholars when he argued the demographic perspective leads to_only the "most abstract and un in formative conclusions." In fact, the demo-

PAGE 479

460 graphic model appeared framed to answer one sort of question: namely, which is more important in determining public policy, environment or politics? This sort of question may appear "abstract" and likely to generate "uninformative conclusions," but if this is the case how is one to explain the fascination with the demographic approach? Part of the answer may lie in the way this approach opens itself up to the use of certain statistical methods. Further, since the perspective relies heavily on expenditure and socioeconomic measures, researchers were able to employ census data to engage in large-scale comparative studies of municipal policy without enormous commitments of time and money. This approach also played to the frequent equation drawn between scientific study and quantitative research which had often been lacking in urban studies. To my mind, these explanations account for only a portion of the attention given this approach. More fundamental, I believe, was the manner in which this approach and the results it yielded challenged the fundamental set of assumptions underwriting much of the research in American urban politics. At issue was the role of political processes in shaping public policy. In this sense, the conclusions of the demographic debate bear heavily upon the earlier controversy over community power. If environmental factors do affect the types of policies which are pursued by a municipality is SO me fundamental way, then the impact of political factors, however conceived, may be irrelevant. The question addressed by the demographic approach may be abstract, but the conclusion cuts to the very heart of the assumptions standing behind the community power controversy. Community power studies such as Presthus ' Men at the Ton had expli-

PAGE 480

461 citly included social and economic factors into their perspective." But such factors acted only to illuminate the status and power of elite figures. For the most part, environmental factors do not play a significant role in notions about the determination of policy. 1 * There is a direct philosophical link between the concerns which animated the community power controversy and those which fired the debate over the demographic model. In fact, these concerns are very much part of American political tradition. Without engaging in an extended argument over the ideological cast of these assumptions, I would like to suggest both debates swirled around the procedural principle of ethics. Namely, in both the community power and demographic debate, the procedural vision of American democracy was at issue. This vision holds that good policy is that which emerges from proper political procedures. This approach dominates the Fe^alis^JPaEers and more recently has achieved renewed prominence in various guises in the works of Rawls, Nozick and Buchanan. For the purposes of this argument, it is sufficient to suggest this view foreswear* elitism as a proper political procedure, sanctions Pluralism, and repudiates the overriding effects of environmental factors on public policy. 15 In each debafce> ^ procedural ^.^ ^ ^ .^^ Of course, there is an alternative interpretation of the intensity of these debates. It might well be argued that political scientists, in particular, sought to validate their own discipline. In this view, what I have called the procedural vision is important not because of its philosophical perspective but because of the need among political scientists to find an important role for politics. While it is possible this interpretation is the correct one, the case for the philosophical basis of the procedural vision is a strong one.

PAGE 481

462 The procedure! vision has a speclal ^^^^^ ^ ^^^^ ^^^^^ ^ SMrCh " ^ " POrtlM ° f "» "^rai — • d »,„ itan conmunityj ,. argued Louis Hartz over 20 years ago "ic i -k , years ago, xs Liberal community . "16 Without enterin 8 into the debate over the character ^ ^^^^ poUtiMi ^ -d t „. fo r m o£ the domlna „ t imge of ^^^ society> i ^ ^ ^ subscribe in lar g e ne asure t „ ^ ^^^ Qf ^^ ^^ jy ^ Particuiar, x share the vie „ that poUticai beuefs h ^ ^ ^^ ^ a bread substantive consensus . And ..^ ^^ ^ ^ ^^ ^ basically libera!, e g ali t aria„, indlvldualistic; populist „ 18 ihis con senses applies to political researchers aj ^ ^ ^ ^ ^ ^_ lace. The underpin, as pec t of these beliefs has affected the character of policy research. The effects of the i m pl icit procedural ^ on ^ ^^ ^ not evident ln every ease. ta policy ls taken as „ ^^ ^ able in pactin g open the co™„„ity, the bastions sidestep the assertions found in the procedure! apptoach. The content of the policy hein g e X a-ed is necassatily ta k e„ as fu„da m ental ; process considerations are of -11 importance. However, „hen policy is taken as ^ ^^ ^ able, the chances for the procedural hias to enter into the calcuiation Of the researcher are substantial. PoUticai procedures, because they are invoiced i„ s hapi„ g publlc pollcy _ are ^ ^ ^ ^^ ^ ^ aeglact of the policies t^eives. As L „„i has re^ed, "the dan g ar U. in faUin g lig btly into the cogent that policy content is irrelevant."" Policy content ts largeiy negicctej ^ ^ ^^ ^ sumptions because it is assumed that poiicy is a refiection of the subjective interests actin g to shape policy. There is noth„• . r y mere xs nothing objective about the policy its P , f -if• Y itself, it is argued, which can be considered apart from the interests involved in shaping it.

PAGE 482

463 U this ar 8 ume„t ls correct , ±t ±s easy to ^^^^^ ^^ innediate i^act of Lowi's ar g „me„t made in .^l^u^ over , decaJe ago 20 His contribution to polIcy research _ ^ yieuing ^^^ ^ ^ ^^ dent variable which conditioned the political „ political process itself. By f Q °" ssinE squarely on the ™ "p-y , Loui provlded an implIcit MJ " f aV ° lding °" '"~ tol — P"oas. Of course, one way of view"8 this approach is as an e.plicit consideration of feedback „„„„., a systems model. Nonetheless, L o„i issued fn a spate of research ^ OP6ned " the —""^ ° f "^ hypotheses oa policy formation based oa the assumptions of several different views of urban politics Speeffically, «, focus can act to move tne investi 8 ation of policy formation oac k to th e point „ he re tne community power controversy divided on the issne of influence in the policy process. IP ^ research on city councils, , have heen concerned with the delation of urhan policy and the _ r in uhich councUs ^ ^ ^ ",I have coined council relations and council decision-„aWn g .odes. Thus this study resembles many of the worhs undertaken in the guise of the procedure! ethic. Onderlyi„ g the entire „ rt has heen a concern with the manner in which polltical ac£s _ ^^ ^ ^^ ^ ^ because I explicitly focused on policy choice,, my conclusions are not strictly in the procedural tradition pm>tradition. Politics is important for policy, but the policies pursued by councils »™ n™<-y councils are contingent on a complex arrangement of factors which include the political n™,, • tne political predispositions of the members, significant community pressm-P^ n, fl „ y pressures, the manner m which members take sustenance from the community and their rWll,, y ana their colleagues, and the nature of the policy choices which present themselves. r„ SO me sense, policies appear to tahe on well-defined meaning „i thin pol itical communities. However,

PAGE 483

•464 the views of the polltical actors Md th£ ongoing reiations ^ ^ ^ tors condition actions on soch policies. This is not to sueeeqir n , H r SSest Low! s posrtton is without flaw. In mv own -search, L Rifled his senate considers!,,. However, in moving ^ f
PAGE 484

465 Notes ^-Lj£Z£t£££££S £ ^T C " ieS! Caparison, in . (New York: General Learning l 9 7l" J. 2 " al Learni ° S Press Re P"nt 2 Heinz Eulau and Kenneth Prpw-it^ t =k • , t^L^Hlkages, Representation Up ?" ^^HL^s^^eraocracy: Adapta3 Ibid. , p . 611. 4 Ibid., p. 612. 5 Ibid ., p. 611. l^tlL^^ly^^^Sll^SiiSSjMis^^ (Englewood Cliff a, „. j. : |_ 2q J Floyd Honrer, Cosslt^ajei^truoture (New York: Anchor, 1963, , 10 James Q. Wilson, "We Need tn Shiftt? cience, ed. Edward H. Buchrie f^nn I Focus," Es^a^s^n^olitical Buchrig (Bloo mi ngton: Indiana IJ^^r^T^oTTp . 131. ^nJLeaS (Be^Hiillf!^^ >rk: Oxford Ifaiversity" S J 9 stlT^ ^ T ° P: A St " dV in C o mmunltv Pow er (New ^^1^^ and Char l es Adrian , ^^^ 63^7^^-— -—-^^L^ng (Philadelphia: University of^inT.iylvInia led: Political « r Goldrich and Bert Swanson, The Rulers and the

PAGE 485

16 Louis Hartz, The Lihpr^l Tt-^--1955), p. 3. " ~ iil^^Revo^ut^^New YoTkTTZTc^uTFT^a^ SternshL, Consensus ^tflill and 27^™* °* tMs thesls ' see ^chard University, -T975TT-" 2£i ^Li|nd J 4nK^^ (Bloomington: Indiana One, t L 8 Two 5 :^ 1 l^^To^T^ ° f ^^ *>""«' Beyond the 1974 >> P. 20. ny ' ^2ii£i£2LJcience_guarterl 2 , Vol. LXXXIX (March!

PAGE 486

APPENDIX A THE RESEARCH INSTRUMENT AND SCALE VARIABLES The questionnaire used in thls research consists ^ _ ^ ^ ^ questions designed to operational^ the exchange theoretical concerns questions were phrased in q,,/-t, P an such a manner as to facilitate easy responses on the part of eounci! members. Few of the questions are original to this research. Most have heen mo dified from the efforts of earlier researchers. The questionnaire is seen in Table A-l. Question one is an adaptation of the final portion of the questionnaire used by „ ayn e Erancis in his i e s islat i ve J ss l ues^he^^ 1 It consciously preserves the format and the intent of his final questions to state legislators. At issue in each of the subdivisions of this question are council interest in an important issue, conflict noted over that issue, the resources involved in the issue, and the results of council deliberation over the ic c ,,o tk~ issue. The resources subdivision explicitly incorporates Cobb and Elder's factors for issue expansion (see Chapter two). Question two reflects the influence of a paper written by Ronald Hedlund and Keith Ha™ on legislator attitudes.^ They employ a similarly simple conservatism-liberalism scale. Their scale was devised bv Clenn Wilson and lota Patterson to eliminate s ,« of the confounding aspects of measures which employ more wordy questions to investigate respondent attitudes. 3 Wilson and Patterson argue prepositional questioning confuses as much as it informs. The researcher „ a „ „ researcher can never be sure what part 467 M

PAGE 487

468 « -' LO (.-, L y. y. £ ^" *-' — ' — ' LwJ uJ <.S = -O T "? .C ? ^ ? cocoocoo Z2 ^2222 « « w « w w y, -i ^ c ssiSSU

PAGE 488

469 3 O 5 5 o six ^ 5 — c — c = "3 C C CK I = = I . £ c 1= s; = e 5 S i ^ H u J! ~ — Z. £

PAGE 489

470 I 2 1 3 j; ;= o O i i C c c L C c.. ^ ~ c £ 5 Ft — t| ! £ ^ 5 I 5 J — i — E = c g 9 c. £ r > r. i ^ £ c n n <-, n r\* t v rv c. CV E : £ E c < , = r * " J jr ~ C t 3 2 e x ^5z E = E c c £• i E £ S S 5 = U G to "5 "C c c c r: c t 5 c c E E o " £ > > 7 «C X c c o o c o 2 Z Z Z Z Z Ifi O O (.1 */, o*, iUlu a. u L_ -* — o " t i c -c c c £ -r t < ? ° c 5 E „ ^ O c K c 2 c £ F S Is | I £ S fcr t* r* (s o o o o c o Z Z 2 2 2 2 "5 w) rt U) tf; wj L' u UJ UJ u UJ > > > > > > 5S ' f: 5:5 : C J T C r 5 o S 5 « , ^. > > ~ — r: i H £ & ? c •£ ^ § 2 c S = | = ^ > c * ; ^ z t > £ I 1 | ) > f = N f I r I 1 1 I = ; M 1 5 £ I y isj: -51 £ = £ s = z ? £ J Jz >* c c. s * 3 3 6J> L L m = = £ c c £ = s; < = o »

PAGE 490

471 c X z r.f I err _ c — c 5 c s i > p £ ** t 6. -f/A'"^ c r fe o. b III S I2l^ » — -

PAGE 491

472 of a question the respondent is responding to. As an alternative, they constructed a series of catch-words or brief labels representing various familiar (or controversial) social issues. 4 " This item formatj „ they argue, " is an improvement insofar as it reduces the influence of cognitive processes, task conflict, grammatical confusion and social desirability.'* x have modified the . r 1±st of quest . ons ^ confora ^ ^^ ^ all length of my questionnaire, and have reduced the number of items asked from 50 to 32. Question three confronts the respondent with an extensive list of Policy areas which must be examined with respect to three criteria. Admittedly, this question is complex. And although the length of the question exceeds that recommended by most researchers, the nature of the question's content mitigates this criticism somewhat. The list is composed of substantive areas of council consideration; each member is asked to identify areas of council disagreement, communal actions by the community and areas of individual citizen contact. I„ pre tests using . shorter list, respondents questioned the absence of certain areas of concern and felt compelled to write new areas into the question. The highly selective and salient nature of the type of response this question commanded the pre-test suggested that a longer list would not suffer from severe drop-off effects across the list or down the columns. And in effect, this was the case. Questions f„ ur , seven , ard ten „ ere suggested fcy ^^ ^^ of goals.* Question seven u clearly thc ^^ problemt .^ ^ pre _ tests and in numerous face-to-face interviews with councii M bers, the connotations of priv^te.gain appeared to interfere with straightforward and honest answers.

PAGE 492

Questions five and six were motivated by a concer „ over ^^ attests to settle „ atters coming before the ^ The ^^ ^ ^ Reseat Project . The ^^ ^ ^ ^^ ^ ^ ^ ^ their Appendix c , Questlo „ p ln j a6fflagHJ!£j!!=aaa _ 7 ^^^ ^ was necessary to modify the thrift nf M y tne thrust of the questions somewhat and to add questions on elects develop in Chapter two. !he stimulus fot several of the questions on conncil discord comes fr M an excelled discnssion °f logrolling in Froman's the Le£uaati^e^ro«ss . 8 a. items in question nine „ere derived and modified from several questionnaires written hy the Internationa! City Management association as re P ortad i„ the Urban Data Service Feports and the Municipa! yearbook series Questions edeven and fifteen were not used in t his study, hut a PP ear m work done by the Center for Governments Responsibility on the "Coveminent in the Sunshine" law. in question twelve, the respondent is asked to identify important organizations or groups within his co„unity. The l ist f rom „ hlch £Ms selection takes place was derived, in part , from . similar ^ used by NOP.C in their surveys of ccunity leaders.? However, the P re-test res P ooses or former members also P layed a role in developing the question. Question thirteen evolved in a similar fashion. An initial list of problem areas was pared and then supplemented by the suggestions of f„ ra er council members. The 32 lte»s in question fourteen are an attempt to elicit responses about the internal dynamics of the council itself. S „ en different M . peers of council l Ife „ere examined through the use of a modified for.

PAGE 493

of the questionnaire developed by John Hemphill. „h lle these Items „ ere specifically designed for use in seal! groups in a community settlng , se _ veral of the questions had to be severely modified to fit the council context and the length of the questionnaire. This demanded that Helm's series of items be reduced drastically." However, several questions "ere asked in each of the council characteristic areas; the focus was intended to provide information on the amount of participation among council members, the stratification of the council, the degree of council p„lari 2 atio„, the intimacy of the council members, the pressures felt by ambers from sources in the community, the degree to which the members function as a unit (viscidity), and the perceived autonomy of the council fro, outside forces. Each of these items captures selected aspects of the internal and external environments discussed in Chapter two. The questionnaire's final question, number sixteen, mas asked to get members to record their impression of the instrument. Following Wayne Francis, I attempted to give the respondent an opportunity to vent his feelings about the questionnaire." This is an immediate form of re _ ciprocity recommended by several sources. 12 The questionnaire was prepared and pre-tested on 12 former city council members during the summer of 1,74. m November , 1974 _ ^ ques _ tionnaire was sent ot each of the 444 members of city councils over 10,000 population as listed in the directory of council members published by the Florida League of Cities, of these members, 439 proved to be still in office at the time of the first mailing. 0ne hundred seventy-three questionnaires were finaliy completed. A breakdown of the responses by wave follows:

PAGE 494

4 75 First wave response 95 Second wave response 43 Personal interviews 35 Total 173 In other words, the sample of council members collating the questionnaire represents 39.5% of the total population. However, only 31.5% of the questionnaires were returned by mail . In order tQ facil±tate g ^ fully representative sample in at least one of the environmentally constrained groups, intensive interviews followed the two mailouts. During January and February of 1975, 35 interviews were recorded with council -mbers in the highest SES grouping (Group F) . A breakdown of the response rate by environmental grouping produced Table A-2. Table A-2 Response Rate By SES Group jjumber of Respondents Response Rat e A B C D E F 24 14 31 32 14 58 27

PAGE 495

476 A B C D E F Table A-3 Councils Represented In Each SES Group By Number Of Council Members Responding Number of Councils SES Group Represented 12 (17 possible) 6(6 possible) 14 (18 possible) 11 (15 possible) 6 (10 possible) 12 (12 possible) TOTALS 61 (78 possible) Number of Councils With 2 or

PAGE 496

the effects of response rate in thic ct A thxs study 1S quite difficult precisely because the population characteristics of r h • 1CS ° f the Clt y council member are virtually unknown. To my knowledge no en -i dp . 8 ' ° COm P llat ^n of such commonly used deSlgn ^iables as education, age ge, occupation or race exists. Only the sex of the members is ea^il,, a is easily determined fl nrl n, c , ac , en . f . . , ' ^ th£ USeiulness <* this characteristic i S l imited b the ber ° f W ° men in ^e overall populatxon. A check n f t,u . rTahle ^ shows that for this ^^^^^ France hetween the sample and the populatio „. Table A-5 A r„™ rable A-5 Co m ™ ° f -*« and Population Characterlstics By sex Sample Population "" 156 399 Women 16 41 Z = .523 However, this ic i-u„ t . cms is the barest possible indication t-ho . ndication the sample may be charac^erxstxc of the entire population> Without systematic criteria for n, for the entire population, typical attests to establish validity are futile. The SES er lhe SES g^oupxngs appear to r; an ait —*«— -. — 7 „ ithIn Khlch — nad e about the varlance of council CIMracterlstics _ *— *. SES „ 8S have been constructed t0 dlf contexts. Therp i c t,~ no i^ reaso „ to expect ^ poUticai ^ istics or the Individual eharaeteristies to h • -, *>*, However, the s ES „ • 8r ° UPln8S "" " e — *» -J-tion with ta own political aspects of each of the council, . • „, a , counczls. An inspection of the historidevelopment cf council „e„hetship be t„een 19 6 and 19 74 yIe lded a — of compatItio „ f „ r ^ _ ^ __ ^ ^ ^ ^

PAGE 497

4 78 of the number „ f termB served per member ot ^ ^^^^ ihe quest . on _ naire asked eaeh conncil m e„ber to give the „„„ber of years experience he had accrued o„ the council. Thus the experience of each „et,ber In the sample can he compared with the amount of experience typical of eac h council. However, when the two different measures are compared, the results are less than clear cut. In short, the sa„ pl e hears s„e relation to the population, hot the relation cannot he precisely specified. Unfortunately, my knowledge of the population of council members will support little more than this tenuously drawn conclusion; „ amely , there is no evidence ^ sug _ gest the population and the sample arp Hiffor.nf pie are different on most design variables. The Scale Variahlpg Much of the analysis in this investigation involves the use of scaled variables. The scales are composites of items designed to tap a number of the aspects of council dynamics and member attitudes. The usefulness of the scales depends upon the manner in which the scales are produced and the reliability of the scales. The reliability of a measure concerns the degree to which measurements are repeatable T> -r«= icpeacaDie. it is a necessary but not sufficient condition for the validity of a measurement. Traditionally, the reliability of scales has been tested with the split-runs tests. Recently, Cronback's alpha has come to replace this measure." Under the so-called "domain-sampling" model, the measurement error in a scale can be considered of a random sample of items from a hypothetical domain of items. If all items in the domain ^ correlated> ^ ^ existence of a common core would be demonstrated in a narrow dispersion of correlations about the mean. Cronback's alpha measures this concentration. It assumes reliability depends entirely on the average correlation a-ng items and the number of items in the scale. Reliability increases

PAGE 498

A 79 as each of the two components increases . ^ coefficIent ^ ^^ using the following formula: a = pr / [ 1 + Y (p i) ] Where r = mean inter-item correlation P number of items Ar^or suggests the usual model for ^^ rellaMllty> a product ^ psychological test theory> is ±napproprlate for ^^^^ ^ ^^^^^ He argues most sociological or polUical questlonnariea ^^^ from the±i Psychological counterparts since the conte „ t „ f ^ ^ ^ ^ ^^ -ill usually differ by . substantial ^ Moreover _ ^^ typi ^ ^ "*" "" aU lte " S In a «— •"• P-1U1 or measure a singU -derl yin8 scale property equally „ lu „ ften be ^^ ^ _ ^^ ^ ciological or political data set. That is "it-„„* mat is, items may measure a single property hot do so unequally the !te™ ™ u airy, the items may measure two or more indepenoent P ro P erties either equally or unequally. "U ^^ ^^ ^ no help in the second case. As a consequence Armor suggests the use of ractor scaling techniques. A set of items can he subjected to a factor analysis, and scales he derived in any »er of „ays." Perhaps ^ ^^ ^^ ^^ a varima* rotatIon „ hich ^^ ^^ ^^ ^ ^^ ^ their loadings can then he used to form scales. Armor describes the process as follows: abilitier 3 '^ 111 ^^ sln g le f -tor case, even better reliitemf h:t C lo n a 6 hi hlf 6 n d ^ f0rm1 ^ S " les -** oni r y r t e ne even more important Jn ru * g f en / actor This method is existence ^ P °" ant ln the multiple-factor case since the each TaTtol w ill h" ^ ^^ faCt ° rS *™ntees that The set of items thlt S ^f/^ loadi ^ close to zero. 0-40 or so with no hlner 1 H § Y " ^ faCt0r (sa ^ above

PAGE 499

480 A -re involved technique also uses factor analysis, but uses the first principal component in a principal components analysis. High loading are again taken to for, the scaie (although factor scores can he used as well) and the reliability of the scale can be ascertained by calculagenvalue of the first component h i„ the follo „ ing ^^ » [p/(p-D] [1(l/Aj)] Seven q oestio„s on the survey are made op of items amenable to the types of analysis explained above. Ques tio„ two, five, six, nine, twelve thirteen and fourteen were designed to measure several different aspects of council Ufa and member attributes. The factor technique suggested by Amor were used to build summed scales from selected subsets of items "ithin each of the questions. Each scale will be explained briefly below. The reliability coefficients for each of the scales are presented in Table A-7. The scale memoIlics are presente
PAGE 500

Table A6 Scales and Their Mnemonics 481 Scale Name Mnemonic A. Member Predispositions Political Conservatism Religious Conservatism Social Conservatism Summary (Total) Conservatism B. Influential Groups People-oriented Groups Business-oriented Groups Governmental Agency Groups Media Groups C. Information Cues Group Information Cues Citizen Information Cues Governmental Agency Cues D. Council Relations Participation Stratification Polarization Intimacy Pressure from the Community Autonomy Viscidity E. Problem Perceptions Human Resource Problems Governmental Problems Growth-related Problems Financial Problems Service Provision Problems F. Council Performance and Activity Citizen Orientation Frequency of Hearings Total Council Conflict Evaluation of Performance CONSPOL CONSRJEL CONSSOC CONSTOT PEOPLE BUSINESS GOVT INF SMALL INF GROUPQ CITIZENQ GOVTQ PARTICPT STRATIF POLARIZE INTIMACY PRESSURE AUTONOMY VISCIDITY HUMPROB GOVPROB GROWPROB FINPROB SERPROB CITIZEN HEARINGS CONFLICT PERFORM

PAGE 501

Table A6 Continued Scale Name G. Policy Disagreement Mnemonic Generalized Policy Disagreements SPLITGEN Idiosyncratic Policy Disagreements SPLITIDO Particularized Policy Disagreements SPLITPAR H. Policy Agreement 482 Bargaining or Logrolling AVOIDCON Informal Patterns of Agreement INFORMAL Consensual Patterns of Agreement ALLONE Amicable Agreement CONSENT

PAGE 502

483 Welfare Amnesty Mixed Marriages Socialized Medicine School Bussing Gun Control Disarmament This combination of controversial issues appeared to share a political basis. Each of the items relates to the ability of the state to intervene in social policy. The second factor to emerge in the varimax rotation was designated as the CONSERVATIVE RELIGIOUS PREDISPOSITION (CONSREL) scale. It consisted of highly loaded items such as: Death Penalty School Prayer Military Discipline Inborn Conscience Chaperones Right to Work Laws With the exception of right to work laws, the items of this scale appear related to religious fundamentalism. The emphasis on punishment, severe discipline and inborn conscience describe well the communion of beliefs which characterized many of the most fundamentalist sects. CONSERVATIVE SOCIAL PREDISPOSITIONS (CONSSOC) appear to capture the sense of the third scale. The third factor was highly involved with such items as: Birth Control Divorce Abortion And while these items might well be associated with the constellation of beliefs surrounding Roman Catholic dogma, their general acceptance among several groups within the U.S. population prompted a characterization which turned on their pervasive social component. Throughout this study this particular scale is not relied upon heavily because of the small var-

PAGE 503

ri.tta.lt displays across the distribution of cou „ cil ^^ tespon ^ These three dispositional scale variables use exactly half of the If in the Wilson-Patterson scale. Rather than alio., 50 percent of the information on council member predispositions to 8 „ unused, I resorted to the second teohni qU e suggested b, A r „or. The principal covenants of the 32 items in q „ es tio„ two were calculated, and the first factor was ta«e„ to represent a generalized form of conservatism. I designated it the SUGARY or TOTAL CONSERVATIVE PREDISPOSITION (CONSTOT) scale. It is a »—t y measure which includes the following items: Death Penalty Welfare Working Mothers School Prayer Sabbath Observance Communal Living Military Discipline Amnesty Modern Art Nudist Camps Corporal Punishment Mixed Marriage Socialized Medicine Busing Gun Control Integration Inborn Conscience Right to Work Laws Disarmament The inter-item scare correlations for each of the scare variables is suite high, CronbaCfs alpha is greater than .91 for each of the measures. The theta suggested b y Armor was calculated for the CONSTOT scale and is reported in Tablp A7 a in lable AJ_. A gaxn, the reliability is quite high. Group Influence Scal es The group perspective on American politics has long been one of the -st usefu! analytical tools at the disposal „f thoughtful observers of the African scene. its worth in urban politics has been explored b y

PAGE 504

485 Table A7 Reliability Coefficients for Scale Variabl es Scale VariabJp. NflTng A1 . — — . Alpha Coefficient I. Conservatism-Liberalism Seal es CONSPOL . 94 CONSREL [g 3 CONSSOC CONSTOT II. Group Influence Seal es IV. Policy Agreement Seal es V. Citizen Activity Seal es VI. Council Performance Scale PERFORM VII. Information Cue Scales .91 •94 [6 = .87] PEOPLE 92 BUSINESS \ 92 GOVT INF \ 9l MEDIA * 93 SMALLINF " 90 III. Policy Disagreement Scales SPLITGEN

PAGE 505

486 Table A-7 Continued Scale Variable Name Alpha Coefficient VIII. Council Characteristics Scales PARTCIPT .77 STRATIF .76 POLARIZE *84 INTIMACY .88 PRESSURE .92 AUTONOMY .92 VISCIDITY *80 IX. Problem Perception Scales HUMP ROB .91 GOVPROB .88 GROWPROB . 89 FINPROB .91 SERPROB .89

PAGE 506

487 numerous scholars. Consider the works of Betty Zisk or Linton Freeman. 18 Moreover, forms of the group perspective have dominated much of the literature in urban analysis swirling around the elitist-pluralist debate. Question 12 of the mail survey (see Table A-l) asked each member to name the most important groups or organizations in his community according to their role in community decision making. The responses of the 173 council members provided the basis for five distinct scale variables. PEOPLE-ORIENTED GROUPS dominate the first factor (PEOPLE). The following items are highly involved in this factor: Minority and Ethnic Groups School Employees Health Interests Welfare Organizations The second variable is summed from organization and institutions clearly identified with the business community. Several items load highly on the BUSINESS-ORIENTED GROUPS scale (BUSINESS) including: Bankers and Executives of Financial Insitutions Chamber of Commerce Church Leaders Real Estate and Land Developers Retail Merchants While church leaders may not appear a clear case of business-oriented involvement, the reliance of church leaders on the business community for financial support and fundraising activities provides some rationale for their inclusion in the factor. Perhaps a more convincing reason for the appearance of church leaders in this factor lies in the "natural" grouping which characterizes each of the items. As with the business community, church leaders are likely to exist in every community and to have interests which will require their appearance before the city council. In this sense, the business community (and church leaders) has a natural status in community life that other groups in a community will lack. IM fEy c n i B B j sW|i> l »!" m » w kmtrttt*-

PAGE 507

488 For city council members, there is another such natural grouping. The members' performance on the council and their recruitment onto the council depends upon their individual attentiveness to community politics, Ultimately, they must attend to the governmental institutions of the community. The responses of the council members reflect this perception. The third factor involves a summed group of items called GOVERNMENTAL AGENCY GROUPS ( GOVT INF ) . This scale includes: Municipal Employees Heads of Local Government Agencies City Council Manager /Mayor The perceptual side of the council members' responses appears in the fourth factor as well. The scaled variables consist of only two items; both are related to the influence of MEDIA GROUPS (MEDIA). These items are: Newspapers TV/Radio Stations A fifth factor emerged from the varimax rotation. Loading highly on this dimension were three items which members report are seldom important in their communities. For this reason and because of the position of the factor in the analysis, the scale variable has been designed as the SMALL INFLUENCE GROUPS ( SMALL INF ) . This scale includes: Bar Association Labor Unions Motel/Hotel Associations The reliability of each of the scale variables is quite high. None is lower than .90. As with the conservatism-liberalism scale, the group influence scale has inter-item scale correlations.

PAGE 508

489 Policy Disagreement Scales In order to gauge the reasons behind disagreements among council members, each member was asked to account for the manner in which his council divided. The seven possible answers found in question six were subjected to Armor's technique and three distinct factors emerged. The first scale variable sums items in which members report systematic differences account for their policy disagreement. These differences are likely to extend across a number of issues; they are quite general. As a consequence this scale variable has been designated as GENERALIZED DISAGREEMENT (SPLITGEN) . It includes the following items: Basic Policy Differences Background and Experiential Differences Personality and Ambition Differences Differences in the Constituencies Served Of course, it is not necessary for councils to differ in systematic fashion. In fact, most council votes are unanimous and appear, superficially, to reflect no dissension at all. Many city council members recognize this, and the character of their perceptions of council disagreement is best labelled as IDIOSYNCRATIC DISAGREEMENTS (SPLITIDO) . The items forming the summed scale are: Splits simply do not occur Splits do not generally occur If splits do occur, it is due to momentary and fleeting personal actions One item remains; it refers to the presence of PARTICULARIZED DISAGREEMENTS (SPLITPAR). This item is: Splits are due to differences of opinion on particular issues It is used as the one indicator of disagreements that occur typically enough to be registered by council members, but which have no generalized basis in the actions of the council.

PAGE 509

490 The three disagreement scales refer to the different ways in which disagreements are perceived to happen. They explicitly encompass the members' experience on the council and summarize numerous decision-making situations. They ask the member to determine how, in general, decision disagreements come about. The reliability of these scales is lower than those previously examined. However, the Cronback's alpha for each is still high and neither of the calculated values is lower than .85. Policy Agreement Scales While political scientists often discuss politics in terms of the conflicts associated with it, a major factor in most city councils is their degree of unanimity. The City Council Research Project directed by Eulau and Prewitt attempted to gauge such consensus in a series of items that I have included in my questionnaire. 19 These items along with some questions prompted by the work of Lewis Froman were subjected to an analysis similar to that used to construct most of the scales discussed in this section. Four scales were generated; they summarize the four dimensions of council agreement found in the responses of the council members. The first scale is composed of several items suggesting members often engage in logrolling or bargaining which emphasizes conflict avoidance. Prejudicial disagreements are avoided since every member realizes the council will continue to meet and each vote of another member may be necessary at some later time for a personal project. This is a form of anticipated logrolling done on an implicit basis that I will call BARGAINING (AVOID CON). The scale is constituted of the following items: We go along with the others to get support for our own proposal

PAGE 510

491 We go along with the others even if you disagree a bit YouT 6 n I UnCOmf ° rtable to be in the mi „o?"y You usually know how everyone is going to vote ^ you are goxng to lose in any event, you go along Rather than upset the 'apple cart ' you just go flong The second scale reflects, even more, the informality o/the relationship in most city councils. This scale has been designated i TOL AG ™ T (INTOL) The *«"** "« compose the scale variables: lZelV° ir ° n diff — S OUt ivfo-llj before the issue There is an unwritten rule that it-,•<, u -*. to disagree 1S better to a § re e than Everyone here is a friend, so unless the issue is verv i portant, you go along 1S Very lm Council members, thls scale suggests _ ^^ ^^ ^ .^^ ^ one another „ hl e„ affect the M „ner i„ „ hich they organiEe ^^ for making policy choices. StiU another pattern emerges from the member responses. Many members « disagreement on the issnes coming before the council . The ^^ perceive one another to be of lik e minds; choices are made by CONSENSUAL AGREEMENT (ALLONE) . This scale is composed of two items: UnanL,? T"^ °° dlsa Sreeme„t on many issues b^er^e^f"^^ Sl " Ce " G ™„t ^ the Sunshine" The second item is negatively correlated with the factor. Thus in the scale variable the values en this particular item are inverted. For Ufa. -inded council members, the Sunshine ba„ appears to pose no special difficulty. They agree on the issues. No consultation is necessary to produce agreement. Eor many councils agreement is won through palaver. Members work toward consensus through discussion. The unwritten rule calls for AMICABLE AGREEMENT (CONSENT). There is only one item to indicate such a process :

PAGE 511

492 We usually talk an issue over during the meeting and try to reach an agreement This is a much observed happenstance at many group meetings. Its importance should not be diminished by the paucity of indicators available. Scales of Vari ous Council Activities Each council goes about its public duties in a slighly different manner. Their ways of doing business manifest themselves in the attention paid to common council activities. Each member in the survey was asked to gauge the frequency with which eight different council activities occurred in his or her council. These responses provided the basis for a factor analysis which yielded three distinct factors. The first factor to emerge involves items relating to the frequency with which the council attends to citizen groups and individual requests for assistance. Designated the CITIZEN-ORIENTED ACTIVITIES (CITIZEN) scale, this meausre is the sum of three items. These items are. Council members make public statements explaining votes Council members provide assistance to citizens on specific matters Business groups supply information and testify before the co^nc" Most council meetings involve numerous requests by the members for more information. Agenda items are investigated and discussed at general meetings in Florida. The pervasive effects of the open meetings law demand most items be discussed in full view of the public. The members are prompted to justify their actions publicly, to question their colleagues, and to ask the city manager (or mayor) for more detailed explanations. These characteristics are manifest in the second scale which I have termed TOTAL CONFLICTUAL COUNCIL ACTIVITIES (CONFLICT) . It summarizes the following items: Counc"il P ° liCy diSagreements between Manager/Mayor and the

PAGE 512

493 2=2 =2= S?S^ .£= sX^-SThe third SC aie variable imolves tuo ^^ uhich ^^ ^ ^^ conventions of P „bHc nearlngs . DMlgnated hearing activiiies ((BnB) this scale is composed of the following items: sees in Table ^ ^ ^^ ^ ^ ^ ^ ^ ^ -is seo,e is belo„ that of the other t„o activity scaies, this al pha still reflects a substantia! degree of reliability. The evaluation of cit y councU performance is measured in a compo _ -*. scale consisting of tbe seven items asled i„ ques£ ion 9B. „„ like -t of tbe ntber q uestio„s as k ed of tbe councii members, tbis one „as -signed to produce an overall evaluation or Just one score. Xbe scaie constructed using tbe second technique suggested by armor, P rinci pal -ponents anaiysis. The scale consists of tbe factor scores produced -Tag tbe ioadings of tbe various items on tbe first factor and tbe origana! values on tbe items. Designated the PERFORCE (PERFORM) scale this measure has a theta vaiue of .80. and an aipha re lia bilit y coefficient of .94. Information Cue Scale^ Studies of ccngress have demonstrated the importance of information -as fer legislative P oiic y positions. 2 " Prhan researchers have demonstrated information sources are often surrogates for p0 „erful agents ^ tbe _ ty . Council members turn to those „itb e xp ertise, and aieng »ltb such s k iil comes a certain perfective on tbe form of municipai pol icy.

PAGE 513

494 Three scales, designed to tap members' use of various information sources, were constructed on the basis of loadings of factors produced in a factor analysis of the items in question 9C. The first scale was characterized as the GROUP CUE (GROUPQ) scale variable. It consists of the following items: Information from Experienced Business Leaders Information from the Mass Media Information from Outside Consultants Information from Civic Leaders Information from Luncheon Groups The items include a variety of information sources, but the loadings on the first factor indicate the highest correlations with items such as experienced business leaders and luncheon groups. A second pattern emerges from the responses of the council members. Council members indicate citizens themselves are a distinctive information source. In a scale denoted CITIZEN CUES (CITIZENQ) , the following items were involved: Individual Citizens at City Council Meetings Citizens Within the Community While the cynical observer might attribute this report to the conventions of local politics, it should be noted neither of these items were especially high or low in their frequency counts. Rather they displayed a consistently moderate pattern across all councils. The third source of information cues for council members comes from governmental sources. Involved as they are in a nexus of governmental agencies, this report is hardly surprising. Two items compose the GOVERNMENTAL CUES (GOVTQ) scale. They are: City Manager and Appointed City Officials Fellow Councilmen and Other Elected Officials

PAGE 514

495 Thxs scale, together with each of the other information cue measures, has a high reliability coefficient (see Table A-7) . £°H2£iL_R£l£y£ns_Scales Eulau and Prewitt, ln the f^j^ths^D^^ argue ^ ^ significant factor f„ the individual cou „ cilma „, s ^^ ^ ^ ^ CO " nCil " M a " d SlMlar1 ^ «» -' significant facto, ln che council , s political environment is the body of citizens it represents and governs."" These interrelationships, they maintain, mahe the problem of reflective adaptation on the part of the council members the most difficolt problem confronting any conncil. Tbe council is a small group with a difference. Unlike small gr „ ups studied in social psychological research, this group is elected hy a constituency to whom it is heholding for purposes which are not always clear. Selected aspects of the collective adaptation of the council were investigated using the items developed by Hemphill foI studyIng ^ groups in actual social settings.^ The items in q „estio„ U were included to tap the participation of the members in the group, the degree Co which the council is stratified by influence, the extent of polarization among the members over purpose, the amount of personal intimacy among the members, the general air of pleasantness or hedonic tone of council meetings, and the viscidity or amount of consensus among the membership over council actions. However, rather than accept these items as indicative of the council members under consideration in this study, empirical scales were derived using the items as a whole. Ibis was done because 1*11., entire list of items bad been shortened as a result of the pre-test experience where respondents became irritated over the extent of the questions framing this measure.

PAGE 515

496 Seven scale variables were constructed. The first was designated MEMBER PARTICIPATION (PARTICPT) . It consists of three items: There is a degree of council participation on the part of members The council has a reputation for not getting much done Each member of the council is on one or more active committees The second of these items loads negatively with the first factor and was included in the summed score by inverting the original values. STRATIFICATION (STRATIF) was the term labeling the second factor. It becomes a scaled variable when the following items are summed: Each member of the council has as much power as any other member Experienced members of the council are in charge Certain members have more influence on the council than others The opinions of all members are considered as equal The first and last items above were inverted in the construction of the final scale variable. The third factor of the varimax rotation loaded most heavily on those items dealing with council POLARIZATION (POLARIZE) . These items are: The objective of the council has never been clearly recognized The council is directed toward one particular goal Certain members meet for one thing and others for a different thing The council has major purposes which are to some degree in conflict The fourth factor includes items which deal with council INTIMACY (INTIMACY) . The intimacy scale variable is formed by s ummi ng the following items: All members know each other very well Members are in daily contact Members of the council do small favors for one another Members of the council are personal friends A scale variable gauging the amount of community pressure that each

PAGE 516

497 of the members perceived their council was under formed the fifth measure of council characteristics. It was designated as COMMUNITY PRESSURE (PRESSURE). The following items were used in constructing the scale: The council is under outside pressure The things the council does must be endorsed by groups within the community There are two or three members of the council who generally take the same side on any council issue There are tensions among subgroups that tend to interfere with the council's decisions The members were asked several questions about their ability to pursue objectives independent of community or individual influences. These items emerged in what I termed COUNCIL AUTONOMY (AUTONOMY) . The items forming this scale include: There are individuals in the community that really determine what the council does Plans of the council are made by other groups in the community Members often grumble about the work they do for the council The final council characteristics for which a scale was constructed was denoted by the label COUNCIL VISCIDITY (VISCIDTY) . How "thick" are the members with one another; what degree of consensus exists? The viscidity scale is made up of items bearing on such questions. These items are: Members work together as a team Certain members of the council have no respect for other members Certain members are hostile to other members The council has the support of the entire community Members of the council may suffer in some way for their council actions at the hands of an outside group The work of the council is well divided among the members Personal dissatisfaction with the council is too small to be brought up Almost everyone of the original items in Hemphill's list appears in the scales. The exceptions are items referring to the general hedonic tone of the council and one item referring to the amount of participation (SBfflWfWKT-CTiflTnnr miUIJIMIfaB

PAGE 517

498 on the part of the council members. The participation and stratification scales exhibit the lowest reliability coefficients of those calculated. However, an inspection of Table A7 will reveal even these are not especially low. Each are calculated to be more than 75 percent reliable. Scales of Problem Perceptions Each council member was asked to evaluate the seriousness posed by various problems within his community. These perceptions fell into four distinct patterns; each pattern formed the basis of a scale variable. The scale represents the seriousness a member assigns to a problem area in his community. In a sense, each scale is a reflection of the sensitivity of the council member to a host of problems. The labels, of course, serve as surrogates for a greater complex of phenomena than a mere name would allow. This is because each of the items within each scale is in itself quite complex. Problem areas were listed in question 13 in an undifferentiated fashion and were not supplemented by additional information. Thus the response to each area involves cognitive and affective components. This response might well go beyond the problems existing within a given city and strike at a member's selective cognitive processes. Each label triggers a complex reaction involving a member's predispositions and the objective reality in which he exists. Five distinctive patterns emerged from the members responses. The problem perceptions were characterized as follows; HUMAN RESOURCE PROBLEMS (HUMPROB), GOVERNMENTAL PROBLEMS (GOVPROB) , GROWTH-RELATED PROBLEMS (GROWPROB), FINANCIAL PROBLEMS (FINPROB) , and SERVICE PROVISION PROBLEMS (SERPROB) . The human resource problems consist of the following items: »?* » »» » ' »n

PAGE 518

499 Crime Education Health Care and Hospitals Poverty Urban Redevelopment and Housing The governance! probes area consists of itecs deaiing primarily „ith inter-governcental reiations. However, the sig„i£i ca „oe of traffic and transportation problem caused m e to refrain free Unking the label directly to governmental interaction dif f icniaties. I his sca le lncludes the following items: Local-federal governmental relations Local-state governmental relations Transportation and traffic problems The scale variable, growth-related problems, consists of three items: Zoning Environmental Concerns Area Growth The scale variable, service provision problems, includes two items: Public Utilities Recreation and Entertainment Finally, the financial problems scale was made up of the following items: Employment Opportunities Governmental Finance Industrial and Economic Development Racial and Ethnic Relations Tax Base

PAGE 519

Notes 1 Wayne L. Francis, Legislative Issues in the Fifty State s: A Compar ative Analysis (Chicago: Rand McNally, 1967). Ronald Hedlund and Keith Hamm, "The Effects of Institutional Change^on Legislator Attitudes: Procedural Reform and Member Perceptions, prepared for delivery at Annual Meeting of American Political Scxence Association (Chicago, 111.: August 29-September 2, 1974). ^ 3 Glenn Wilson and John Patterson, "A New Measure of Conserva' lsm ' British J ournal of Social and Cl inical Psychology, Vol. VII (1968), pp. 264-269. " 4 Ibid ., p. 265. 5 Ibid . 6 Richard F. Fenno, Jr., Congressm en in Committees (Boston: Little, Brown, 1973), p. 1. 7 Heinz Eulau and Kenneth Prewitt, Labyrinths of Democracy : Adaptations, Linkag es, Representation, and Policies in Urb an Politics (Indianapolis: Bobbs-Merrill, 1973), p. 684. 8 Lewis Froman, The Congressional Process: Strateg ies. Rules and Procedures (Boston: Little, Brown, 1967), pp. 22-27. 9 See Appendix A in Linton C. Freeman, Patterns of Local Comm unity Leadership (Indianapolis: Bobbs-Merrill, 1968). 10 Thomas W. Madron, Small Group Methods and the Study of Politic s (Evanston, 111.: Northwestern University, 1969), pp. 75-86. 11 Francis, op. cit . 12 For a good discussion of the importance of reciprocity in field research, see Myron Glazer, The Research Adventure: Promise and Problems of Field Work (New York: Random House, 1972), pp. 125-192. 13 For a readily understandable discussion of these issues, see Jum C. Nunnally, Psychometric Theory (New York: McGraw-Hill 1967) pp. 172-205. 14 David J. Armer, "Theta Reliability and Factor Scaling," Sociological Methodology: 1974. ed. Herbert Costner (San Francisco: Jossey Bass, 1974), p. 23. * w*. T-... . r|nr ..ii,ii_, i ,,. n.^,^,

PAGE 520

501 15 Ibid . 16. Ibid. , p. 38. 17 Wilson and Patterson, op. cit ., p. 265. 19 Eulau and Prewitt, op. C it ., p. 684. 21 Eulau and Prewitt, op. cit ., p. 81. 22 Madron, op. cit .. p. 75.

PAGE 521

APPENDIX B ANALYSIS OF FINANCIAL DATA FOR FY 1972 FOR CITIES The complexities of expenditure analysis are many. m Florida, the difficulties are compounded by the restrictions placed on municipalities by the state government and the manner in which state and federal funds figure in local coffers. Unfortunately, a detailed consideration of these problems is beyond the scope of this study. However, this appendix is intended to give the reader an overview of the financial situation of Florida cities. It relies on the results published by the Florida Census of Local Government Finances published in 1973, extracts of which are reproduced below. 1 Some of the significant features of the (Florida) cities by class as compiled by the Local Government Commission are presented below: 1. Population density increases dramatically with each larger population class. 2 ' w^\ nUmb ! r ° f Clty em P lo y ees increases significantly Imount a o C f P T latl ° n Cl3SS ' and ac ^ dingly the average amount of salaries increases. 3. The average operating millage increases slightly with each larger population class. 4 " with 3 ^^ PerC6nt ° f PaV6d StreetS leases slightly with each larger population class. 5. The percent of the population served by municipal or other public sewer system increases for each populadua! decline"! * 25 '°°° ^^^ *« «*"» * gra6. The percent of the population serviced by municipal refuse or other public refuse collection system increases for each population class up to 50,000 population, then shows a slight decline. ? " l h tir rC T° f P °P ulation se ™e d by municipal water or other public water systems is greater than 75 percent for all population classes and for some it is above 90 per 502

PAGE 522

503 CN C7\ to to m rH c_> 3 o 3 ,13 -H to o rH 3 o 3 E o i4-< 3 u CD 3 CD O U CD H > O O o o c rH O O O O O in C CM in I •H O O O O O O lo rH CM 3

PAGE 523

504 The most significant revenue trends overall are: 1. The amount of state assistance on a per capita basis to cities remains relatively constant at around $29 Over 53 percent of the state assistance is sent to the 14 cities with a population of 50,00] and over. 2. Federal financial assistance comprises an insignificant portion of total revenues available to cities except for the largest population class (50,001 and over) where it accounts for almost 5 percent. Cities with a population of 50,001 and over receive almost ' ten times the amount of federal assistance as the next smaller class of cities (25,001-50,000) 3. 4 5, Revenues from "other local governments" (generally through interlocal agreements) comprise an insignificant portion of total revenues available to cities ' except for the largest population class where it accounts for slightly over 5 percent. Revenue from local sources is the most significant source for all classes of cities, accounting for over 75 percent in all classes except the largest and the smallest. On a per capita basis, the total revenue available to cities increases generally as the population ?nnnn S1 ^ ^ Creases e * ce Pt for the classes 5,00110,000 and 50,001 and over. Total revenue per capita is as follows: Class Per Capita 0-2500 $158 2501-5000 190 5001-10,000 186 10,001-25,000 249 25,001-50,000 290 50,001 & over 250 'Utility Sales Revenue' and 'Ad Valorem Taxes' are the two largest sources of local revenue available to cities for 50 0m »T ° f ^ ?°P ulation c ^sses (5,001-10,000 and 50 001 and over) . Utility Sales Revenue , acco ; nts f rece p a ts 8e from erCent f ^ M ™" «* ±S water" sewer TT ^^ ° f UtlUty Rations; such as, ,*; '. SeWer ' ele ctnc, gas and transit. Therefore if Utility sales Revenue' is compared with 'Ad Valorem Taxes' on an individual basis, 'Ad Valorem Taxes' will account for the largest percent of local revenue.

PAGE 524

505 When viewed from a per capita standpoint, total expenditures for current operations show an increasing trend as the population class size increases except for the classes 500110,000 and 50,001 and over. The category "Public Safety" accounts for the largest expenditure category for all classes except 2501-5000 and 10,001-25,000; for these two classes ' the Utility' category is the largest. When viewed from a per capita standpoint total expenditures for capital outlay show an increasing trend as the population class size increases except for the classes 5001-10 000 and 50,001 and over. The category 'Sanitation' accounts'for the largest expenditure for cities in the classes 0-2500 and 2501-5000. The category 'Utilities' accounts for the largest expenditure for cities in the classes 5001-10 00010,001-25,000 and 25,001-50,000. For cities in the largest class size (50,001 and over), the categories 'Sanitation' Utilities' and 'Urban and Urban Renewal* accounts for the largest expenditure with each having equal amounts. Cities in the population class 2501-5000 have the highest total per capita indebtedness with $903. The city of Punta Gorda has the highest per capita indebtedness for the class with $986 and accordingly the highest per capita interest payments at $86. Cities in the population class 25,00150,000 have the highest per capita indebtedness for General Obligation Bonds with $111. The city of Sarasota has the highest per capita indebtedness in the class for General Obligation Bonds with $367. Also, the class 25,001-50,000 has the highest per capita interest payments with $26. For fiscal year 1972, the net change in long-term debt increased tor all population classes except the class 0-2500 where it decreased by $141 per capita. 2 While AdJ/alprem Taxes are typically among the most important sources of revenue for Florida cities, several factors affect the level of revenue derived from such sources. First, any analysis of such revenues should consider Florida's constitutional provision authorizing a $5000 Homestead tax exemption. The amount of this exemption is substantial; about a quarter of all property falls into this category. Second, municipalities can levy a millage on municipal properties for operating, debt retirement, and other purposes ranging from one to over 20 mills. However, "the 1968 Session of the Florida Legislature further restricted the real property tax by provi-

PAGE 525

506 ding that after January 1, 1970 cities could not levy in excess of 10 mills for operational purposes unless additional millage was approved by referendum." The pattern of millage adoptation among Florida cities has been quite complex. The 1969 edition of the Municipal Data Handbook cites the following figures : Fifty-three Florida municipalities reported levying ten or more mills for general operation of their local governments in fiscal 1968. These total mill levies ranged from ten to as high as 20 mills. Breakdown of these millages were as follows: (10) Thirty cities; (11) Thirteen cities; (12) Nine cities; (13) Nine cities; (14) Five cities; (15) Five cities; (16) Two cities; (19) One city; and (20) One city. In fiscal 1969, 218 cities reported levying an operating millage ranging from a low of .22 to a high of 16.52 mills. Of the major sized cities with 10,000 or more population 67 levied operational millage for municipal purposes. Of these cities, 22 levied 10 or more mills in fiscal 1969. The four large cities not levying operating millage were Largo, Plantation, Pembroke Pines and Miramar. Millages now being levied for retirement of bonded indebtedness varies considerably between cities. The recent League S t n^ in fl iC Mf d that m111 leVi6S f ° r debt service ran § e from 0.06 to 8 mills. There were 98 cities that indicated levying millage for debt retirement which totaled 201.55 mills. Nineteen municipalities reported levying a tax millage on municipal properties in 1968 for purposes other than general operational and debt retirement of improvement programs. A review of municipal property taxes should consider at least 6 major factors: 1) The total municipal revenue derived from ad valorem taxes; 2) The total tax millage on properties; 3) The total property valuations; 4) The level or percentage of the assessed or taxable value compared to the actual value of the property; 5) Legal and practical limits on millage; and o) Exemptions.3 One should not overemphasize the importance of the ad valorem tax, however. For every population group, ad valorem taxes constitute less than one half the total revenue raised for localities. Revenues available to finance Florida municipal government are from nine principal sources (in order of importance):

PAGE 526

507 1Ad Valorem taxes 2. Municipally-owned utilities and user fee for gas, electricity, water, sewerage, garbage and trash disposal service 3. Cigarette tax 4. Utility taxes 5. Franchises 6. Licenses and permits 7. Fines and forfeiture 8. Road and bridge funds 9. Parking meters In short, Florida cities have several ways in which they raise monies. And while I am not concerned with the determinants of revenue patterns in this study, there appears to be no systematic link with expenditures other than those already noted. In particular, the millage limits placed on ad_ Z alorem taxes levied for operational purposes do not appear to constrain expenditures dramatically.

PAGE 527

508 Notes 1 Florida, Department of Administration, Revenue and Expendi ture Data for City, County and Special Districts, Fiscal Year, 1972 (Tallahassee, Fla. : Department of Administration, 1973) . ~ 2 Ibid ., pp. 28-30, 37, 44, 57. 3 Florida League of Municipalities, Municipal Data Book for Flor ida Officials 1968-1969 (Jacksonville, Fla.: Florida League of Municipalities, 1970).

PAGE 528

APPENDIX C DEMOGRAPHIC VARIABLES USED IN STUDY I. Socioeconomic Status Variables Variable Name Variable Number 005 008 009 010 011 013* 014* 015 016* 017* 018* 019 029* 030* 037 045 046* 048 053 055 056 057 069* 070 071* 072 074* 075 IIFamily S Variable Number Persons per household, 1970 Percent female in labor force older than 16 years 1970 Percent employed in manufacturing, 1 970 Percent employed as government workers, 19 70 Median family income