Healthy policy, healthy babies

 Title Page
 Table of Contents
 List of Tables
 List of Figures
 1. Introduction
 2. Review of literature
 3. Methods and results - state-level...
 4. Methods and results - substate...
 5. Discussion and conclusions
 Appendix A. Healthy start expert...
 Appendix B. Substate scattergr...
 Appendix C. Substate arima...
 Biographical sketch

Material Information

Healthy policy, healthy babies the use of the Box-Jenkins ARIMA time series analysis to determine levels of success in Florida's Healthy Start program
Physical Description:
x, 221 leaves : ill. ; 29 cm.
Garner, Michael W
Publication Date:


Subjects / Keywords:
Political Science thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Political Science -- UF   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1999.
Includes bibliographical references (leaves 205-220).
Statement of Responsibility:
by Michael W. Garner.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 021557902
oclc - 43637846
System ID:


Material Information

Healthy policy, healthy babies the use of the Box-Jenkins ARIMA time series analysis to determine levels of success in Florida's Healthy Start program
Physical Description:
x, 221 leaves : ill. ; 29 cm.
Garner, Michael W
Publication Date:


Subjects / Keywords:
Political Science thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Political Science -- UF   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1999.
Includes bibliographical references (leaves 205-220).
Statement of Responsibility:
by Michael W. Garner.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 021557902
oclc - 43637846
System ID:

Table of Contents
    Title Page
        Page i
        Page ii
        Page iii
    Table of Contents
        Page iv
        Page v
    List of Tables
        Page vi
    List of Figures
        Page vii
        Page viii
        Page ix
        Page x
    1. Introduction
        Page 1
        Page 2
        Page 3
    2. Review of literature
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
    3. Methods and results - state-level analysis
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
    4. Methods and results - substate survey and analysis
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
    5. Discussion and conclusions
        Page 166
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
    Appendix A. Healthy start expert survey
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
        Page 177
        Page 178
    Appendix B. Substate scattergrams
        Page 179
        Page 180
        Page 181
        Page 182
        Page 183
        Page 184
    Appendix C. Substate arima analyses
        Page 185
        Page 186
        Page 187
        Page 188
        Page 189
        Page 190
        Page 191
        Page 192
        Page 193
        Page 194
        Page 195
        Page 196
        Page 197
        Page 198
        Page 199
        Page 200
        Page 201
        Page 202
        Page 203
        Page 204
        Page 205
        Page 206
        Page 207
        Page 208
        Page 209
        Page 210
        Page 211
        Page 212
        Page 213
        Page 214
        Page 215
        Page 216
        Page 217
        Page 218
        Page 219
        Page 220
    Biographical sketch
        Page 221
        Page 222
        Page 223
        Page 224
Full Text







I would like to extend my sincere gratitude to several individuals who helped to

make this dissertation a reality. First, the numerous Florida County Public Health

Departments representatives taught me how a major public health effort like Healthy Start

is implemented at the street-level. These wonderful professionals truly care about the

clients they serve and hope for better lives for the children they bring into this world.

In addition, I would like to thank Randall Remmel, Ph.D., formerly with the

University of South Florida, for his continued patience with my requests for dataruns and

spreadsheets. Without Randy's assistance in acquiring this data, it would have been

impossible to complete this dissertation.

I must also ext-nd my gratitude to my former employer, Carol Gormley, at the

North Central Florida Health Planning Council. Carol taught me how to interact with

competing agendas in public programs in a way that ultimately led to the interest of the

client population being of primary concern to all involved in the implementation of the

program. She demonstrated amazing diplomacy in the most difficult of situations and

continually showed how a true professional in the public policy arena should conduct him

or herself.

I owe a tremendous debt of gratitude to my current employer, Blue Cross Blue

Shield of Florida. Melissa Rehfus, Ph.D., and Pamela Smith Martin, Ph.D., continually

supported the completion of this dissertation through both personal encouragement and

generosity in the time they allowed me over the last three years to complete this work. I

know there were many times that they must have wondered to themselves whether it

would ever end. I appreciate their patience.

I do not know where to begin to thank the academic team that led me through my

undergraduate and graduate degrees at the University of Florida. I have spent so many

years in the Political Science Department that I feel at home there like few other places.

Thanks to the office staff (Debbie, Marty, and Hazel) for all of their help with long-

distance registration and insight on the upcoming football season. I would also like to

thank the wonderful faculty who always created a stimulating learning environment that is

a credit to the whole university. Most importantly I would like to thank the members of

my doctoral committee for their patience and guidance in this endeavor. David Hedge,

Ph.D., and Michael Scicchitano, Ph.D., truly earn the monikers of mentors and friends.

Their combined push for excellence in this project and in my personal life has been a

steady hand of guidance in seas that were often rough. I only hope that this work meets

their high expectations of quality and is a contribution to the academic field worthy of the

time they have given me over the years.

Finally, I would like to thank my partner, my best friend, and my love, Lisa. You

have been a constant source of caring and support through this process, and your own

accomplishments in academia served as an inspiration when I needed it most. To my

children, Brit and Karen, I am sorry for the time that this work has taken away from you.

I can only hope that you will see this work as something that made me a better person and

helped your mother and me to give you the type of life you have had growing up. I also

hope this demonstrates that education is a lifetime commitment for all of us.



ACKNOWLEDGMENTS ................................... ............... ... ii

LIST OF TABLES ....................................... ..................... vi

LIST OF FIGURES ........................................ .. ................ vii

ABSTRACT ....................................... ....................... ix

CHAPTER 1: INTRODUCTION .................. ... ................. 1

CHAPTER 2: REVIEW OF LITERATURE ........................................ 4

Policy Evaluation as an Administrative and Political Tool ................. 4
Policy Innovation in the States .................. .................. 6
States and Health Policy ................................... ....... .......... 15
States and Health Policy Analysis/Evaluation ............................... 19
Issues in Policy Implementation ............................................ 21
Summary of Policy Literature ................... .............. .............. 34
Public Policy Problem: Low Birthweight Births ............................ 34
Low Birthweight Births: Incidence, Causes and Impacts .................. 36
Medical and Psychosocial Interventions ...................................... 61
The Perinatal Paradox .................................................... 69
Programs to Reduce LBW:National and State ............................... 71
Maternal and Infant Health Policy: State Initiatives ........................ 93
Intervention Programs to Reduce LBW in Florida .......................... 97
Florida's Approach to Maternal and Infant Health Programs ............... 100
Central Hypothesis ....................... .... ........ .................. 112


Hypothesis ................................... ..... ................ 114
Dependent Variable ..................................... .................. 125
Procedure ......................................................... 126
Results ........................................................ 127
ARIMA: Identification Process ................................................ 130

ARIMA: Estimation and Diagnosis ......................................... 134
ARIMA: Intervention Component ......................... ................. 138
Results Summary: State-Level Analysis ................................. 144

ANALYSIS ..... ....................... ... ................... 145

Survey of Healthy Start Experts ....................... ................. 145
Participants ................................................... 146
Survey Results .......... ...................................... ................. 147
Description of Survey Respondents ........................................ 147
Implementation Differences in the Surveyed Coalitions .................... 150
Decision-maker Results Summary ........................ ............. 158
Methods and Results: Substate ARIMA Analysis .......................... 158
Hypothesis ....................................... ........ ........ 158
Dependent Variable ....................................... ............... 159
Procedure ............................................. ........... 159
Results of Substate ARIMA Analysis ..................................... 160
Dataset Validity ...................................... .............. 160
ARIMA: Identification Process ........................................... 163
ARIMA: Estimation and Diagnosis ......................................... 164
ARIMA: Intervention Component ........................................ 164

CHAPTER 5: DISCUSSION AND CONCLUSIONS ............................. 166

Summary of Research Findings ........................................... 166
Possible Reasons for Policy Failure .......................................... 168
Additional Areas of Possible Research ................................... .. 171

APPENDIX A: HEALTHY START EXPERT SURVEY ....................... 172

APPENDIX B: SUBSTATE SCATTERGRAMS ................................. 179

APPENDIX C: SUBSTATE ARIMA ANALYSES ............................... 185

REFERENCES ....................................... .. ............... 205

BIOGRAPHICAL SKETCH ...................................... ........... 221


Table page

1. Federal Healthy Start programs by state: 1998 .............................. 87

2. Maternal and infant care programs by state ................................... 94

3. Florida Healthy Start coalitions and their respective counties ............. 102

4. PROC PRINT output of Healthy Start dataset ............................. 128

5. Description of surveyed counties ........................ ................ 148


Figure page

1. Pregnancies and their outcomes in the United States: 1980-1992 ......... 38

2. Infant mortality rates in the United States: 1985-1996 ...................... 38

3. Infant mortality rates by age of mother, United States: 1996 .............. 40

4. Infant mortality rates by educational attainment of mother, US ............ 41

5. Infant mortality rates by smoking status of mother, US: 1996 ............. 43

6. Infant mortality rates by trimester when prenatal care began .............. 44

7. Percentage of live births with late or no prenatal care, US .................. 45

8. Infant mortality rates by sex of infant, US: 1996 ............................. 46

9. Infant mortality rates by period of gestation, US: 1996 ..................... 47

10. Infant mortality rates by period of gestation > 26 weeks ................... 47

11. Infant mortality rates by leading causes of death, US: 1996 .............. 49

12. Infant mortality rates by birthweight, US: 1996 ............................. 49

13. Infant mortality rates by birthweights > 2,000 grams ...................... 50

14. Low birthweight birth rates in the United States: 1970-1995 .............. 52

15. Low birthweight birth rates in Florida: 1986-1996 ......................... 99

16. Infant mortality rates in Florida; 1991-1996 ................................ 100

17. Healthy Start prenatal screening rates: 1992-1997 ......................... 107

18. Healthy Start infant (postnatal) screening rates: 1992-1997 ................ 109

19. PROC GPLOT output of NWLBWR by SASDATE ....................... 130

20. PROC ARIMA output for the Healthy Start dataset ........................ 131

21. PROC ARIMA output of the Healthy Start dataset with estimates ....... 135

22. PROC ARIMA output with intervention components ...................... 139

23. Means for surveyed coalition datasets ........................ ............... 161

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



Michael W. Garner

December 1999

Chairman: David M. Hedge, Ph.D
Major Department: Political Science

Within the study of public policy implementation are the principles that the success

or failure of a policy is never a given and that it is difficult, if not impossible, to causally

link the actions of various actors in the process of implementation to a particular outcome.

This research examines the Florida Healthy Start program in order to determine if

the program made a significant impact on its outcome objectives. The study used a

combination of Box-Jenkins autoregressive integrated moving average (ARIMA) time

series analysis of monthly low birthweight rates in Florida between January 1988 through

December 1996, and a qualitative telephone survey of experts in the Healthy Start

program. The results of the ARIMA time-series analysis indicate that an intervention

impact resulting from the implementation of the Healthy Start program in 1992 could not

be determined, apparently indicating that the program has not had an impact on low

birthweight birth rates in Florida. In addition, the studies found that although the

coalitions are a combination of rural and urban counties across the state, the

implementation of the Healthy Start program has been very similar across coalitions. In

conclusion, the methodology allows researchers to directly examine if a public health

program has the ability to impact a target population, beyond the larger trend of change

that would otherwise be present. These findings add to the policy innovation, policy

implementation, and medical literature in several ways.

Additional research in this area might compare the impact of Florida's maternal

and infant care programs to that of other state programs, or similar programs administered

at the federal level. Research may also, and should, be conducted on the cause behind the

consistent seasonal pattern of low birthweight births in nonwhite women in Florida. This

research should determine if the seasonal pattern is unique to Florida's nonwhite

population; whether the pattern is consistent over a larger span of years; and whether the

cause of the seasonality is biological, environmental, or a result of the predominant use of

the public health care system in Florida by nonwhite women for accessing perinatal health



In the federalist system of the United States, the states are increasingly responding

to social crises within their jurisdictions by implementing their own innovative policies,

rather than waiting for federal intervention. This is occurring because, in general, states

have undergone significant changes in their institutions allowing for greater

representation of new ideas which, in turn, have led to a greater willingness to pursue

new policies (Bowman and Kearney, 1986; Van Horn, 1993; Hedge, 1998). In terms of

health policy, states have truly served as "laboratories of democracy" by their testing of

new approaches to solving health problems and "highlighting consequences [of these

innovations] before national imposition" (Weissert and Weissert, 1996). Specifically,

the states have introduced many health policy reforms over the past twenty years

including providing universal health coverage, reforming small-group health insurance

markets, and creating new programs to address serious public health concerns (e.g.,

AIDS, low birthweight births and the resulting infant mortality) (Weissert and Weissert,

1996; Leichter, 1997).

While there is agreement that the states have become more active in creating new

polices, the question still remains whether they can effectively and efficiently implement

these polices. Gauging a state's ability to implement a particular health policy or

program has proven problematic. Additionally, determining whether the states are better

able to accomplish the goals of these innovative policies than the federal government has

not been proven conclusively. Within the study of public policy implementation are the

principles that the success or failure of a policy is never a given and that it is difficult, if

not impossible, to causally link the actions of various actors in the process of

implementation to a particular outcome (Scheirer and Griffith, 1990). For this reason,

some of the most significant divisions in the literature on implementation remain: (1)

trying to determine which factors most directly impact implementation; and (2) whether

the causal link between a program's outputs and outcomes can be established. In the end,

the inability to develop an empirical research model that can explain success or failure of

public policy in non-anecdotal terms (i.e., not case studies), and the ability to link

implementation outputs to outcomes has proven elusive (Goggin, 1986; Younis, 1990;

Lester, et al., 1987). The shortcomings in the research literature present problems for the

policy and political science fields, as well as for policymakers who need to understand

why programs succeed or fail.

In light of these continuing problems, this analysis attempts to determine whether

the Florida Healthy Start program has made an impact on the stated program goal

(outcome) of reducing low birthweight rates. Specifically, the purpose of this dissertation

will be to determine three things:

1. Has there been a change in low birthweight rates within the target
population in the state of Florida, as a whole, since the introduction of the
Healthy Start program? And if so, what has been the impact?

2. Are there differences in the way the Healthy Start program has been
implemented among the programs' administrative regions (coalitions)?

3. Have any identified differences in implementation had a significantly
different impact on reducing low birthweight rates in the target

This study is divided into several chapters. Chapter 2 is a literature review that

examines policy innovations within the states; the role of health policy analysis and

assessment in state policy innovation; implementation factors that may explain the

success or failure of the implementation of a public policy; the problem of low

birthweight births in the U.S.; and the federal and state programs created to address the

issue. Chapter 3 examines the impact of Florida's Healthy Start program at the state level

using the Box-Jenkins ARIMA time series analysis methodology. Chapter 4 examines

the impact of the Healthy Start program at the substate (coalition) level to determine if

differences in policy implementation would demonstrate any change in low birthweight

births within those substate areas. The final chapter discusses the research findings and

their implications for determining why public policies and programs often fail to succeed

to meet their stated goals.


This chapter is a review of the literature pertaining to policy innovation in the

states; factors that impact the implementation of public policies and programs; and the

incidence and causes of low birthweight births, as well as the interventions that have been

tried to reduce the problem. The goals of this chapter are: (1) to demonstrate why it is

important to focus on health policy at the state level; (2) to describe some of the

implementation factors which may be important determinants of whether a program

succeeds or fails; and (3) to educate the reader of the problem of low birthweight births in

the U.S. and Florida; and (4) to describe how this problem has been addressed both

nationally and at the state level.

Policy Evaluation as an Administrative and Political Tool

If politics is society's chosen method of allocating benefits among its citizens

within the context of legal powers (Easton, 1953), then public policy is the primary tool

of the political process. Policy is the expression of the body politic and is used to direct

the allocation of resources through various programs. In this process, interested parties

attempt to influence the allocation of these resources based on their interpretation of what

is "good" public policy. By the nature of the process, there are invariably "losers" and

"winners," and each will need to support their respective position on a given policy issue.

Because the establishment and continuous functioning of a public program is a

constant reminder of the choice that was made regarding the allocation of society's

resources, the ability of a public program to address a policy issue will be constantly

scrutinized by both advocates and opponents. One tool that is used by both sides of an

issue is a policy evaluation. Advocates evaluate the program's success so that they can

argue for continued, or increased, funding during future budgetary cycles, while

opponents evaluate a program in order to lobby for the end of the program, or for

significant modifications to its operation. Additionally, an administrator may evaluate a

program in order to determine whether it works and for whom, or to determine the

priority the policy should receive among competing programs. For all of these reasons,

policy evaluation is often a political tool that is frequently used by various interest in the

policy debate to support their particular view.

However, there are several problems that hinder interested parties from using

policy evaluations to determine whether a program has been successful and whether it

should be continued. In the United States, the policy process is complicated because the

allocation of resources takes place at all levels of our federalist system. As described by

Anton, "[t]he American system of federal governance, then, is one in which several levels

of government typically share responsibility for policy and administration -- and often

funding as well" (1997, p. 701). Whether a particular policy issue is more appropriately

administered at the federal or the state level is, itself, a constant area of debate in political

science and society. Another problem with examining public programs to determine if they

have had the desired effect (and to see if the allocation of society's resources are

appropriate), is that it is still not certain what implementation factors may improve the

chances of success and which factors may doom the program to failure.

Health policy is an issue that clearly reflects these struggles. While most areas of

health care and health policy have traditionally been addressed at the local level, certain

societal desires have created a role for the federal government in the delivery of health

care (e.g., Medicare, Medicaid, Clinton's health care reform debate in 1993). At the same

time, it is not known conclusively which factors most directly impact successful

implementation. As society increases the debate on whether health care is better

administered at the federal or state level, the ability to understand whether a particular

program can have an impact on health care issues and which factors are key for their

success will become more important.

Policy Innovation in the States

The role of the states in developing policies in the American federalist system

began to receive considerable attention during the 1980s. As the Reagan Administration

launched its "deevolution" initiatives, critics of his policies began to ask whether the states

had the ability to handle the social and economic problems that were being shifted to state

responsibility. In Bowman and Kearney's, The Resurgence of the States (1986), the

authors argue that during the late 1970s and the early 1980s, the states became revitalized

actors within the federalist system. They go on to suggest that this revitalization has

produced innovative policy in a number of areas (i.e., economic development, education,

and the environment). In order to better understand this resurgence, it is necessary to

look at the conditions of the states prior to Bowman and Kearney's research.

The original role of the federal government was primarily one of providing

common defense, protecting against barriers to interstate commerce, conducting foreign

affairs, and regulating territorial expansion (Bowman and Kearney, 1986). All other

functions were left to the states and this state-dominated federalism continued until the

turn of the twentieth century. This system went through a significant transformation after

the stock market crash of 1929. States were unable to handle the social and economic

problems that they faced and so they abdicated many of their policy powers to

Washington. The result was a federally dominated system of federalism.

What did the states look like during this time? Terry Sanford described the

condition of the pre-reform states in his book, Storm Over the States in 1967 (most of the

reforms in the states took place after 1965) (Gray and Eisinger, 1991). He opens his book

with a list of the charges against the states including that they were indecisive, antiquated,

timid, ineffective, not willing to face their problems, and not interested in cities.

Governors were seen as "Good-time Charlies" with few institutional abilities or

responsibilities (a feature that had existed from post-Revolutionary times, although some

reforms had begun around the turn of the century) (Sabato, 1983), legislatures were often

seen as corrupt institutions where "good-ol'-boy-politics" reigned (Bowman and Kearney,

1986), party competition in many states was non-existent, and executive agencies were

"bastions of mediocrity" (A.C.I.R., 1985). Sanford concluded his book with a list of

reforms that could have been the blueprint for the reforms to come.

Bowman and Kearney's argument suggests that the states have developed over the

last thirty years to a point that they should no longer be seen as weak links in the federalist

system. The authors describe a model of resurgence where the national government

became unwilling or unable to deal with many of the local social problems. In turn, state

governments, which had undergone various capacity building reforms, were able to

reclaim the policy powers that they had abdicated during the New Deal era. The authors

concluded that these reforms have moved the country away from nation-centered

federalism back towards federalism where states are on a more equal footing.

How has this model of resurgence held up and why we would expect it to change

in such a short period of time? To answer the second part of this question first, Bowman

and Kearney offered three factors that could slow, or reverse, this resurgence. Among

these three factors, the authors stated that,"the spirit of state resurgence could again (and

permanently) be dampened by worsening money problems unless federal, state, and local

officials are able to confront head-on the need for restructuring intergovernmental

financial relations" (1986, p. 35). Coming out of the recession of the late 1980s, it

appeared that Bowman and Kearney's predictions may come true. In accessing the

resurgence of the states in the early 1990s, Van Horn summed up the challenge to the

resurgence model by saying, "the boom years of increasing state revenues were replaced

by the gloomy years of budget deficits, tax increases, and program reduction. Federal

officeholders passed the buck but not the bucks to handle a broad array of public

problems.. .State officials and institutions now are being tested like never before" (1993,

p. 2).

Is the resurgence model still valid? A review of the literature suggests that the

model has held up rather well, although resurgence has slowed. Gray, Jacob, and

Albritton (1990) provide evidence that the states not only were able to survive the decline

in federal programs during the Reagan administration but also were able to continue or

expand many of the program. Furthermore, Gray and Eisinger (1991) note that even

though the value of federal aid to the states fell 39% in the years between 1980 and 1987,

the states have become more economically self-reliant. However, as the recession took

hold in the early 1990's, states were forced to cut expenditures and increase taxes to offset

the effects of the economy. The state resurgence slowed, but the federal government did

not intervene as it had done in the past [Raimondo, 1993 (in Van Horn, 1993); Van Horn,

1993 (in Van Horn, 1993); Nathan, 1993 (in Van Horn, 1993)]. The net result was that

the gains made by the states during the 1980's were not diminished relative to the power

of the federal government. As we come out of the economic downturn, the states appear

to have actually weathered the storm better than the federal government. Where the

federal government increased the deficit in response to this problem (a situation that is

prohibited at the state level), the states made tough policy choices and are now in the

forefront of many of the policies on the national agenda.

Hedge (1998) finds significant evidence that the states have continued this

resurgence well into the 1990s through both reforms on the demand-side and supply-side

of state governments. The author examines the links between citizens and policymakers

(the demand-side) and reforms that have impacted their interaction. These demand side

reforms (i.e., reapportionment, civil rights policy, electoral reforms, the use of initiatives

and referendum, interest group politics, and the resurgence of the political parties) have

had a mainly positive impact on statehouse democracy, although there have been some

negative consequences associated with the reforms. Likewise, Hedge finds a

strengthening of the supply-side institutions (i.e., procedural, professional, demographic,

and administrative changes within the executive, legislative and judicial branches of state

government) that have increased the ability of states to govern.

How has this shifting of policy innovation between the federal and state

governments changed the basic responsibilities of each? This change can be demonstrated

by examining the nature of expenditures within the federalist system and by the areas of

policy innovation. Peterson (1995) describes changes in spending patterns for the federal

and state governments between 1962 and 1990 in order to look at their respective

dominate policy areas. Peterson divides expenditures into two primary categories:

redistributive and developmental. The author defines redistributive expenditures as those

policies that shift economic resources from one group to another (from business and

prosperous individuals to the elderly, the disabled, or the poor) and defines developmental

expenditures as those that facilitate the expansion of economic resources in general (social

and physical infrastructure). Over the thirty-year timeframe (1962-1990) analyzed by

Peterson, the federal government's expenditures for redistributive policies grew at a faster

rate than the expenditures for developmental policies'. At the same time, the opposite has

occurred at the state level. Developmental expenditures increased dramatically at the state

level (at a level more than double the federal expenditure for developmental policies) while

redistributive expenditures realized only modest increases. Peterson argues that this

indicates that the federal government is increasingly concerned with redistributive polices

while the states are primarily focused on developmental policies.

' Peterson indicates that the federal government's redistributive expenditures grew from 4.8% of the gross
national product (GNP) in 1962 to 10.3% in 1990. Over the same time period, the federal government's
developmental expenditures increased from 4.2% to 5.2% of the GNP.

In terms of specific innovative policies, the "resurgence" described by Bowman

and Kearney looked at several specific areas where there was an increase in the innovative

policy making at the state level. They examined economic development policy, education

policy, and hazardous and nuclear waste policy. The authors suggest that these policies

provided new methods of dealing with these issues and that they were readily adapted in

other states and even at the national level. Other researchers have supported this

argument. The Council of State Governments (1990) described innovation at the state

level in health, the environment, and economic development. Other areas that have seen

increased state action include welfare reform, health care reform, corrections, and anti-

crime measures (A.C.I.R., 1985; Van Horn, 1993; Gray and Eisinger, 1991; Hedge,

1998). Welfare reform began in more than thirty states in the early 1990s and increased

exponentially after Clinton ended the Aid to Families with Dependent Children program in

1996. "Get-tough-on-crime" legislation (including waiting periods for the purchase of

handguns, the banning of assault weapons, and mandatory-sentencing measures) began at

the state level, and comprehensive health care reform is being tested first at the state level

[in 1993 alone, 1,850 health care reform proposals were introduced in state legislatures

(Gray, 1994)]. Many of these policies are just now finding themselves on the national

agenda, in some cases, years after they were initiated at the state level (health care reform

is one of the best examples of this phenomena).

Having considered whether state governments are the appropriate level for policy

interventions, it is important to better understand what determines whether governments

will respond with new, or innovative, policy to any particular problem. By innovative

policy, researchers are describing policy that is qualitatively different from existing policy.

Wilson (1989) argues that innovation occurs when there is a "fundamental" change in a

"significant" number of"tasks" conducted by the organization. Innovative policy may be

new and comprehensive and introduced to address new or existing problems (possibly the

policies of an in-coming administration that represents a different ideological party dealing

with new and existing problems) or policies may address new, evolving problems using the

existing programs to implement the changes (the response to AIDS in the late 1980s).

Desveaux, et al. (1994) label the former innovation as "deep policy innovation" and the

latter as "shallow policy innovation." The authors suggest that governments will react to

a problem with innovative policies if they perceive that the problem is unique and that

incremental changes to existing programs will not be sufficient to "relieve either the

problem or public pressure [to address the issue]" (1994, p. 500). The authors go on to

define policy innovation as occurring when "it deals with a complex problem in a

comprehensive fashion... [and] it relieves public pressure, relaxes the constraints on

politicians, and permits scarce political resources to be redirected to other pressing

challenges" (1994, p. 502).

The research on policy innovation suggests there are certain times when state

governments, in particular, will adopt innovative policy. Frances Berry (1994) describes

four models to explain what causes state governments to introduce innovative policy:

internal determinants, regional diffusion, national interaction, and a hybrid of these

models. According to Berry, the internal determinants model suggests that state

policymakers will primarily (almost exclusively) consider factors inherent to the state when

deciding the types of policies that should be adopted. In this model, the policymakers are

seen as "fully independent" from outside influences. The second model, regional

diffusion, predicts that a state is increasingly likely to adopt a new policy as the number of

closely surrounding states enact similar policies. This is particularly true when the region

has a similar political, social, and economic culture (Walker, 1969). Closely related to the

regional diffusion model, the third model, national interaction, predicts that state

policymakers are more likely to adopt a policy as the number of national contacts with

officials from other states (who have already adopted a similar policy) increases. Finally,

Berry argues that a hybrid of these models is necessary to reflect "both [the] internal

[state] determinants and diffusion among states" that occur when states adopt new

policies. In earlier articles, Berry and Berry (1990, 1992) introduce a methodology (a

pooled time-series analysis called "event history analysis") that takes these factors into

account and allows researchers to empirically test what drives innovation. The author

demonstrates that this hybrid approach tends to be the most accurate for explaining why

and when states adopt innovative policy. Oliver and Paul-Shaheen (1997) test these

models by looking at comprehensive health care reform in six states. The authors argue

that the national interaction model, when coupled with a supportive internal political

culture, most closely reflects the way the policy innovation occurred in the subject states,

indirectly supporting Berry's hybrid model.

Having examined why and when states develop innovate policies, the next

questions becomes what allows certain innovative policies to be adopted and

implemented? Building on the work of James Q. Wilson and R. Douglas Arnold, Oliver

and Paul-Shaheen (1997) suggest that policies are most likely to be adopted and

implemented if the policy is designed to maximize perceived benefits while minimizing

costs. The authors argue that the policy leader, or entrepreneur, must employ both

technical skill and political intelligence to create concentrated benefits (to develop eager

customers), dilute concentrated costs (ease economic, procedural, or symbolic burdens),

or convert diffuse benefits into concentrated benefits (by making them more tangible or

more immediate). The authors go on to describe how the environment in which these

policies are developed (especially health care reform policies) is a primary factor in

whether a policy will be developed, adopted, and implemented. They suggest a model to

understand this environment that centers around two broad categories of factors:

contextual conditions and dynamic factors. Contextual conditions are seen as the

institutional factors that are relatively stable over time. These conditions include

socioeconomic, political, and other institutional factors of the state (e.g., levels of income,

political culture, partisanship distribution, type and location of health care delivery system,

historical health care reforms). The second category, dynamic factors, include conditions

that tend to be individually-centered including leadership style, leadership ides, and

individually-created power structures (e.g, personal commitment to challenge the status

quo, the drive to exploit opportunities to solve problems, and the power to generate

resources to translate ideas into action). The authors argue that health care policy

innovation tends to be facilitated, rather than activated, by the contextual conditions

within the state. Institutional factors (e.g., partisanship, historical health care policy

activism) facilitate "broader reform by (1) expanding the scope of conflict on the issue to

include individuals and institutions more favorably disposed toward major reform than

those who regularly participate in the state health policy community; (2) providing an

opportunity to incubate ideas for innovation and test their political feasibility; (3)

modifying self-interest and partisanship through information and deliberation; and (4)

building on the substantive and political progress of earlier initiatives" (1997, p. 739). As

important as these conditions are for health policy innovation, the authors' analyses

indicate that "the ultimate determinants of a state's capacity for [health] reform are the

relatively dynamic factors of leadership, ideas, and power" (1997, p.741). Leadership was

critical for its ability to generate attention to the issue and the commitment of other

important actors in the system (including executive and legislative members) to addressing

the issue. But, in order to maintain the attention of these political players, the policy

leader must also be able to generate ideas that can define the current situation and future

trends as well as guide the development of innovative solutions. Finally, the individual

must have enough power to secure the appropriate resources to translate the ideas into


States and Health Policy

If asked whether health policy should be developed at the state or federal level,

most health policymakers would be unable to provide a definitive answer. There are solid

rationale for policy development at both levels. State or regional boundaries do not

restrict many health problems, and solutions to combat these problems may achieve

greater economies of scale by having federal administration. Another reason for federal

policy development includes calls by the health care system for national standards of

quality or technological regulation. Finally, the federal government may be given

responsibility when there is apparent market failure (e.g., minorities in the U.S. still find

access to care difficult or substandard in some areas). Even state policymakers have

difficulty choosing between state and federal health policy development. Health care

expenditures can wreak havoc on state budgets and the federal financing of health

programs is extremely attractive to state legislators trying to balance a budget. However,

states have consistently shown that they are unwilling to abdicate their authority over

health issues to the federal government and many of the current health policy debates at

the federal level (focusing on quality and access in health care systems) are challenged

under the charge that federal agencies are "stepping on" traditional state functions (Sparer,

1993). Other reasons for state input into health policy development is that it provides

incentives to look for innovative solutions and prevents program stagnation. Additionally,

state involvement allows "different parts of society different amounts of time to adapt to

change because some may be ready to accept it sooner than others" (Davidson, 1997).

The literature on health policy innovation often includes this pull between the state

and federal government (Leichter, 1997; Weissert and Weissert, 1996). For example,

Peterson, when examining the role of the states in the development of health policy asks,

"What can and should we, then, expect of states as the crucibles of health policy. .Are

they capable of playing a dynamic role in health policy making, and if so, to what end..

.here do they [the states] fit in the American federalist system, so transformed by practice,

law, financing, and judicial interpretation since the ratification of the Constitution more

than two hundred years ago.. .What are the sources of innovation.. What limits do states

still face as potential policy leaders.. .What are the dynamics of health politics within the

states?" (1997, p. 688). These questions are reasonable, and provide a good framework

for trying to understand the role of the states in developing health policy. However, it can

be cynically argued that that these are rhetorical questions, because the fact is that the

states have traditionally been the source of health policy innovation.

Irrespective of how the states are functioning overall and which factors are most

important to ensure effective policy implementation, the role of the states in developing

and implementing innovative health policy is hard to dispute. States have traditionally

been responsible for many aspects of health policy including the regulation of insurance

and medical providers, public health, and tort law concerning medical malpractice. In fact,

the U.S. health care system is defined by variations between states. In the few areas

where there has been significant federal involvement, primarily social welfare policy [i.e.,

Medicaid, Aid to Families with Dependent Children (AFDC)], significant "deevolution" to

the states began during the mid 1990s (Kondratas, et al., 1998). In the 1990s, the states

were looked to by liberals as the source of innovative policies that could be easily adopted

at the federal level (e.g., expansion of health insurance coverage) and conservatives

looked to the states for ideas that would allow federal agencies to delegate the delivery of

certain health welfare programs to state and local agencies (Sparer, 1998). Even the most

comprehensive federal health care reform legislation to pass in the 1990s, the Health

Insurance Portability and Accountability Act of 1996 (HIPAA), included provisions that

allowed state insurance commissioners to develop the rules and enforcement procedures

for the law, so that they could better meet community needs within each state. As the

century comes to an end, the federal government is relying heavily on the experience and

innovations that have occurred at the state level to develop the next round of national

health care reform legislation.2 Anton goes further and asserts that the "states are now,

and are likely to continue to be, the leading participants in shaping future health care and

2 Virtually all of the health care reform proposals introduced during the 105th Congress (1997-1998) have
previously been implemented at the state level over the last decade. The only exception has been the call
for a federal body that would set and enforce quality standards for health insurance plans.

welfare policy. .[especially] addressing the central issues that need attention: the problem

of currently uninsured populations and the problem of cost" (1997, p. 715).

Sparer (1998) provides a good summary of the evolution of health policy in the

U.S. federalist system. He argues that the states have played a critical role in health policy

development from the beginning of the nation. Most of the early state health policy

concerned the distribution of medical services to those who could not afford care and

public health issues (e.g., the prevention of communicable diseases). Under the influence

of the English poor-law tradition, social welfare programs (including the provision of

health care to the indigent) were considered a local responsibility. This changed following

the Great Depression when the federal government expanded its role in providing certain

welfare programs, although it was not until the 1960s that the federal government

expanded into the realm of health policy with the creation of the Medicare and Medicaid

programs. However, over the next thirty years, the states resumed their role as the

innovators of health policy with the federal government repeatedly defeated in its attempts

to establish a national health care system. As stated earlier, the result has been that the

states continue to dominate health policy in our federalist system with the licensing of

health care providers; the granting of approval for the construction of health care facilities

based on local community need; the regulation of health insurance; the administration of

workers' compensation services; medical malpractice standards; and public health services

(including primary care delivery for high-risk populations, like pregnant women at high

risk for pre-term delivery).

States and Health Policy Analysis/Evaluation

In response to the responsibility for health policy innovation falling on the states,

they are becoming increasingly reliant on sophisticated actuarial and economic estimates

of policy impacts, as well as policy evaluation techniques to determine whether programs

are having their intended effect. Coburn (1998) looked at this use of policy evaluation and

analysis at the state level. He found that state policymakers are increasingly using these

techniques to "help in thinking problems through" and "sorting out the potential policy or

programmatic responses." He goes on to explain that the area of health policy has been a

driving force in the increase in policy evaluation utilization as states are forced to deal with

the restructuring of their Medicaid programs; as they seek ways to deal with health care

payment, distribution, and regulatory problems; and as they develop innovative health care

reform initiatives. He concludes by suggesting increased collaboration between state

governments and academic policy programs in the state.

The collaboration that Coburn calls for may take many forms, however, the key is

to combine the elements of the applied (policy analysis/policy evaluation) and the

conceptual (political science). In complex substantive fields like health policy, the need

for these two areas to be combined is essential. Davidson (1997) argues that one of the

primary reasons for the 1993 Clinton Administration health care reform proposal failure

was that significant effort was put into the policy development and assessment component,

without the same work put into the political considerations. Davidson states that "even

though the analysis [of the reform proposal] was based on a detailed understanding of how

the health care system operates and what it produces, the people who assembled the

pieces into a comprehensive plan failed to allow sufficiently for the complex interplay of

[political] forces including the actions of interests groups, the state of the economy, the

composition of Congress, and other factors that determine whether an idea becomes a

law" (1997, p. 880).

Other advocates also see where social science research in health and medicine

serves several functions. Mechanic (1995) describe several "roles" performed by social

scientist in health including: framing the issues; intelligence; monitoring; evaluation and

assessment; and implementation. In Mechanic's description, social science researchers in

health policy establish a framework for looking at a particular issue, which helps to

capture the attention ofpolicymakers. Much of its success in the policy process comes

over a long-term timeframe through continuous efforts that "stimulate editorial writers,

commentators, journalists, and policy personnel" (1995, p. 1494). Social science

researchers are also important sources of intelligence, that is, they are able to describe

emerging trends and future problems. The third role of the social scientist in health policy

is that of monitoring. Mechanic states that "in the health arena, social scientists design,

administer, and analyze sophisticated surveys that track the prevalence of morbidity, use

of services, expenditures, access and satisfaction, function and disability, and many other

health related parameters" (1995, p. 1495). Finally, social scientists are important sources

of evaluation and assessment of health care programs and policies, and help policymakers

to understand the factors that may hinder or help implementation.

Of the social science fields that contribute to health care policy, the field of

political science may be the most important. Thompson argues that despiteie growing

federal involvement with health issues as well as their aura of life and death importance,

relatively few works have probed the politics of health. .[i]n particular, the role of the

bureaucracy in shaping who gets what from health programs has escaped attention..

thishs neglect is serious, for the results of policy often have as much to do with the

politics played out after a bill becomes a law as the politics in evidence prior to passage"

(1983, p. 2). In his book examining the politics of health care, Thompson reviews eight

health policies to look at their impact and concludes that implementation is critical in

shaping "who gets what" in the health care system. But the author also concludes that the

realization that implementation is critical is a far cry from understanding the process by

which it becomes so important for the allocation of societal resources. The next section

will examine how the process of implementation works.

Issues in Policy Implementation

Policy innovation in the states is not enough to guarantee issues will be addressed

unless resources are also devoted to the proper implementation of the policies. What is

public policy implementation? Pressman and Wildavsky (1984) take the literal definition

from the dictionary and add one word, "policy." Their definition is "to carry out,

accomplish, fulfill, produce, and complete a policy." This is a rather simple, yet attractive,

definition. Other authors expand on this premise and define implementation as "the

carrying out of a basic policy decision, usually incorporated in a statute but which can also

take the form of important executive orders or court decisions. Ideally, that decision

identifies the problems) to be addressed, stipulates the objectives) to be pursued, and, in

a variety of ways, 'structures' the implementation process. The process normally runs

through a number of stages beginning with passage of the basic statute, followed by the

policy outputs (decisions) of the implementing agencies, the compliance of target groups

with those decisions, the actual impacts-both intended and unintended-of those outputs,

the perceived impacts of agency decisions, and finally, important revisions (or attempted

revisions) in the basic statute" (Mazmanian and Sabatier, 1983, p. 20). This is one of the

best summaries of implementation, however, it fails to make the point that implementation

should be thought of as one of several steps in the policy process. Along with policy

formulation and policy evaluation, implementation is one step toward the solution of a


Until Pressman and Wildavsky released the first edition of their book,

Implementation, in 1973, there had been little examination of why public programs

succeeded or failed. It had been assumed that the success or failure of a program was

predetermined when it left the hands of the policymaker and that implementation was a

given. Over the last twenty years, however, others have examined this assumption and the

resulting discussion has produced at least two distinct phases, or generations, of research

on implementation and one phase that has been trying to emerge from the previous

literature to become a third generation since the mid-1980's (Sabatier, 1986; Goggin,

1986; Lester et al., 1987; Williams, 1980; Younis, 1990; Goggin et al., 1990). At this

point, it is necessary to look at each of these phases.

The first phase of implementation research was mainly concerned with "detailed

accounts of how a single authoritative decision was carried out, either at a single location.

S.or at multiple sites" (Goggin, 1986, p. 328). Most of this research entailed case studies

that tried to determine which barriers hindered successful implementation. These studies

had several common features including: a detailed, qualitative investigation of the

problems encountered by the street-level implementers when they tried to introduce a new

project or alter an existing program; a broad conceptual model that tried to incorporate

several economic, political, bureaucratic, organizational, and social factors; a "top-down"

system of analysis where the policy is examined from its inception down to its local

implementation in order to determine the extent to which its goals and objectives have

been met; and generally, a pessimistic outlook (Pressman and Wildavsky, 1984; Derthick,

1970; Bardach, 1974). This early research was very qualitative in nature, although it did

provide some interesting conclusions, such as Pressman and Wildavsky's (1984)

description of the difficulty of successfully implementing a policy when multiple veto

points are present.

The second phase of research appeared during the early to mid-1980's and

involved a move toward the development of models that attempted to conceptualize and

identify factors that helped or hindered policy implementation. This second generation

had the following common features: a more analytical approach; it often used a certain

conceptual model to test across several policy areas; an it attempted to quantify certain

factors for analysis; and it had a slightly more optimistic outlook (Goggin, 1986; Sabatier,

1986; Younis, 1990). Researchers such as Van Horn and Van Meter (1975), Sabatier and

Mazmanian (1980, 1983), and Edwards and Sharkansky (1978) all developed conceptual

models that were used by several authors with varying degrees of success (Lester and

Bowman, 1989; Browning, Marshall, and Tabb, 1980; Rosenbaum, 1980; Goodwin and

Moen, 1980).

Two of the most significant problems that came out of this second phase of

research is what Goggin (1986) calls the "too few cases/too many variables" problem and

the debate over "top-down vs. bottom-up" models in implementation studies (Sabatier,

1986; Younis, 1990). Besides the empirical difficulty in quantifying many of the variables

in these conceptual models (i.e. program specificity, support from leaders), Goggin argues

that the first two generations of implementation literature failed to differentiate() among

types of implementation outcomes, or in specifying the causal patterns associated with

these outcomes, the frequency with which these patterns occur, and the relative

importance and unique effects of each of the various independent variables that are part of

any multivariate analysis of implementation performance" (Goggin, 1986). The result of

these conditions has been a reliance on a multitude of variables with few observations.

The second issue is closely related to problems in intergovernmental relations.

Since many policies are created and implemented in our federalist system, implementation

research began asking what factors help/hinder implementation in this system. Whether

implementation is driven by top-down, national factors (Van Meter and Van Horn, 1975;

Edwards, 1980; Mazmanian and Sabatier, 1983) or bottom-up, "street-level" factors has

been an area of considerable debate (Elmore, 1978; Sabatier, 1986). Krane (1993)

explains that the top-down model of implementation is a "rational-technical" process that

depends on the "marshaling of resources" (i.e., funding, skilled workers, and facilities) to

ensure effective policy implementation. The top-down approach argues that the

implementation process can be directed by the federal actors through, among other

factors, statutory development, proper allocation of resources, and support from the

agency leaders. This approach has been criticized for its inability to demonstrate which

factors are the most important under which circumstances and the assumption that the

policymakers are the most critical actors in the implementation process (Sabatier, 1986;

Lester, et al., 1987). In contrast, the bottom-up approach stresses the importance of the

persons that are responsible for delivering the services and client interaction, or the street-

level bureaucrats, for successful implementation. The literature on the bottom-up

approach argues that these actors are able to effectively distort or resist cooperating with

these policies through various techniques. Because of this, this research suggest that

policy outcomes are determined by interaction and bargaining between agency actors and

the clients (Elmore, 1978; Sabatier, 1986). The bottom-up approach also takes into

account the lower-level bureaucrat's individual commitment, identification, and knowledge

of the policy's goals and objectives (Lipsky, 1977; Hedge et al., 1988).

In order to overcome these deficiencies, Goggin and others (Lester et al., 1987;

Sabatier, 1986) have argued that a third generation of literature needs to be developed that

can better access the factors that affect implementation in order to shake off this

"intellectual baggage" (Williams, 1980). Many of these authors believe that this third

phase of research should involve the development of testable theory, especially by linking

outputs to outcomes, and by demonstrating causal relationships. This is not to say that the

literature does not address this output-outcome link, but that the current literature tends to

focus on its inability to show that the link exist. Most of the literature on this issue tends

to describe the failure of policy outputs to produce the expected outcomes (Morgan,

1996). This dissertation will address these issues and contribute to a third generation of

implementation research.

What are some of the factors that can help or hinder policy implementation? Even

with all of the criticisms of this implementation research, a good conceptual framework

has developed from the literature and some factors have been quantified. There have been

several analytical models (Mazmanian and Sabatier, 1983; Van Meter and Van Horn,

1975; Elmore, 1978) and several case studies that have tried to determine which factors

have the greatest impact on the implementation process (Pressman and Wildavsky, 1984;

Hanford and Sokolow, 1987; Mazmanian and Sabatier, 1981; Edwards and Sharkansky,

1978), but few studies have identified good measures of these factors (West et al., 1990-

91; Lester and Bowman, 1989; May, 1993; Wood, 1992; Hedge et al., 1988, 1989, 1991)

and fewer still have tried to link these factors to outcomes (Durant and Legge, 1992;

Ringquist, 1993). Among the most useful factors that have been examined include:

problem tractability, organizational factors, environmental conditions at the bottom of the

implementation chain, and the attributes of actors at the various levels of implementation.

Each of these factors need a more thorough explanation.

Problem Tractability

A policy is often created to solve a problem. Depending on what type of problem

the policy is trying to solve (social, economic, environmental, etc.), the chances of

implementation success or failure can be surmised. Some problems are simply more

difficult to solve than other problems. Hazardous or nuclear waste is simply more difficult

to manage than solid waste. This is especially true of the problems that come into the

public sector for solutions. These problems have often been delegated into the public

sector because of market failure in the private sector, yet they are problems that must be

managed. Paul Sabatier and Daniel Mazmanian (1980,1981,1983) conceptualized this

idea in a series of publications released in the early 1980's. Several authors used their

implementation model and found the concept of tractability withstood examination

(Goodwin and Moen, 1980; Browning, Marshall, and Tabb, 1980; Rosenbaum, 1980;

Mazmanian and Sabatier, 1981). Under this concept one can conclude that to the extent

the problem has valid technical solutions, technologically accurate instruments of

measurement, specific types and amounts of behaviors to alter, an identifiable and

geographically concentrated population to target, the problem has a higher or lower

degree of tractability that will help or hinder implementation.

Organizational Factors

How well bureaucratic entities function can be one of the greatest indicators of

implementation success or failure, and so bureaucratic capabilities in the form of structure,

communication, leadership, coordination, and resources can make the difference between

success and failure. Pressman and Wildavsky (1984) often cite the inefficiency of the

project approval system of the Economic Development Administration's (EDA) Oakland

project as one of the greatest obstacles in its implementation. The actors involved had far

too many clearance points (or agencies) to go through to gain approval for projects. The

authors calculated that it would often take 30 to 70 steps to gain approval. These "veto

points" caused several delays in implementation and often ultimate failure for particular

projects. Related to this, the structure of the agency and the implementation of the

program can be very important to its function (Pressman and Wildavsky, 1984; Reppucci

and Suanders, 1978; Sherman, 1978; Maynard-Moody, Musheno and Palumbo, 1990;

Hasenfeld and Brock, 1991). Whether the agency's decision-making process is centralized

or decentralized can affect the implementation process through the ability to ensure

accountability or the ability to respond quickly to a changing environment. What type of

departmentalization has been utilized (program, function, client type, or geography), and

how the agency is able to change with its environment may all have an impact on

implementation. Another point of failure in the EDA project in Oakland was that the

agency was not able to change as the obstacles increased. With each delay, the

expectations of the actors involved changed and the EDA could not keep with the changes

(Pressman and Wildavsky, 1984). Effective leadership within agencies is also important to

program implementation (Pressman and Wildavsky, 1984; Sharp, 1981). The value of a

consistently active leader has increased the probability of success for many programs, as

much as the ineffective leadership or departure of an effective leader has brought a level of

decline to many programs. Pressman and Wildavsky (1984) indicate that one of the first

negative turning points of the Oakland project was the departure of Eugene P. Foley as

head of the EDA. Soon afterward, the relationship of the project to other agencies and its

importance within the EDA proved more complicated. However, other authors (Sabatier

and Mazmanian, 1980) point out that "...leadership skill remains a rather elusive concept."

While everyone acknowledges its importance, its attributes vary from situation to situation

and thus it is extremely difficult to predict whether specific individuals will go beyond

what could reasonably be expected in using the available resources in support of statutory

objectives." Resources (in the form of personnel, equipment, information, expertise, and

available funds that can be distributed to clients) can be one of the most important

influences on successful implementation (Van Horn and Van Meter, 1987; Pressman and

Wildavsky, 1984; Mazmanian and Sabatier, 1983; Paul-Shaheen, 1990). Several authors

find this to be an obstacle in the implementation of environmental regulation (Lieber,

1983; Worthley and Torkelson, 1983; McDowell, 1988; Hanford and Sokolov, 1987),

while others find this concern in social policy (Edwards and Sharkansky, 1978;). The level

of resources can be a measure of the policy's importance in the eyes of external actors

(legislatures) and resources can also be used as a measure of internal political factors

because they can reveal which programs can gain priority within the organization

(Ringquist, 1993).

Environmental Factors

This concept tends to be the dumping ground for the factors that tend to include

the political, socio-economic, and demographic factors of the area in which the program is

to be implemented (Crotty, 1988; Mazmanian and Sabatier, 1983; Pressman and

Wildavsky, 1984). In practice, these variables tend to become the control variables in the

models and play a mediating or confounding role in terms of trying to link outputs to

outcomes. These factors would be conceptualized as non-statutory variables under the

Mazmanian/Sabatier model of implementation. Pressman and Wildavsky (1984) found

that the EDA staff cited the troublesome political environment of the city of Oakland

(especially the resistance of the first mayor during implementation) as a serious obstacle to

successful implementation. Other authors (Goodwin and Moen, 1980) found that

economic factors influence the ability to implement some social policies.

Attributes of the Actors

After having taken all of these larger factors into consideration, there still remains

one factor to examine, the role of the individual actor (county health unit administrator,

coalition member, nurse practitioner, and clients). The role of the individual in the policy

process has been an issue of debate since the turn of the century. With the rise of the

"New Institutionalism" concepts in political science, the role of the individual has become

challenged once again (March and Olsen, 1984). Of all the attributes that the individual

has at his/her disposal to shape the behavior of organizations, the following three

attributes are most important: cognitive ability, motivations, and attitudes.

By cognitive ability, the literature is describing the psychophysiological ability of a

person to process information and make decisions. Motivations are those internal factors

that define the reasons for individuals to engage in any activity. Finally, attitudes are those

internal characteristics that shape the type of action to be undertaken and have the ability

to distort the other two attributes. In order to understand how these attributes affect the

shape of implementation process, each must be examined in greater detail. The cognitive

ability of individuals has been one of the earliest factors studied in organizational theory.

Max Weber was one of the earliest individuals that argued the rationality of the decision-

making process was imperative in optimizing efficiency in bureaucracies. To the extent

that an individual can receive and translate information into a decision in a rational

manner, the more likely it is that a policy will be implemented in a manner consistent with

the organization's objectives (Fry, 1989). However, the cognitive ability of individuals to

make purely rational decisions has also been questioned from an early time. Herbert

Simon argued that individuals are limited in their ability to make "rational" decisions

because they are cognitively limited, that is, they are unable to consider all information and

options at the same time. Furthermore, what information is available has been biased by a

limited perception that has been structured by other attributes such as the individual's

attitudes and motivations. The result is a set of decisions that rely only on the biased

information that can be considered at that particular point in time, thus we get Simon's

"satificing man" (Fry, 1989). If we assume that efficiency (the greatest possible output for

the least input) is the rationale for organizing, than this limited cognitive ability of the

individuals within the organization reduces the capacity of the organization to realize

efficiency. This can manifest itself in higher monetary costs or policy outcomes

significantly different from what was intended.

The second attribute of individuals that can shape the behavior of the organization

is the motivation of the individual. Gortner, et.al. (1989) describe how motivations affect

the behavior of organizations. First, motivations are some of the primary reasons that

individuals decide to join and continue to participate in organizations (Rusbult and

Lowery, 1985). Second, motivations can determine the extent to which the individual

agrees to pursue either personal or organizational goals. Finally, motivations determine

the extent to which individuals allow others to direct or control their behaviors. Gortner,

et. al. examine Anthony Downs' typology of bureau official motivation to understand how

motivations shape organizational behavior. Downs describes five types of individuals:

climbers, conservers, zealots, advocates, and statesmen. The first two types are seen to be

primarily motivated by self-interest. The climber seeks to maximize his/her power,

prestige, and monetary reward by "climbing" the organizational ladder as quickly as

possible. The conserver seeks to maximize his/her security and convenience by resisting

change and avoiding actions that may reveal failure. The implication for the former type is

that he/she can help to maximize short-term goals at the cost of alienation of co-workers

and subordinates which results in lower moral, while the latter can impede the

responsiveness in the organization to new problems. The other three types of officials

have motives that are combinations of self-interest and altruistic loyalty. These

motivations can have positive effects such as overcoming inertia, promoting change, and

obtaining resources for the organization, but they do so at a cost. These motivations can

cause alienation among personnel within the organization and between organizations.

They can also distort the overall goals of the organization by redistributing resources away

from product output to institutional preservation.

The final attribute of individuals to be examined are attitudes. These are cultural

and professional norms and perspectives that have the potential to filter the perception of

information (limiting cognitive abilities) and distort organization goals and objectives

(altering motivations). Attitudes are shaped from birth are influenced by the family,

school, peers, and society at large. How do attitudes shape the behavior of organizations?

The clearest way is by distorting the decision-making process and the implementation of

policy (particularly at the street-level). Hedge, Menzel, and Williams (1988) look at five

attitudes that they presumed would have an affect on site-level regulation. They believed

that the extent to which inspectors had certain attitudes (rule orientation, perceptions of

coal operators, etc.) the greater would be the enforcement of the regulation. They found

that some of these attitudes do tend to affect the level of enforcement. They concluded by

suggesting that these attitudes may hinder the ability of centralized authorities to control

subordinates' behaviors, thus distorting the impact of the policy. Lipsky (1976) also finds

evidence that the attitudes of individuals can distort not only the distribution of

bureaucratic outputs but the perception of the organization by the clients as well. With

the problems that these attributes can cause in organizations in mind, the question

becomes to what extent can organizations control these effects? The primary and

traditional form of control has been the structure of the organization itself. The Founders

of the Constitution saw organizational structure as the primary method of control the

excesses of the individual and the masses, thus they constructed an institutional separation

of powers (Hult and Walcott, 1989). Many of the early theorist in organizational theory

(Frederick Taylor, Max Weber) saw the structure of the organization as the primary tool

in controlling individual attributes that could affect efficiency. The effects of cognitive

limitations can be controlled by centralizing decision making so that the effects are

minimized, by standardizing the decision-making process, and by monitoring decision

making. The fewer the number of individuals involved in decision making, the fewer the

opportunities for discretion, and the greater the chance to discover deviations from

objectives, the less the potential for distortion of policy implementation (Kaufinan, 1967;

Fry, 1989; Perrow et al., 1989). Reorganization and the internalization of agency

objectives can be used to refocus individuals' motivations away from unfavorable behavior

toward greater agency responsiveness (Kaufman, 1967; Fry, 1989). Socialization,

professionalization, and recruitment policies have also been promoted as tools to restrain

the effects of attitudes (Kaufnan, 1967; Kearney and Sinha, 1988). All of these tools

have been used at different times, with varying degrees of success, but there are

drawbacks to using these methods.

Organizations that rely on strong, centralized authority in decision-making

processes can be too rigid and unable to respond to new problems. The result may be an

organization that has a dramatic decrease in efficiency or the inability to remain efficient at

a reasonable cost (Chubb and Moe, 1988). Reorganization may produce more political

than organizational benefits and these tactics often have cost associated with them that

make their use questionable (Meier, 1980). Socialization, professionalization, and

recruitment may increase individual responsiveness, but only within certain limits. Hedge,

Menzel, and Williams (1988) point out that these methods of controlling behavior could

be overpowered by proximate influences like local interaction and other attitudes. Other

factors like outside political interest and legalistic constraints may also have an effect

(Moe, 1985; McCubbins, Noll, and Weingast, 1987; Bendor and Moe, 1985).

This is not to say that none of these make any difference in controlling the effects

of these attributes, but rather, they are limited in the extent that they work. No single,

bureaucratic tool can control these individual attributes (at least, not without more serious

consequences), just like no single individual attribute can determine organizational

behavior. It is also possible that these control methods work better in either spatial or

temporal situations. What works at one time may not work at another and what works in

one organization may not work in the next (Tipple and Wellman, 1991). This is why it is

important to continue to study both sides of this equation, the individual attributes and

organizational controls, especially in terms of how the individual affects implementation.

Summary of Policy Literature

As this discussion demonstrates, in order for innovative policy to be implemented

successfully there are many factors that must be considered by the policymaker and the

implementing administer. Policymakers, administrators, and academics continually look at

these issues to understand whether a program has achieved the desired outcomes. The

next portion of this literature review will examine the social health problem of low

birthweight births and the programs created to address the problem.

Public Policy Problem: Low Birthweight Births

From birth until death, humans are faced with potential health problems that affect

both the length and quality of life. Newborns face a beginning predestined by the genetics

and the lifestyle of their parents (especially the impact of the mother's behavior on the

fetus in her womb), and the health care resources of the community. Children grow in an

environment that presents the danger of accidents, limited access to health care services,

potentially fatal illnesses, and a range of devastating social conditions like poverty.

Adolescents continue to develop in a world wrought with accidental death and illness, but

also must face new dangers resulting from high-risk personal behaviors like casual sexual

relations, drug and alcohol abuse, tobacco use, and the prevalence of violence in the

current youth culture. While some of these threats diminish in adulthood, new problems

arise in the form of certain mental illnesses, stress-related conditions, poor dietary habits

and obesity, and occupational injuries. As people approach their later years, the

cumulative effect of the problems faced earlier in life begin to take their toll on health.

Heart disease, cancer, and other health conditions become the primary causes of death,

and dealing with these conditions are compounded by concerns over the financial burden

of immediate and long-term health care. Because health problems facing an individual are

common among most people at the same point in life, society often calls for public

interventions to address these health conditions.

Public interventions usually take the form of public health programs that focus

either on the prevention of illness, the reduction in the spread of disease, or the facilitation

of access to health care. Prenatal care for high-risk pregnant women, childhood

immunizations, epidemiological monitoring of disease outbreaks, and medical services

funding programs like Medicaid and Medicare are all examples of public interventions. In

order to get these programs initiated and continued, policymakers must be convinced of

the existence, severity, and pervasiveness of the problem. This is often accomplished by

utilizing policy analysis and program evaluation. The extent to which these analyses can

be understood and are able to demonstrate that the program will solve the problem for

which it is intended to address, determines the likelihood that the policies and programs

will exist. For example, Yankauer, in his analysis of social policy and the use of social

sciences states "[t]he WIC program like Head Start, Improved Pregnancy Outcome

(IPO), and other large child heath and welfare programs was sold to legislators by

advocacy groups on the grounds that the services provided would achieve gains in health

that could be measured by specific health indicators.. .[t]he strategy by which these

programs were promoted and sold to legislators thus left their advocates in a vulnerable

position if evaluation failed to show that measurable outcomes had been achieved as

promised, or that the program costs exceeded their benefits by a substantial margin"

(1985, p. 181).

This section will look at a public health problem and the programs developed to

address the problem. The first part of this section examines the importance of the issue

(i.e., the incidence, causes, and impact of low birthweight births). Next, the section will

describe the medical and psychosocial interventions that have been introduced to reduce

the incidence of this condition. With this general understanding of the problem, the next

part of this section will describe national and state programs developed to reduce the

incidence and effects of low birthweight births. Finally, this section will describe the

prevention program developed in Florida on which the rest of this research will be


Low Birthweight (LBW) Births: Incidence, Causes, and Impacts

In the United States, just over half of all pregnancies to women between the ages

of 15 and 44 years result in a live birth (Figure 1). The rest of these pregnancies are ether

terminated through induced abortions or are the result of fetal demise (in fact, many

pregnancies spontaneously terminate before the woman is even aware that she was

pregnant). For those pregnancies that do result in a live birth, the threat of the infant

dying prior to its first birthday is still a serious concern. For this reason, a great deal of

health policy is focused on understanding what factors increase the risk of infant mortality.

Infant Mortality and the Low Birthweight Birth Rate Link

Infant mortality has been a major health policy issue in the United States since its

creation. In order for a nation to grow, its population must have sustained growth

through natural birth and immigration. In addition, the personal impact of an infant death

is one of the most traumatic experiences most people will ever endure. For these and

many other reasons, infant mortality is a serious public concern. Although Many

significant strides have taken place over the last century. In the United States, overall

infant mortality has declined from 140 deaths per 1,000 live births at the turn of the

century to 7.2 deaths per 1,000 live births in 1996 (see Figure 2). Even with the decrease

in overall mortality rates, each year there are still approximately 22,000 newborn deaths

within the first 27 days after birth (Adams,1995).






7 40

2 20

1981 1983 1985 1987 1989 1991
1980 1982 1984 1986 1988 1990 1992

STotal pregnancies Live births
Induced abortions Fetal losses
Soa:US Ntod cwmc 5ikbr HeSmIa nai s*ivafrnk S4llUdS

Figure 1 -- Pregnancies and their outcomes in the United States: 1980-1992


15 Black


-- --- - -------- ~ A ll-races-

s White

S All Races ------ White -- -- Black
19 1 8 I 190 I 1992 1994 s 199
1985 1987 1989 1991 1993 1995

Figure 2 -- Infant mortality rates in the United States: 1985-1996

As Figure 2 indicates, overall infant mortality rate has been declining steadily,

although there remains a significant difference in mortality rates between whites and blacks

in the United States. Between 1950 and 1991, infant mortality for white infants declined

by 3.23% per year, while infant mortality rates for blacks in the U.S. only declined by

2.89% annually (Singh and Yu, 1995). Birth outcome disparity between whites and blacks

in the U.S. is an area of constant attention in the public health policy field. Considerable

research has examined the racial disparity in infant mortality and other poor birth

outcomes, however, no conclusive evidence points to any of the known correlated risk

factors as the sole reason for the disparity. The racial difference is constant throughout

the data examining the factors that contribute to infant mortality (Frisbie et al., 1997;

Shiono, et aL, 1997). An interesting note on this subject is that the racial disparity in birth

outcomes only seems to apply to U.S. born blacks. Persons of African descent, whether

immigrating directly from Africa or from another part of the world, consistently report

birth outcomes similar to the white population in the U.S. and considerably better than

U.S.-born blacks. This remains true even when many of the other risk factors (e.g., low

educational attainment, poverty) are present. Why this disparity in birth outcomes exist in

the U.S.-born and African immigrant populations continues to be studied (Wasse et al.,

1994; David and Collins, 1997).

The health policy literature has examined many factors in order to determine what

has the most significant impact on the chances of infant mortality. Some of these factors

examined include: mother's age; mother's education; mother's smoking status during

pregnancy; the trimester when prenatal care began; the sex of the infant; the gestational

age of the infant; and the infant weight at birth. As demonstrated by Figures 3 through 14,

all of these factors are seen to have some impact. Each will be examined here in greater


Figure 3 examines the effect of the mother's age on the risk of infant mortality. As

seen in this data from 1996, the mother's age has a predictive value for identifying infants

with a higher risk for mortality before their first birthday. Live births to women younger

than 20 years of age or greater than 35 years of age are more likely to result in an infant


All races







<20 20-24 25-29 30-34 35-40
Age (Years)

Figure 3 -- Infant mortality rates by age of mother, United States: 1996

mortality than those births to women between 25 and 35 years of age. Morgan (1996)

argues that states with an abnormally large group of teen mothers is the greatest predictor

of low birthweight births and the resulting infant deaths. This is true across races,

although the overall mortality rates for blacks are higher (these findings are constant

across time). Public policy programs have used these findings in several ways. The risk of

an infant death has been one of the primary messages used in teenage pregnancy

prevention programs. Additionally, this information has been provided to women trying to

balance their career and personal lives. As a woman chooses to delay childbirth, she will

increase the risk of a live birth resulting in a infant mortality. Public education of these

issues are important as women make certain choices in their lives.




14 -

112 -

I:1 ---;"------

14 -

---- Allraces ------- White Black
Oto 8 9 to 11 12 13to 15 16+
Educational Attainment (Years of Formal Education)
SOBECUS NaunitCmfor KHd6musBa uit*tcv o AIfanIdsna
Figure 4 -- Infant mortality rates by educational attainment of mother, United States:

A mother's educational status has also been examined to determine its impact on

the risk of infant mortality (Din-Dzietham and Hertz-Picciotto, 1998). Figure 4 shows

that the mother's educational attainment (as defined by the number of years of formal

education completed) is inversely correlated with the risk of infant mortality. As

educational attainment increases, the risk of infant mortality decreases. Educational

attainment's effect on mortality is likely related to the ability of the mother to better

understand her prenatal health care needs and the access to resources that come with

greater education. In addition, the results are confounded by several other environmental

factors. A woman under the age of 20 is unlikely to have more than some college

education; the pregnancy, itself, may be the reason for low educational attainment; and a

woman with low educational attainment may be plagued by poverty and the devastating

factors associated with it. Finally, the reduction in risk associated with higher maternal

educational attainment has been found to be of less significance for blacks, possibly

indicating diminishing returns on educational investment for this population (Din-

Dzietham and Hertz-Picciotto, 1998).

One of the most frequently studied factors affecting the risk of infant mortality is

the mother's use of tobacco during pregnancy. Smoking during pregnancy greatly

increases the risk of infant mortality. Figure 5 demonstrates that women who smoke

during pregnancy are twice as likely as those who do not smoke of having a live birth that

later results in an infant mortality. As a public policy issue, public health efforts in the

United States are heavily focused on educating women about the risk of smoking during


Another factor having a significant effect on reducing infant mortality is a woman

beginning prenatal health care during the first trimester of her pregnancy. Much of the

public health efforts around the prevention of infant mortality is focused on educating

women about the need for early prenatal care and providing access to these services for

* Smoking Nonsmoking

All races White Black

Figure 5 -- Infant mortality rates by smoking status of mother, United States: 1996

women unable to afford the health care in the private sector. Figures 6 shows the

devastating effects of late or no prenatal care. As this data indicates, infant mortalities are

three to four times more likely to occur if no prenatal care is utilized. For black women,

the risk of an infant mortality increases between four and five times if no prenatal care is

provided. The importance of these findings receives significant attention from many health

professionals, public policymakers, and women of child-bearing age.

Infant Mortality Rates By Start of Care: U.S. 1996
Trimester of Pregnancy Prenatal Care Began
First Second O Third M Nocar



All Races White Black

Figure 6 -- Infant mortality rates by trimester when prenatal care began, United
States: 1996

Figure 7 demonstrates that educational efforts advocating early prenatal care seem

to be effective in encouraging women to seek early prenatal care. The percentage of

women receiving late or no prenatal care in the United States is at its lowest point in

history. The proportion of women beginning prenatal care in the first trimester and

continuing regular visits throughout their pregnancies has increased between 1981 and

1995, with more than 23 million prenatal visits occurring in 1995 (Kogan et al., 1998).

Many of the public health initiatives associated with pregnancy have the promotion of

early prenatal care as one of their main objectives. This apparent success, unfortunately,

has not resulted in lowered rates of poor birth outcomes. This draws attention to a

paradox that will be discussed later in this research: interventions are being used but poor

birth outcomes remain the same or are getting worse.




4 ---- ----
All races ------ White -- -- Black

1985 1987 1989 1991 1993
1984 1986 1988 1990 1992 1994

Figure 7 -- Percentage of live births with late or no prenatal care, United States:

In addition to factors associated with the mother, there are many factors related to

the pregnancy and the infant, itself, there are associated with increased risk of infant

mortality. One of the factors is the sex of the infant. Research is still inconclusive about

why there is a difference between male and female infants' respective risks for mortality,

but the risk is present across time and races. Although it is a small difference, male infants

are at a statistically higher risk for death before one year of age (see Figure 8).

The infant's sex may have only a slight impact on its risk for infant mortality, but other

infant factors have some of the most significant impacts on this risk. Figure 9 shows how

one of these factors, short gestational age, may cause one of every two live births to result

in infant death. Across races, infants that are born at less than 28 weeks of gestational age

(calculated as the number of weeks since the start of the last menstruation), have almost a

50% chance of dying before its first birthday. The survival rate increases dramatically

after this point, as seen by Figures 9 and 10.

0 Male B Female

All Races White Black

Figure 8 -- Infant mortality rates by sex of infant, United States: 1996

<28 weeks U 28-36 weeks
37-41 weeks I 42+ weeks



200 -


All Races White Black

Figure 9 -- Infant mortality rates by period of gestation, United States: 1996

S28-36 weeks 37-41 weeks 1 42+weeks


All Races White Black
m:Figur Nfat rtait b rid of getation > 2 wee ite tate 99

Figure 10 -- Infant mortality by period of gestation > 26 weeks, United States: 1996

So, which factors have been found to have the most significant impact on infant

mortality? Figure 11 illustrates the five most common causes of infant mortality.

Congenital anomalies, or birth defects, are the primary cause of death for most infants;

however, this is not true across races. Within the black population, disorders related to

low birthweight birth or short gestation are the primary cause of infant mortality. Across

all races, low birthweight births and short gestation (premature birth) occur in 11% of all

pregnancies and are responsible for the majority of infant deaths in the neonatal period

(less than 28 days after birth) (Adams, 1995; Goldenberg and Rouse, 1998). For all races,

sudden infant death syndrome (SIDS), respiratory distress syndrome (RDS), and

conditions related to maternal complications round out the main causes of infant mortality

(Singh and Yu, 1995).

The effects of low birthweight births on the chance of infant mortality is an

important relation. When mortality rates are adjusted for factors such as social class,

maternal age, parity, and race, the greatest predictor of infant mortality is birth with a low

birthweight [less than 2,500 grams (or 5 lbs. 8 oz.) at birth] or a very low birthweight [less

than 1,500 grams (or 3 lbs. 5 oz.) at birth] (Newberger et al., 1976). Births under 500

grams have over a 90% mortality rate and infants born under 1,000 grams account for

between 50 and 60% of all neonatal deaths (Goldenberg and Rouse, 1998). The risk of

infant mortality for those infants with low birthweights between 1,000 and 1,500 grams

has decreased from around 50% in 1960 to about 5% in 1997. After birthweights of

2,500 grams, the risk of infant mortality reduces sharply until the trend begins to climb

again with very high birthweights (exceeding 4,500 grams) (see Figures 12 and 13).


2 50


All races White Black

Figure Innt mortality rates by leading causes of death, United States: 1996
Figure 11 -- Infant mortality rates by leading causes of death, United States: 1996






I 200

All Races






600-749 1 100-1240 1000-111 1 2S00-2999 300.-3999 4600-409
< 600 780-.l 1250-.1499 2000-2499 3000-3409 4000-440 15000+

Figure 12 -- Infant mortality rates by birthweight, United States: 1996

All Races ------ White -- Black



2 White -

---------- -- ------ -w hits -------
2800.2999 36 -30499 4600-489B
2000-24S9 3000-3499 4000-4499 6000*

Figure 13 -- Infant mortality rates by birthweights > 2, 000 grams, United States:

Low birthweight births are usually categorized into two broad categories: preterm

births and small-for-gestational age births. Preterm births are those births that occur

before the end of the 37th week of gestation. Preterm births can further be divided into

two categories: spontaneous preterm births and non-spontaneous preterm births

(Goldenberg and Rouse, 1998). Preterm births are a serious problem in the U.S. with

approximately 74% of all neonatal deaths occurring among preterm infants (Adams,

1995). Spontaneous preterm births are those births that follow both spontaneous labor

and spontaneous rupture of the membrane. To date, the best predictors of spontaneous

preterm births are: 1) a previous preterm birth; 2) black race; 3) low maternal

height/weight ratios (i.e., the mother being underweight for her height); 4)bacterial

vaginosis; 5) fetal fibronectin in the vagina or cervix; and 6) a short cervix a determined by

ultrasound (Goldenberg et al., 1998). Non-spontaneous preterm births occur when a

medical care provider decides that continuing the pregnancy would result in unacceptable

risk for either the mother or fetus or both. The most common reasons for non-

spontaneous preterm births are preeclampsia and presumed fetal growth restrictions.

Another form of low birthweight births are small-for-gestational births. These births have

birthweights lower than the 10th percentile for any gestational age based on some standard

population (Goldenberg, 1994). This condition usually results from a congenital anomaly

or from a slowing or temporary interruption of intrauterine growth.

How prevalent are low birthweight births in the United States? As of 1994, 788

babies a day were born with a birthweight under 2,500 grams (March of Dimes, 1997).

Figure 14 illustrates the pattern of low birthweight births by race in the U.S. between 1970

and 1995. As shown in Figure 14, the difference between whites and blacks remains

pronounced over the years examined; however, since 1990, black low birthweight birth

rates have attained a steady decline while white low birthweight rates have begun to

slowly increase from the all-time-lows seen during the 1980s.

The Impact of Low Birthweight Births

The most significant impact of a low birthweight birth is the death of the infant.

Low birthweight and preterm births (which are usually born with LBW) occur in between

7% and 15% of all births (depending on the population), yet they account for between

60% and 85% of all infant mortalities; 50% of long-term neurological morbidity has been

attributed to these births; and .5% to 1% of very preterm births (those born at less than 28

weeks gestation and less than 1,000 grams) account for 50% to 60% of all neonatal

mortalities and about one-third of the long-term morbidity (Goldenberg, 1994). There



- - -- Alrraces

M 40

Il 20

- All Races ------ White ------ Black
1971191 73 1975 1977 1979 1981 1983 198S 1987 1989 1991 1993 1995
1970 1972 1974 1t76 1978 1980 1982 1984 186 1988 1990 192 1994

Figure 14 -- Low birthweight birth rates in the United States: 1970-1995

are many other long-term effects for an infant born with a low birthweight. Low

birthweight births are more likely than other births to have both medical and development

problems. Low birthweight births are more likely to have medical problems such as:

respiratory problems (due to underdeveloped lungs); saline imbalances (causing possible

brain damage); hypoglycemia (low blood sugar); jaundice skin (which indicates liver

problems); iron deficiencies resulting in anemia; low body fat ratios (which make

maintaining body temperature difficult); bleeding into the brain (resulting in brain damage

or death); heart problems (the ductus arteriosus, a large artery which allows blood to

bypass the fetus' non-functioning lungs during pregnancies, may fail to close after birth);

inflammation of the intestine (necrotizing enterocolitis); and poor vision or blindness

(called reinopathy which results from an abnormal growth of blood vessels in the eye). In

addition to these medical problems, low birthweight infants tend to have learning and

other developmental problems.

Finally, there are immediate and extreme financial costs associated with childbirth

and especially with treating high-risk pregnancies and preterm/low birthweight births.

Health care associated with childbirth is one of the primary costs in the U.S. health care

system. The care associated with childbirth (or "perinatal care") includes: prenatal care,

delivery services (those provided to the mother at delivery), newborn care (services

provided to the infant from birth to the initial hospital discharge), and infant care (care

delivered from the initial discharge through the first birthday) (Long et al., 1994). The

perinatal care for each mother-infant pair averages approximately $7,000 per event for an

overall cost of more than $30 billion annually in the U.S. Research has indicated that for

every $1.00 dollar spent on prenatal care in the U.S., there are corresponding savings of

between $1.70 and $3.38 (this research makes a number of significant assumptions that

have been questioned in later studies) (Huntington and Connell, 1994). Feldman and

Wood (1997) found that high-risk pregnancies and the resulting births used a mean of

$20,933 (Mean +/- 1 SD of $38,391) in perinatal care, as compared to lower risk births

with a mean perinatal cost of $7,624 (Mean +/- 1 SD of $14,454). While low birthweight

births represent approximately 7% of all births in the U.S., they represent over 35% of

perinatal health expenditures or an additional $15,000 per birth (Becker et al., 1998).

Many public health prevention programs have justified their creation and continuing

funding with these cost-savings arguments.

Causes of Low Birthweight Births

There are many factors that have demonstrated an association with increased risk

of low birthweight births and, not surprisingly, many of these factors are similar to the

factors that affect infant mortality. By 1988, over 40 factors had been identified which

increase the risk of LBW, many of which are preventable and/or treatable (Sepkowitz,

1994). Technically, the primary causes of LBW births are preterm labor in developed

countries and intrauterine growth restriction (IUGR) in underdeveloped countries;

however, these conditions have underlying etiologies that must be taken into consideration

when looking at LBW (Sprague, 1993). Environmental factors such as mother's

socioeconomic status, age, and utilization of prenatal care have all been associated with

variations in birthweight. Additionally, conditions related to the pregnancy are also

factors that have been identified to explain low birthweight births. These conditions

include congenital anomalies, multi-fetal pregnancies (twins, triplets, etc.), problems

caused by placental attachment, and the mother's general health (including blood pressure,

diabetes, vaginal/intrauterine infections, organ problems, or a structurally abnormal

uterus/cervix). For classification purposes, the principle risk factors for low birthweight

births include: demographic variables; maternal medical risk predating pregnancy; maternal

medical risk in current pregnancy; maternal behavioral and environmental risks; and health

care services access risks (Sprague, 1993).

Demographic variables

There are many demographic factors that have been associated with increased risk

for low birthweight births. Women less than 17 or greater than 34 years of age have

increased risk for LBW births. Morgan (1996) argues that the single greatest predictor of

a community's LBW rate is the proportion of pregnant women under the age of 18

relative to the overall pregnant population. Race is a factor in LBW risk. In the U.S., a

baby born to black mother is three times as likely to suffer from LBW than white infants

and conditions relating to short gestation and LBW are consistently the primary cause of

infant mortality for black infants (Children's Defense Fund, 1992, Sprague, 1993, March

of Dimes, 1997). Marital status is a factor with unmarried women being at higher risk for

LBW births. Low educational attainment is associated with a higher risk of LBW births,

although blacks do not seem to obtain the same preventive benefit from increased

education as their white cohorts. Although each of these individual-level factors are

continually found to be associated with LBW risk, some researchers have argued that

larger, social risk factors may be just as important. These authors argue that macrolevel

social factors in the community in which the mother lives and works (i.e., per capital crime,

unemployment rates, average wealth, per capital income, etc.) have a significant impact on

the amount of risk for LBW births (O'Campo et al., 1997; Roberts, 1997). If this is

accurate, it suggests that larger community interventions may be necessary to address this


Maternal medical risks predating pregnancy

Certain medical histories may indicate a propensity for the mother to have an infant

born with LBW. Parity, or the number of previous live births to a particular woman, may

indicate whether the woman is likely to have an infant born with LBW. Women with no

previous live births or who have more than four previous live births are at increased risk

for LBW in future births. Genitourinary anomalies/surgery is related with an increased

risk of LBW births. The presence of certain chronic conditions (i.e., diabetes, chronic

hypertension) can increase the risk of a LBW birth. Nonimmune status for selected

infections (i.e., rubella) may increase the risk of a LBW birth primarily because the mother

has no protection from this illnesses which are known to have damaging effects on fetal

growth. The presence of a poor obstetrical history (i.e., previous preterm labor, previous

LBW births, infant mortalities) may indicate a propensity for poor birth outcomes in the

future. Finally, maternal genetic factors (i.e., LBW at own birth) may increase the risk of

a LBW birth.

Maternal medical risk in current pregnancy

Certain maternal conditions occurring during the current pregnancy are useful

indicators of an increased risk for low birthweight birth. Poor maternal weight gain during

pregnancy is associated with higher incidence of LBW births. Women are two to three

times more likely to have a LBW birth if they gain less than 10 kg (or 22 lbs.) during

pregnancy (Sprague, 1993). Because black women in the U.S. are twice as likely to gain

less than 16 lbs. during pregnancy, this factor is seen as one of the reasons for the

discrepancy between white and black birth outcomes. Short interpregnancy interval, the

time period between pregnancies, are associated with higher risk for LBW. Women who

space their children less than one year apart are more likely to have an infant born with

LBW. Hypotension, a condition in which the arterial blood pressure is abnormally low,

and pregnancy associated hypertension (preeclampsia), a condition in which there is an

elevation of arterial blood pressure beyond a normal range, are both associated with LBW

births. Preeclampsia, is related to certain conditions in pregnancy that are responsible for

most cases of induced preterm labor often resulting in LBW. Selected infections [i.e.,

symptomatic bacteriuria, rubella, and cytomegalvirus (a form of the herpes virus that can

cause birth defects)] contracted during the pregnancy can increase the risk of LBW births.

Problems with the placenta (i.e., placental previa, abruptio placentae) are factors also

associated with an increased risk of preterm labor and LBW births. Severe vomiting early

in the pregnancy, known as hyperemesis, may increase the risk of LBW. Abnormally low

or high levels of amniotic fluid (known as oligohydramnios and polyhydramnios) may

indicate poor fetal growth or birth defects which will likely result in LBW birth. Anemia,

or other hemoglobin problems, is associated with LBW births. Isoimmunization, the

development of antibodies to the developing fetus, is associated with LBW. Finally, a

structurally-abnormal uterine or cervix ("incompetent cervix") may increase the risk of

preterm labor and subsequent LBW.

Maternal behavioral and environmental risks

Many maternal behavioral and environmental factors are associated with a higher

risk ofpreterm labor and LBW births. One of the most significant risk factors for LBW

birth is the mother's use of tobacco during pregnancy. According to the U.S. Surgeon

General, in 1990, between 17% and 26% of all LBW births in the U.S. are attributable to

maternal smoking, on the basis of the average prevalence of smoking during pregnancy

and the overall relative risk of LBW birth for smokers (Barnett, 1995). Smoking greatly

increases the risk of retarded fetal growth and subsequent LBW birth and is the most

clearly established preventable risk factor associated with LBW (Sprague, 1993;

Hellerstedt et al., 1997). One study in Britain found an average difference in birthweights

between smokers and nonsmokers of 241 grams at birth (Brooke et. al., 1989). Because

this effect has been demonstrated in non-smoking pregnant women exposed to significant

amounts of "second-hand" smoke, the effect may even be larger if they are controlled for

in the comparison (Nafstad et al., 1998). Nicotine and carbon monoxide produced in

tobacco smoke contribute to chronic lack of oxygen to the uterus causing a reduction in

fetal growth (Sprague, 1993). The use of alcohol is also associated with fetal growth

problems and LBW birth. The reasons for this effect may include: 1) alcohol freely

passes to the fetus in concentrations equivalent to that in the mother's blood stream; 2)

fetal elimination of alcohol is less efficient, especially in the first trimester; and 3)

placental dysfunction is common in women who abuse alcohol (Sprague, 1993). The use

of other drugs during pregnancy increases the risk of birth defects and LBW births. The

use of certain drugs, cocaine for example, has been widely reported in certain areas (e.g.,

inner city hospitals) with up to 45% of pregnant women using illicit drugs (Kline et al.,

1997). How substance abuse negatively affects fetal development is unclear and

confounded by environmental and behavioral factors that in themselves are correlated with

an increased risk of LBW. For example, it is unlikely that a pregnant woman using

cocaine is only using this substance; it is unlikely that she will enter early prenatal care for

fear of detection; and it is unlikely that she will adhere to a balanced diet. A factor closely

associated with alcohol and substance abuse is violence toward the pregnant woman and

her fetus. Various studies indicate that between 1% and 20% of women experience

physical violence during pregnancy, as compared to between 9.7% and 29.7% in the

female population as a whole (Ballard et al., 1998). Women who are abused are less likely

to access prenatal care, are more likely to have multiple injury sites, and have significant

abuse directed to the abdomen. Poor nutrition has been associated with increased risk of

LBW, in particular the deficiency of certain minerals and vitamins. While it is known that

certain vitamins and minerals (i.e., folic acids, iron, zinc) are important during pregnancy,

the necessary amount of nutrients and the timing of their introduction is still unclear. In

general, caloric intake in itself seems to be as important as the dietary level of any

particular vitamin or mineral. Psychological conditions (i.e., stress, anxiety, depression)

are also associated with a higher risk of LBW. The theories behind the effect of

psychological factors on pregnancy are rooted in the understanding of the role of

catecholamines (the hormones that are released during stress causing the "fight or flight"

response) in the body (Sprague, 1993). The belief is that these hormones can cause early

uterine contractions and impaired fetal growth. Finally, several other external factors have

been associated with increased chance ofpreterm labor including: high heat-humidity

index; working conditions; physical activity (exercise), etc. (Lajinian et al., 1997).

Health care services access risks

The inability or lack of desire to access early, comprehensive prenatal care has

been associated with an increased risk of LBW births. Research continuously

demonstrates that there is a positive linear association between month of initiation of

prenatal care and low birthweight rates. Low birthweight rates increase as the month of

pregnancy when prenatal care began increases. Furthermore, there is a inverse correlation

between the number of prenatal visits received and low birthweight rates. As the number

of prenatal visits increase, low birthweight rates decreases (Kotelchuck, 1994).

Proponents argue that early and regular prenatal care provides opportunities to: 1) identify

certain maternal risks; 2) educate the woman about environmental and behavioral risk

factors; 3) implement specialized care, and 4) detect preterm labor at an earlier point.

Considerable resources have been put into the provision of prenatal services and it is a

component of most social programs to reduce infant mortality and poor birth outcomes

rates. However, the substantial increase in private and public spending for prenatal care

over the last two decades has not resulted in substantial decreases in the birth outcomes

rates, resulting in a paradox (Mustard and Roos, 1994; Sepkowitz, 1994)). Research is

still unclear about what particular aspects of prenatal care are most essential. Kogan et al.

(1994) found that women who reported receiving insufficient health behavior advice as

part of their prenatal care were at higher risk for a LBW birth. In another study,

researchers indicated that the physician's qualifications and experience has an impact on

birth outcomes (Haas et al., 1995). Still other research has shown that the woman's

proximity to the obstetrical care, irrespective of actual utilization, has an impact on LBW

births with rural women who travel out of town to receive the care have poor outcomes

(Nesbitt et al., 1990; Nesbitt et al., 1997). Differences in the ability or desire to access

health care services has been indicated as a possible explanatory factor for the racial

disparity in poor birth outcomes. In one study, 63.4% of white women received adequate

prenatal care as measured by the Prenatal Care Utilization Index compared with only

51.9% of black women (Kotelchuck, 1994). In fact, black women are less likely to

practice primary preventive health behaviors than white women in general (Duelberg,

1992). As with many risk factors examined, research on prenatal care utilization and birth

outcomes is hindered by methodological problems. One of the key problems is that there

are differences in the way women seek and use prenatal care, and these differences are not

random and are difficult to measure (Frick and Lantz, 1996; Liu, 1998). Since, for ethical

reasons, we are unable to conduct randomized control studies assigning women to

different levels of utilization, conclusive link to improved birth outcomes may never be

known. As a result, a public health policy that heavily relies on the promotion of prenatal

care utilization to reduce the incidence of LBW births does not have conclusive empirical


Medical and Psychosocial Interventions

Many interventions have been developed with the hope of reducing low

birthweight births and the resulting infant mortalities. These interventions can usually be

divided into one of three categories: medical, psychosocial, or a combination of

interventions. The medical interventions include: utilization of prenatal health care;

surgical interventions to prevent preterm births; pharmaceuticals to stop preterm labor;

nutritional interventions; bed rest and hydration; and efforts to reduce vaginal infections.

Psychosocial interventions include: risk-based scoring systems to determine service needs;

cessation programs for tobacco, alcohol, or other drugs; psychological services to reduce

stress, anxiety, and/or depression; and economic assistance programs. Each of these

interventions need greater description in order to understand the rationale for using the


Medical Interventions

The use of medical interventions to address problems in birth outcomes are often

referred to as a "medical-model" approach. (Sprague, 1993). The medical-model

approach assumes that poor birth outcomes can be eliminated through medical

interventions at the individual level. While this model acknowledges that there are social

factors that increase the risk of a poor birth outcome, it advances the use of medical

technology, pharmaceuticals, and information transfer in the provider-patient relationship

as the appropriate venue for reducing these risks.

Early and comprehensive prenatal care is the cornerstone of the medical-model

approach to reduce the risk of low birthweight births. The basis for this approach is

research that indicates that women who receive early prenatal care or have more prenatal

visits are less likely to have a low birthweight birth. Additionally, the Institute of Medicine

estimates that for every dollar spent on prenatal care there is a corresponding savings of

$3.38 resulting from the reduction of medical cost associated with low birthweight births

(Children's Defense Fund, 1992). The medical care required to care for an infant born in

the U.S. cost an average of $2,500 per day, or approximately $8,000 per neonatal

intensive care unit admission (Sprague, 1993). Although the correlation between late or no

prenatal care and increased LBW births has been repeatedly demonstrated, this is not to

say that there is a causal link between early/extensive prenatal care and birthweight.

Goldenberg and Rouse (1998) conducted an extensive review of prenatal care

interventions that changed the inception or intensity of prenatal care and found little

impact on LBW rates. The authors concluded,"...because the enhancements to prenatal

care have varied from study to study and because the associated reductions in preterm

births have been inconsistent, it is not clear which specific additions to prenatal care, if

any, are likely to result in a reduction in preterm births." Other researchers conducting

meta-analyses of the research literature on prenatal care and low birthweight have

concluded that early prenatal care has not consistently reduced the incidence of fetal

growth retardation, very low birthweight, or very preterm delivery (Alexander and

Korenbrot, 1995). They conclude by stating that changes in provider practice patterns and

public policy are needed to allow the full benefit of early pregnancy prevention to be


Surgical interventions (cervical cerclage) are sometimes used to prevent preterm

delivery. Between 1 in 200 and 1 in 1,000 pregnant women are diagnosed with a

structurally weak, or "incompetent", cervix. This diagnosis is usually based on a medical

history of spontaneous second-trimester birth without the detection of uterine

contractions. In later pregnancies, the physician may choose to place one or more

circumferential stitches (cerclage) in the cervix. Although common, research has been

unable to conclusively determine the effectiveness of the procedure because the histories

of women undergoing the procedure are difficult to differentiate between an incompetent

cervix or unrecognizable preterm labor (Goldenberg and Rouse, 1998).

Nutritional interventions are often used to reduce the risk ofa LBW birth based on

research demonstrating that women who are underweight at the beginning of pregnancy

and those who gain little weight during the pregnancy are at increased risk for preterm

birth and LBW (Goldenberg and Rouse, 1998). These interventions usually involve a

combination of: counseling; protein supplementation; caloric supplementation; and vitamin

and mineral supplementation. Little proof exists to show that nutritional counseling

affects the eating behaviors of pregnancy women and even less evidence that it prevents

preterm births or LBW. Contrary to research showing that women with low-protein diets

are at higher risk for preterm births, programs that promote protein supplementation often

result in adverse outcomes. One of the most prominent U.S. programs designed to

prevent preterm births and LBW births is the Special Supplementation Program for

Women, Infants, and Children (WIC). The basis of the program is to provide calorically-

enriched diets to low income pregnant women. This, and similar programs, have

demonstrated small increases in birth weight. These programs do not tend to prevent

preterm birth, but do effect fetal growth. Finally, vitamin and mineral supplementation

interventions are based, once again, on research demonstrating that anemia, low maternal

zinc blood levels, the absence of folic acid from maternal diets, and other vitamin

deficiencies are associated with congenital anomalies, retarded fetal growth, LBW, and

preterm births. Programs supplementing vitamins and minerals have had mixed results,

often depending on the initial deficiencies of the pregnant woman.

Tocolytics, a class of pharmaceuticals, are used to interrupt or stop uterine

contractions in the event of spontaneous preterm labor (Goldenberg and Rouse, 1998).

These drugs include: beta-mimetic agents; magnesium sulfate; calcium-channel blockers;

oxytocin antagonists; and nonsteroidal anti-inflammatory agents. Although widely used,

these drugs have only been shown to effective for a relatively short period of time

(approximately up to 48 hours). With such a temporary interruption of preterm labor,

what is the benefit of using these interventions? Most health care professionals agree that

tocolytics, alone, are beneficial for "buying" the extra time necessary to transfer the

mother to a facility specializing in methods to care for preterm deliveries (Sprague, 1993;

Bronstein et al., 1998). Benefits are also gained when tocolytics are used in conjunction

with corticosteroids (hormones that help the lungs work more effectively). The most

commonly used corticosteroid for this purpose is surfactant. By 1991, both synthetic and

animal surfactant were approved for use in the United States (Strobino et al., 1995). The

tocolytic agents may help increase the antenatal period long enough for the corticosteroids

to reduce the risk of neonatal respiratory distress syndrome, intraventricular hemorrhage,

and the resulting neonatal mortality. Additional randomized studies are needed to prove

the effectiveness of the using these drugs in coordination3.

Bed rest and hydration has been advocated for years for women facing preterm

labor. Like many of the other interventions discussed, bed rest and hydration has little

evidence to support it as an effective intervention during preterm labor. Ironically, as with

many of the other commonly used techniques, research has demonstrated that there may

actually be an increased risk of preterm labor for women using this intervention. In

addition, pregnant women proscribed bed rest are at increased risk for other adverse

outcomes including venous thrombosis and pulmonary edema (Goldenberg and Rouse,


Treatment of vaginal and intrauterine infections is another intervention based on

research showing women with certain infections are at increased risk for preterm labor.

Up to 80% of preterm births are associated with the presence of vaginal and/or

intrauterine infections, although programs designed to introduce antibiotics in women with

histories of preterm labor have proved ineffective in preventing early labor in later

pregnancies (Goldenberg and Rouse, 1998). Irrespective of supporting evidence,

interventions to reduce infections are becoming more common in prenatal care programs.

Psychosocial Interventions

In the last two decades, there has been a movement away from a purely "medical-

model" approach to preventing poor birth outcomes to a more holistic approach that

3The use oftocolytics and corticosteroids have been indicated as factors explaining the reduction of infant
mortality rates in the U.S. However, while infant mortality rates are declining overall, white mortality
rates are improving at a faster rate than blacks. Recent research indicates that differences in the use of
these drugs may be the reason for the growing disparity between whites and blacks with African American

includes psychological and social (psychosocial) interventions. Rosenblatt (1989) argues

that the medical model is ill-suited for large social problems like low birthweight births.

Additionally, he argues that because our current obstetrical style uses a "maximin

approach" (manage the patient to prevent the worst possible outcome), the care focuses

attention on the search for deviation from the physiological norm and diverts attention

away psychosocial interventions that may be more appropriate. These interventions

usually include cessation programs, psychological counseling, economic assistance, family

planning, domestic violence prevention, and home visits by health or social workers


Cessation programs for tobacco, alcohol, or other drugs are often advocated for

the prevention of low birthweight births, and most psychosocial intervention programs

include cessation programs. Research has demonstrated that as the overall prevalence of

smoking declines in the population, there is a corresponding decrease in LBW rates

(Cnattingius and Haglund, 1997). While the use of these substances during pregnancy is

linked with restricted fetal growth and congenital anomalies, the effectiveness of these

programs in reducing low birthweight births at the individual level has not been shown.

Part of the problem in determining program effectiveness in reducing poor birth outcomes

is that they often result in low cessation rates. Persons using these substances during

pregnancy rarely stop completely, although some do reduce their use. This is especially

true for tobacco cessation programs. Psychological services to reduce stress, anxiety,

and/or depression are also included under many psychosocial intervention programs. The

provision of these services is based on research that indicates that women with unusual

mothers and infants less likely to receive the interventions. The reason for the disparity in drug utilization

stress, anxiety, or depression have an increased risk of low birthweight births. Like many

of the other interventions, research has failed to demonstrate that these interventions help

reduce the overall rates ofpreterm birth or low birthweight births. Finally, economic

assistance programs have been advocated as a way to ensure women receive many of the

previously mentioned psychosocial interventions through the private sector if public

assistance interventions are not available or limited. Unfortunately, research indicates that

economic assistance programs, and psychosocial interventions as a whole, do not seem to

have a positive impact on poor birth outcome prevention (Villar et al., 1992; Sepkowitz,


Combined Medical and Psychosocial Interventions

Because of the correlation between the various factors associated with preterm

labor and LBW births, many of the programs designed to reduce the risk of both often

include a wide variety, if not all, of the intervention discussed in this section. Many of

these programs are based on either a "risk-scoring" system or a community-wide "full-

population" approach in order to determine resource allocation.

Risk-scoring systems, where either self-reported factors are collected through a

special survey instrument or information is obtained during the initial prenatal visit, are

often used to determine how many factors correlated with increasing the risk of low

birthweight births are present for a given pregnancy. Each factor identified is usually

given a weighted value (i.e., smoking status may have a higher score than mother's age)

and a score is determined. These systems usually have a scoring "threshold" under which

the pregnancy is seen as having a low risk for LBW and under which expanded services or

patterns is still unknown (Bronstein et al., 1998).

interventions are not offered. Conversely, scores that surpass the threshold indicate a

significantly higher risk of poor birth outcome and usually qualify the mother for special

assistance considered essential to reduce this risk. In general, these systems tend to have

the problem of identifying too many women as high-risk who do not eventually have a

LBW birth, and fail to identify many women who are not considered high risk and yet

have a LBW birth (Sprague, 1993). After reviewing these systems, Goldenberg and

Rouse (1998) argue that they tend to consistently identify those pregnancy with two or

more times the risk of a normal pregnancy but offer little in ensuring better health

outcomes. The authors go on to argue that scoring systems seem to only increase the use

of interventions with unproven effectiveness or the use of unwarranted interventions at a

high financial cost and possible negative impact.

Full-population programs are focused at community-level interventions.

Proponents of the community-based intervention programs argue that it is more logical to

affect larger societal and environmental factors in preventing LBW births, than directing

increased resources to a particular individual. These proponents use an analogy called the

"upstream-downstream" model to describe the benefits of a community-wide approach

(Sprague, 1993). The upstream-downstream model describes a problem of a river in

which persons continually fall in and need saving. To solve this problem, three points of

intervention could be utilized. First, rescuers can rush into the water to save the individual

as they come downstream, using whatever resources are needed, many which are very

expensive. This is the equivalent of treating LBW infants after birth to prevent potential

mortality. The second intervention point is to go upstream to the point where people are

falling into the river and either teach the people how to swim or catch them before they

fall into the water. This is equivalent to identifying and providing services to high-risk

women through the risk-assessment programs. Finally, the third intervention point is to

either fix the bank where people are slipping in, build a barrier between the walkway and

the river, or build a bridge over the river to help people cross. In doing so, the

community is provided with protection and people, irrespective of their risk, benefit

equally from this intervention. The third point of intervention is the equivalent of the

community-wide interventions for preventing LBW births. Community-wide interventions

to prevent LBW births usually support universal screening of pregnant women for risk

factors and allow all women to access any services provided. The belief is that all women

can benefit from the services and that a broad approach to interventions will capture

pregnancies that would have been categorized as "low risk" yet which may result in LBW

births. Because until recently the medical model of interventions has been the primary

choice in the U.S., there has been little research examining the effectiveness of

community-wide approaches. What is known is that the U.S. continues to expend

considerable resources to prevent preterm/LBW births and yet conditions remain the same

or have worsen within some populations.

The Perinatal Paradox

Before examining federal and state approaches to coping with poor birth

outcomes, it is important to summarize the paradox that is presented by the current

research on this issue. Research on the causes of LBW birth and the interventions that

have been developed to combat the problem make it clear that there is a disconnect

between the knowledge and the implementation of the interventions. This disconnect has

been noticed by other experts as well and called the "perinatal paradox" (Rosenblatt,

1989; Sepkowitz, 1994; Goldenberg, 1994; Kliegman, 1995; Kramer et al., 1998). Many

researchers and child health policy advocates point out that although the amount and types

of interventions have increased, birth outcome indicators remain the same year to year,

and indicators for some populations have worsened. While expenditures for obstetric and

neonatal care continues to rise, neonatal intensive care units (NICU) continue to

proliferate, and regionalized perinatal care programs become commonplace, poor birth

outcomes continue (Rosenblatt, 1989). Goldenberg states ".. .despite the utilization of

ever increasing amounts of research and clinical care resources aimed at achieving a

reduction in low birthweight, there is limited evidence to date that any intervention or

practice has had a major impact on preventing preterm birth or growth retardation..

.neither the behavioral [psychosocial] approaches (smoking or drug cessation programs,

nutritional counseling or supplementation, provision of social support, etc.) or medical

approaches (enhanced prenatal care, tocolytic agents, etc.) adopted to date have had a

large or even statistical impact on the rate of preterm birth or growth retardation" (1994,

p. 630).

Most research that has demonstrated a change in infant mortality rates attributes

the change to preterm infants' increased chances for survival, not in the overall reduction

of LBW or preterm labor rates. In fact, research indicates that neonatal technology has

been a driving factor in perinatal survival. More than 35 studies have shown that the use of

certain corticosteroids reduces the chance of infant mortality by between 30% and 40%.

Up to 60% of infant mortalities attributed to respiratory distress syndrome have been

eliminated with their use. The primary corticosteroid intervention, exogenous surfactant

therapy, "consists of artificial or animal-derived surfactant (a combination of lipids with

surface tension-lowering properties) being administered through an endotrachael tube

directly into a newborn's lung, usually within minutes of the birth"( Kliegman, 1995).

Kramer et al. (1998) suggest that the increasing poor birth outcomes rates are actually

related to changing medical practices. The authors look at preterm births in a Canadian

hospital over a eighteen-year period beginning in 1978. They found that the increased

rates of premature births were "largely attributable to the increasing use of early

ultrasound dating, preterm induction and preterm cesarean delivery without labor, and

changes in sociodemographic and behavioral factors."

In summary, the "perinatal paradox" appears to be the result of three interacting

phenomena. First, multiple preventive interventions are being used more widely than ever

before, although there appears to be little evidence that any of these interventions have

had any impact on preventing preterm or LBW births. Second, what reduction in poor

birth outcomes have been identified seem to be related to survival rates ofpreterm/LBW

infants, rather than prevention efforts. Finally, physicians and other medical providers

appear to be willing to risk premature, induced labor in certain pregnancies because of the

availability of medical interventions which increase the survivability of these infants

(Joseph, et al., 1998). If these assumptions are accurate, it is unlikely that poor birth

outcome rates will decrease in the future (irrespective of preventive interventions used in

society) as even more technological developments allow infants to survive at smaller

birthweights and extreme prematurity.

Programs to Reduce LBW Births: National and State

As previously discussed, the prevention of LBW births and the resulting infant

mortalities have been primary health policy issues in the U.S. for decades. This being the

case, there is still a long way to go. Every day in the United States, 81 infants die, 443

infants are born to mothers who had late or no prenatal care, and 781 infants are born with

a birthweight under 2,500 grams. With these statistics, it is understandable why

intervention programs are continuously created and maintained as a primary public health

policy issue. This section examines the history of child health programs in the U.S., the

current federal intervention programs, and similar state policy innovations dealing with

poor birth outcomes.

Early Federal Intervention Programs: Turn of the Century

Poor birth outcome rates, while still considered a serious problem in the U.S., are

considerably lower than they were at the turn of the century. In 1900, 10% of infants died

in the first year of life. The primary causes of infant death at the turn of the century were

infectious diseases and diarrhea resulting from poor water and sanitation conditions.

Urban areas were affected the most with rural infants having a higher chance for survival

to their first birthday. Not surprisingly, the first federal and state interventions affecting

infant mortality rates involved sanitary reforms in the cities. The resulting decline in infant

mortality over the next thirty years was also partially attributed to the pasteurization of

diary products, educational programs that taught childcare to new mothers, birth control

education (although still illegal in much of the United States) which helped women to

space their births, and the development of the first antibiotics and immunizations. In 1921,

Congress passed the first federal maternal and child health program, the Sheppard-Towner

Maternity and Infancy Act (Cooper, 1992).

Federal Maternal and Child Health Policy and the New Deal: 1935-1953

The next significant change in federal policy resulted from the passage of the

Social Security Act of 1935. Part of the "New Deal" programs introduced by the

Roosevelt Administration, Title V of the Social Security Act provided grants to the states

to improve maternal and child health and Title IV-A included a provision entitled Aid to

Dependent Children (ADC). As originally designed, the ADC program provided direct

financial support to orphans or children cared for by a widowed mother, or when the

father was unavoidably absent from the home. Coupled with the earlier provisions of the

Sheppard-Towner Act, these new programs helped states to increase health care system

capacity for maternal and child health (Schlesinger and Kronebusch, 1990). Although the

Department of Health, Education, and Welfare (later renamed the Department of Health

and Human Services) was created and elevated to a cabinet position in 1953, federal

health policy related to maternal and child health continued under these programs through

World War II and did not receive priority attention again until the social welfare initiatives

of the 1960s.

Federal Maternal and Infant Health Policy and the Great Society (1962-1965)

During the 1960s, three of the most important federal policy initiatives related to

maternal and child health were created or modified from earlier policies. These policies

include: 1) changes in the ADC program; 2) creation of Maternal and Infant Care (MIC)

Projects; and 3) the creation of Medicaid. Because most of these programs are

partnerships between the federal government and the states, these three programs became

the cornerstone of maternal and child health policy in the U.S. Each one of the programs

will be discussed in greater detail.

Aid to Families with Dependent Children (AFDC)

More easily recognized as the Aid to Families with Dependent Children (AFDC)

program, the ADC program was funded by the federal government and administered by

the states. While initially focused on general child welfare, eventually AFDC came to

include assistance for medical care, cash financial assistance, and other social services to

children in need because of a major family crisis such as divorce or the death, disability, or

desertion of a parent. The program, in conjunction with Medicaid, became an important

source for support and funding for medical care for pregnant women and children (Miller

et al., 1990). Because states determined many of the requirements of the program, there

was wide variability among the states in who receives what level of assistance. Many

states even allowed pregnant women to apply for AFDC benefits on behalf of the fetus. In

addition, only half of the states joined the program prior to 1970, with the number of

participating states declining over the years.

The AFDC program expanded twice since its creation. In 1962, the Kennedy

Administration proposed and signed into law changes to the Social Security Act

expanding eligibility to include payments to the mother in the family and making AFDC a

separately-funded program. This first expansion was approved because of the argument

that "if assistance is provided only to children, it is inconceivable that the mother or other

adult caring for them will forego food or other necessities in order that the money will

only benefit the children" (Gray et al., 1990, p. 421). The second expansion occurred in

the 1980s allowing payments to go to families with able-bodied, unemployed adult males.

As a result of the two expansions, federal eligibility requirements for AFDC stated that an

eligible family must have a dependent child who is: 1) under the age of 18; 2) deprived of

parental support because of a parent's death, continued absence, incapacity, or the

unemployment of the principal family earner in a two-parent household; 3) living in the

home of a parent or other specified, close relative; 4) a resident of a state participating in

the program; and 5) a U.S. citizen or an alien permanently and lawfully residing in the U.S.

This second expansion changed removed the program so far from its original

purpose that it became an issue of considerable public debate. While originally designed

to provide direct financial support for food, housing, and medical care to orphans and

children in families without a father, the second expansion of the program allowed

payments to family units with unemployed males and those headed by mothers who had

the children "out-of-wedlock." For this reason, AFDC became one of the most costly and

controversial social welfare program in the U.S. (Bowman and Kearney, 1990). Even in

the 1930s, before the more controversial expansions, one legislator from Illinois

consistently referred to AFDC as the "aid to bastard children program" (Gray et al.,

1990). Critics argued that the people should not receive public funds if they are able to

work and funds should not be used to support families in which the men were unwilling to

provide support. Other critics argued that the design of the program gave an incentive for

married couples to divorce or for couples not to get married in order to obtain or keep

their benefits. Finally, opponents to the program believed that its design provides a

disincentive for people to look for jobs, because they could lose the benefits yet make less

money in the lower-income jobs available to the recipients (Wilson, 1992). Some statistics

support these arguments. Between 1964 and 1969, the number of "welfare mothers" grew

by more than 60%, and by the 1990s over 90% of AFDC payments were to single-parent

households in which the living father was divorced or separated from or never married the

mother (Wilson, 1992; Bowman and Kearney, 1990). The percentage of AFDC children

who were recipients because the parents never wed increased from 33.8% in 1977 to

48.9% in 1986 and the percentage of adults receiving assistance that are women was

approximately 88% by the end of the 1980s (Gray et al., 1990). The AFDC program

continued to be a political "hot-potato" throughout the 1970s and 1980s. With public

support for AFDC never high and with support declining rapidly through the early 1990s,

the Clinton Administration and the 104' Congress pushed through significant welfare

reform in 1996 (P.L. 104-193) effectively eliminating the AFDC program.

Maternal and Infant Care (MIC) Projects

Created in 1963, Maternal and Infant Care (MIC) projects consist of federal

funding to states to reduce poor birth outcomes in impoverished, urban areas. The

program was designed to provide actual prenatal and maternity services in these areas.

These services are provided to eligible pregnant woman and infants up to one year of age.

Maternity and medical services provided through MIC projects include: diagnostic and

preventive services; nursing assessments; nutritional assessments and counseling; case

management; medical exams; physician and nurse midwife visits; family planning services;

delivery support assistance; and parenting classes.

MIC projects, at one point, provided some of the most comprehensive, direct

service programs available for pregnant women and their infants. MIC projects were not

as widespread as later programs (i.e., Medicaid, WIC) and there were only 56 independent

projects in the U.S. at the height of the program (Miller et al., 1990). Few of the projects

are still in operation today.


Following the first expansion of the AFDC program, the federal government

embarked on a more direct approach to provide aid to pregnant women and their infants.

In 1965, the Johnson Administration proposed significant amendments to the Social

Security Act. These amendments resulted in the addition of Title XIX to the act,

commonly called Medicaid. Another joint program funded and administered by both the

federal and state governments, Medicaid requires certain basic medical services be

provided to low-income individuals including: inpatient hospitalization; outpatient hospital

services; laboratory and radiology services; physician services; and maternal care. The

program is the primary source of funding for long-term institutional care for the physically

and mentally disabled. Specifically related to maternal and infant care, Medicaid provides

free prenatal and hospital care, and postnatal care for child and mother if needed. Persons

already receiving AFDC or Supplemental Security Income (SSI) are automatically

qualified for Medicaid

As part of the federal-state partnership, states have the ability to change eligibility

within certain parameters set by the federal government and may select additional benefits

not required by the federal government. States may also determine the scope of services

that can be offered and may require limited co-payments, deductibles, or coinsurance

(Gray et al., 1990). The program also allows certain people who would not otherwise be

eligible for public assistance, to receive public assistance for medical care. This optional

"medical needy" program is being offered by more than half of the states, and has become

the leading provider of perinatal care for low-income women and children.

Federal Programs for Women and Infants and Republican Administrations: 1972-1991

Federal interventions to improve pregnancy outcomes and prevent infant mortality

and morbidity were not only Democratic initiatives. Between 1972 and 1991, several

changes were made in the existing federal programs and new ones were signed into law by

Republican administrations. At the same time, spending by the federal government was

shifted to greater administration by the states and declined overall, and many states

developed their own programs in response.

Special Supplemental Food Program for Women. Infants, and Children (WIC)

Aimed at reducing malnutrition (a factor linked to low birthweight births and poor

maternal weight gain) the Nixon Administration promoted the inclusion of a $20 million

amendment to the Child Nutrition Act of 1966. The amendments passed and established

the Special Supplemental Food Program for Women, Infants, and Children, or more

commonly known as WIC. Administered by the U.S. Department of Agriculture (USDA),

the program works through state health departments to distribute food and food vouchers

on a monthly basis to pregnant women, nursing mothers, their infants and children up to

the age of five. The programs were quickly combined with other programs focusing on

maternal and child health. Screening systems in either WIC and one of the other programs

usually help to establish eligibility for others. It is not uncommon for pregnant women and

their children to be receiving services from AFDC, SSI, Medicaid, and WIC all at the same


By 1989, over three million people were enrolled in WIC at a cost of almost $2

billion annually (Miller et al, 1990). In response, the USDA argues that the program has a

high return on investment saving as much as $3 in health care expenses for every dollar

spent in WIC (Cooper, 1992). Many studies have examined the impact of WIC in

preventing infant mortality have shown the program has limited success (Kotelchuck et al.,

1984; Ahluwalia et al., 1998; Moss and Carver, 1998). The program seems to have the

most significant benefit for women who are poor, black/or in their teenage years.

There are criticisms of the WIC program. WIC may have helped some women,

but overall participation in the program of those eligible is less than one-fourth. Critics

argue that this may be related to the social stigma that has become attached to the

program. Additionally, critics argue that the program is more an agricultural subsidy than

a maternal and child welfare program. The foods directly provided by WIC come

primarily from agriculture surpluses. Nutritionists see the value of the foods provided to

women and their infant as high in fat (i.e., milk, cheese, etc.) and low in fruits and

vegetables. The high fat foods are effective in promoting weight gain in women and

infants, but the long-term health effects of a high-fat diet make the program's claim to

preventing LBW births of questionable benefit, especially for the mother.

Improved Pregnancy Outcomes (IPO)

In 1976, the federal Bureau of Child Health Services initiated the Improved

Pregnancy Outcomes (IPO) program as part of its overall child health strategy. The IPO

projects were designed to develop "state-based systems of care for mothers and children."

The allocation of the funds to the states is based on the severity of the infant mortality and

morbidity rates. The funds were designed to encourage states to upgrade their systems of

perinatal care. The first IPO projects were directed to indigent pregnant women trying to

help them access comprehensive outpatient prenatal care, delivery services, and

postpartum care (Miller et al., 1990). States had considerable flexibility in designing their

programs and some, like Florida, incorporated their existing MIC projects into the new

program. Because of their high infant mortality rates compared to other counties in

Florida, the following eight counties were selected to receive the funds: Despot, Glades,

Hardy, Hendry, Lake, Lee, Marion, and Putnam

When the federal funds were cut in 1982, the Florida Legislature approved using

the state's general revenue funds to support and expand the program throughout the state

(Clarke et al., 1993). By 1985, all of the counties in Florida had operational IPO

programs. With this expansion of the program in Florida, its purpose was defined as: "the

development of health programs throughout the state which improve the health of mothers

and children, improve the outcome of pregnancy of mothers and children, and improve the

health of newborns, infants, preschool, and school-age children, adolescents and young

adults. The mission is to identify, organize, and coordinate all available resources to

develop intervention programs which will ultimately result in the prevention of morbidity

and mortality in mothers and children" (Miller et al., 1990, chapter 5, p. 6). The basic

services provided through this expanded program included: prenatal care provided by

nurses, midwives, and physicians; nutritional counseling; parenting classes; home visits;

limited delivery arrangements; "well-child" check-ups; family planning; WIC; and

immunization services.

The efficacy of these programs in many states has been difficult to determine.

Research on the IPO programs in Florida after the state took responsibility for funding has

demonstrated that the programs may have only marginal impact on infant mortality (Miller

et al., 1990; Clarke et al., 1993). Most research indicates that IPO program have

increased the number of prenatal visits that these low-income women receive, however,

there has been little evidence of a corresponding decrease in poor birth outcomes. Some

researchers have argued that the programs are effective but that the women are in worse

condition than the general population, so comparison of their relative birth outcomes is

not valid. The researchers believe that the true impact of the program is "masked" by the

poor condition of the women in the program (Miller et al., 1990). However, based upon

closer examination of these studies, it could be argued that a larger issues has been

overlooked. In order to determine the impact of the program, Miller et al. (1990) and

Clarke et al. (1993) created a matched comparison group of IPO participants and other

pregnancies with similar sociodemographic characteristics as the IPO group. In

comparison, the IPO group has slightly better outcomes. What is assumed by the authors

is that the relative risks are the same between both groups (or that the IPO group may

have worse health conditions upon entering the program). In reality, it is possible that the

comparison would be more likely to have poorer health status. When one considers that

members of the comparison group are unlikely to have private health insurance, the

question has to be raised as to why they would refuse to participate in a free health

program like the IPO program? One possibility is that these individuals have certain

health behaviors (e.g., illicit drug use) or other legal issues that may be a deterrent to

seeking public health official interventions. These risk factors would appear to put this

group at a significantly higher risk for poor birth outcomes.

One finding in Miller et al.'s (1990) study which received no attention in later

published research (Clarke et al., 1993) on the IPO program was related to the authors use

of time series analysis (ARIMA) to examine changes in the LBW rates. The authors found

a seasonal pattern in LBW rates in nonwhite women in Florida. The authors only mention

the finding as part of the methodology discussion. As will be demonstrated later in this

research, this phenomena does not simultaneously occur in the white population in Florida.

This raises a number of questions because the seasonal differences could relate to

increased infant deaths in nonwhite births at certain points of the year.

Maternal and Child Health Block Grants

The Omnibus Budget Reconciliation Act of 1981 was the next significant change

in maternal and child health policy. The Act changed many aspects of the Medicaid

program including amendments to Title V of the Social Security Act establishing Maternal

and Child Health Block Grants for states to provide health care to needy women and

children. States were required to provide matching funds equal to 75% of the federal

allocation. At the same time, welfare eligibility for the working poor was tightened and

many lost Medicaid coverage. During the late 1980s, federal funding for the MCH grant

fell by over 40% (when adjusting for inflation). Because of the cuts and the approaching

recession, services overall were reduced in most states by the end of the 1980s.

Medicaid Eligibility Expansion

Prior to 1989, states were required to offer Medicaid coverage to pregnant women

and young children with a family income less than 133% of the federally-established

poverty level ($18,500 for a family of four in 1989). Congress took action based on two

primary assumptions: 1) financial limitations produced barriers to prenatal care; and 2)

increased prenatal care may reduce the incidence of infant morbidity and mortality (Dubay

et al., 1995). As part of the Omnibus Budget Reconciliation Act of 1989, states were

allowed to raise the eligibility standard to 185% of the poverty level for pregnant women

and their infants. Another expansion of Medicaid in 1991 allowed coverage to expand to

the mother a full 60 days after birth and to her infants up to their first birthday. Florida

was one of the first states to expand its program under the new laws.

Because of the structure of the program, approximately 30 million receive some

form of Medicaid assistance. Of these, more than half-a-million deliveries are paid for by

Medicaid each year, resulting in over $1 billion of costs annually (Miller et al., 1990).

Over time, expenditures for their share of Medicaid has become the largest expenditure for

states out of their budget. Medicaid expenditures have increased at an annual rate of 6.2%

since 1965 and currently account for 36.4% of all state spending (Schulman et al., 1997).

The high cost of the program coupled with the increasing cost of medical care in general

created a budgetary crisis for many states. As a result of the fiscal strains, many states

decided to move their Medicaid recipients (sometimes voluntarily and other times by

mandate) into newly created managed care plans with over 7 million joining the plans by


Studies have shown that Medicaid improves access to medical care, but has not

been shown to reduce the incidence of poor birth outcomes (Dubay et al., 1995).

Schlesinger and Kronebusch argue "...recent Medicaid expansions will not significantly

reduce low birthweight and infant mortality, even after states fully implement these

reforms. Medicaid eligibility increases prenatal visits, but Medicaid's beneficial effects are

limited by the program's inability to encourage women to seek care earlier and its failure

to induce better birth outcomes, once access has been established" (1990, p. 105). The

Alpha Center (1995), a division of the Robert Wood Johnson Foundation, conducted an

analysis of studies looking at the effect of the expansion of Medicaid. The research

indicates that the expansions of Medicaid have led to greater enrollment of eligible women

earlier in their pregnancies, although there has been no marked decrease in poor birth

outcomes overall.

In the early 1990s, in an effort to hold down medical costs associated with the

program, the federal government and many states actively encouraged Medicaid

beneficiaries to enroll in managed care plans (health maintenance organizations or HMOs).

This was also encouraged because better health outcomes were hoped for by using the

preventive philosophy of managed care plans. However, enrollment in managed care plans

has had no effect on reducing poor birth outcomes (Schulman et al., 1997) even though,

overall, similar care and resources are provided to women receiving perinatal care through

Medicaid as those with private insurance (Dobie et al., 1998).

Interestingly, the expansion of Medicaid in Florida appears to have had a greater

benefit for its Medicaid pregnant recipients. In particular, the low birthweight rate for

white women enrolled in Medicaid due to the expansions declined from 67.9 to 61.8 per

1,000 live births, while the rates for low income, white women with private insurance

remained the same (Long and Marquis, 1998).

Federal Healthy Start

During the Bush Administration's first couple of years in office, federal spending

for primary care services, including the Maternal and Child Health Block Grant, increased

by 51%, totaling more than $6 billion in federal fiscal budget year 1992. Of particular

concern for the Bush Administration was the problem of infant morbidity and mortality.

Bush set a goal for the U.S. to reduce overall infant mortality by 31%, to no more than 7

deaths per 1,000 live births, by the year 2000 (Cooper, 1992). Early in his administration,

Bush appointed a commission to examine the problem and recommend solutions. The

task force's report was never released to the public, however, among the 18

recommendations developed by the task force was the expansion of Medicaid eligibility

and the creation of national demonstration projects providing comprehensive maternal and

infant services to reduce infant mortality. The program emphasizes community

involvement reflecting the views of two members of the Bush administration, the assistant

secretary for Health, James Mason, and Robert Harmon, Health Resources and Services

Administration (HRSA) administrator (former public health directors for the states of

Utah and Missouri, respectively) (Howell, et al., 1998).

Entitled "Healthy Start," Bush's program was enacted in 1991 and 15

communities were chosen for participation in the program. The program immediately

created controversy within the states. The task force's recommendation would have cost

over $500 million, but Bush was hesitant to increase federal spending to that level and

recommended taking funds for the program from the Title V Maternity and Child Health

Block Grants and the Migrant Health Programs. States saw the potential for significant

funds to be taken away from their intervention programs to support a small number of

demonstration projects. Maternal and child health advocates in the states argued that

focusing that much effort on such a small number of areas would not significantly reduce

the incidence of infant mortality. Rejecting the funding plan, Congress appropriated about

half of the amount requested by the Administration for fiscal 1991 and 1992 for Healthy


Administered by the HRSA of the U.S. Department of Health and Human Services

(HHS), the federal Healthy Start program targets comprehensive perinatal services in

those communities with the highest infant mortality rates in the United States. An internal

HHS work group outlined the community approach to delivering perinatal services in a

document released to potential applicants in 1991 entitled, Guidancefor the Healthy Start

Program. Applications were due by July of 1991 for funding in September of 1991. The

guidance emphasized "substantive and informed" consumer participation. This was to be

accomplished by the creation of consortium of community members who would be

responsible for planning and the implementation of the program. This use of consortia and

"community empowerment" has been considered by some analysts as "the feature of

Healthy Start that most distinguishes it from previous maternal and child health programs"

(Howell et al., 1998).

Eligibility for the program required that the project area have an average infant

mortality rate for the five-year period between 1984-1988 of at least 150% of the national

average. The area must also have at least 50 but no more than 200 infants deaths per year.

Eligible applicants included: local or state health departments; publicly-supported

providers; tribal organizations; private, nonprofit organizations; or a consortium of the

above groups with the approval of the chief elected official of the city or county, or by

tribal leaders. In the first year, the approved projects were expected to use the grant to

develop a comprehensive plan for their community program. Forty proposals were

submitted, and fifteen communities were eventually provided funding (Howell et al.,


The fifteen communities originally selected for participation applied for a five-year

period between 1992 and 1996. These communities and their relative infant mortality

rates (infant deaths per 1,000 live births) in decreasing severity are as follow: Detroit, MI

(26.3); New Orleans, LA (23.3); District of Columbia (23.2); Philadelphia, PA (22.3);

Cleveland, OH (21.3); Pittsburgh, PA (20.2); Baltimore, MD (20.1); Chicago, IL (19.6);

New York, NY (19.4); northern Plains Indian communities in South Dakota, North

Dakota, Iowa, and Nebraska (18.7); Birmingham, AL (18.4); Oakland, CA (17.9);

Boston, MA (17.1); Lake County, IN (16.2); and Pee Dee region of eastern South

Carolina (16.1). Each site received between 12 and 23 million dollars over this time

period in order to meet an unprecedented goal of a reduction in infant mortality rates of

50% (Strobino et al., 1995; Howell et al., 1998). After the initial round of funding, other

programs were added to the federal program until reaching the current number of 60

programs with an additional 14 planned (MCHB, 1998). Table 1 lists the current

programs by state.

Table 1 -- Federal Healthy Start programs by state: 1998

Alabama Mobile TEEN Center

Birmingham Healthy Start
Arizona Phoenix Arizona Health Care Cost
Containment System / Baby Arizona
Arkansas Blytheville Mississippi County Arkansas EOC,
California Fresno Healthy Start

Los Angeles SHIELDS for Families, Inc./ARK:
Healthy Start Program

Oakland Oakland Healthy Start

San Bernardino San Bernardino County Healthy Start
II Project
Colorado Aurora Metropolitan Denver Provider
Connecticut New Haven Healthy Start New Haven
Delaware Newark Wilmington Healthy Start
District of Columbia Washington, D.C. District of Columbia Healthy Start

Table 1 -- continued.

Florida St. Petersburg St. Petersburg Healthy Start Federal

Tallahassee Florida Panhandle Healthy Start

Tampa Central Hillsborough County Healthy
Start Project
Georgia Augusta Augusta-Richmond County Healthy

Dublin Heart of Georgia Healthy Start

Savannah Chatham Savannah Healthy Start

Stone Mountain Center for Black Wellness, Inc.
Illinois Chicago Chicago Healthy Start

Southside Chicago Healthy Start Southside

Westside Chicago Westside Healthy Start

Park Forest Aunt Martha's Healthy Start Program

East St. Louis Southern Illinois Healthcare
Foundation's East St. Louis Initiative
Indiana Hammond Northwest Indiana Healthy Start

Indianapolis Indianapolis Healthy Start
Iowa Des Moines Healthy Start Des Moines
Kansas Wichita Northeast Wichita Healthy Start
Kentucky Louisville Healthy Start

Williamsburg Voices of Appalachia Healthy Start
Louisiana Monroe North Louisiana Area Health
Education Center

New Orleans New Orleans Healthy Start/Great
Maryland Baltimore Baltimore City Healthy Start
Massachusetts Boston Boston Healthy Start

Table 1 -- continued.

Michigan Detroit Detroit Healthy Start

Lansing Michigan Departments of Community
Health Bureau of Child and Family

Nazareth A Healthy Start in Kalamazoo

Sault Ste. Marie "Maajtaag Mnobmaabzid" A Start of
a Healthy Life

Saginaw Saginaw Cooperative Hospital, Inc.
Mississippi Jackson Delta Futures: A Healthy Start
Missouri Kansas City Kansas City Healthy Start

St. Louis Missouri Bootheel Healthy Start
Nebraska Omaha Omaha Healthy Start
New Jersey Pennsauken Camden Healthy Start

Trenton Essex County Healthy Start
New York Buffalo Buffalo Healthy Start Initiative -
Phase II

New York Downstate Healthy Families
Connections Project

New York Healthy Start / NYC

Rochester Healthy Start Rochester

Syracuse Syracuse Healthy Start
North Carolina Pembroke The University of North Carolina at

Raleigh Healthy Start Baby Love Plus
Ohio Cleveland Greater Cleveland Healthy Family /
Healthy Start
Oklahoma Oklahoma City Community Health Centers, Inc.

Tulsa Tulsa Healthy Start Initiative
Oregon Portland Healthy Birth Initiative

SWhite City Rogue Family Center Healthy Start

Table 1-- continued.

Pennsylvania Chester Healthy Start, Chester

Philadelphia Philadelphia Healthy Start

Pittsburgh Pittsburgh/Allegheny County Healthy

West Chester Healthy Start Initiative for Chester
County, PA
Puerto Rico San Juan Puerto Rico Healthy Start Project
South Carolina Columbia Low County Healthy Start

Richland County Columbia Richland Healthy Start

Florence Pee Dee Healthy Start
South Dakota Aberdeen Northern Plains Healthy Start
Texas Dallas Dallas Healthy Start

Fort Worth Catholic Charities, Diocese of Fort

Galveston Sisters of Charity Health Care System,
Galveston Ministry

Houston Neighborhood Centers, Inc.
Virginia Richmond Virginia Healthy Start Initiative
Wisconsin Lac de Flambeau Honoring Our Children with a
Healthy Start

Milwaukee Milwaukee Healthy Beginnings

As part of the guidance provided to local communities, the federal government has

recommended the use of one or more of nine models of intervention in their respective

Healthy Start programs. These models reflect the strategies that seemed to work best in

affecting poor birth outcomes in the initial demonstration project communities. These

models include: community-based consortia; care coordination/case management;

outreach and client recruitment; family resource centers; enhanced clinical services; risk