Citation
The relationships between selected socioeconomic variables and local tax effort to support public schools in Vermont

Material Information

Title:
The relationships between selected socioeconomic variables and local tax effort to support public schools in Vermont
Creator:
Fabian, Edward Joseph, 1937-
Publication Date:
Language:
English
Physical Description:
vii, 84 leaves : ; 28 cm.

Subjects

Subjects / Keywords:
Agricultural population ( jstor )
Educational administration ( jstor )
Employment discrimination ( jstor )
Mathematical variables ( jstor )
School districts ( jstor )
Schools ( jstor )
Socioeconomics ( jstor )
Taxes ( jstor )
University administration ( jstor )
Womens studies ( jstor )
Community and school -- Vermont ( lcsh )
Dissertations, Academic -- Educational Administration and Supervision -- UF ( lcsh )
Educational Administration and Supervision thesis Ph. D ( lcsh )
School districts -- Finance -- Vermont ( lcsh )
City of Gainesville ( local )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 80-83.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Edward Joseph Fabian.

Record Information

Source Institution:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
025567461 ( ALEPH )
03039828 ( OCLC )

Downloads

This item has the following downloads:


Full Text












THE RELATIONSHIPS BETWEEN
SELECTED SOCIOECONOMIC VARIABLES AND LOCAL
TAX EFFORT TO SUPPORT PUBLIC SCHOOLS
IN VERMONT


by

EDWARD JOSEPH FABIAN














A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA


1976
















ACKNOWLEDGMENTS


This writer wishes to extend his appreciation to Dr. Kern Alexander, Chairman of his Supervisory Committee, for his counsel, assistance and support throughout this study. Thanks are also expressed to other members of his committee for their assistance.

A debt of gratitude is owed Mr. Earl Blekking of the University of

Florida for his guidance throughout the array of statistical procedure and to Mrs. Connie Hebert of the Vermont State Department of Education for the technical assistance she provided.

The researcher expresses the deepest gratitude to his wife, Martha, and children, Charles, Kevin, Chris, Edward, Jr. and Julianne for their many sacrifices and encouragements throughout the writing of this study.










TABLE OF CONTENTS

Page

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

LIST OF TABLES ................................................. iv

ABSTRACT ....................................................... vi


CHAPTER

I. INTRODUCTION ......................................... 1

Statement of the Problem ........................ 3
Justification for the Study ..................... 3
Assumptions ..................................... 4
Definition of Terms ............................. 4
Procedures ...................................... 5

II. REVIEW OF LITERATURE AND RELATED RESEARCH ............ 8

National Historical Overview .................... 8
Related Studies ................................. 15
Summary ......................................... 32

III. STATISTICAL PROCEDURES AND PRESENTATION OF THE DATA .. 33 The Variables ................................... 34
Statistical Procedure ........................... 36
Presentation of the Data ........................ 39

IV. FINDINGS AND ANALYSIS ................................ 56

V. CONCLUSIONS AND RECOMMENDATIONS ...................... 72


APPENDIX
A. RANK ORDER OF VERMONT SCHOOL DISTRICTS BY TAX EFFORT
INDEX FROM HIGH TO LOW ............................... 74

B. CORRELATION COEFFICIENTS ............................. 77

BIBLIOGRAPHY ................................................... 80

BIOGRAPHICAL SKETCH ............................................ 84


iii









LIST OF TABLES


Page


TABLE I Zero Order Correlations Between the Dependent
Variable and the Independent Variables

TABLE II Relationship of Tax Effort to 24 Socioeconomic
Variables in 201 School Districts (Multiple
Regression Analysis)


TABLE III


Relationship of Tax Effort to 24 Socioeconomic Variables in 67 High Wealth Districts (Multiple Regression Analysis)


TABLE IV Variables Not In the Equation of 67 High Wealth
Districts

TABLE V Relationship of Tax Effort to 24 Socioeconomic
Variables in 67 Low Wealth Districts (Multiple
Regression Analysis)

TABLE VI Variables Not in Equation of 67 Low Wealth
Districts


TABLE VII TABLE VIII


Rotated Factor Matrix Derived from the Independent Variables

Factor Correlation Matrix for 201 School Districts


TABLE IX Relationship of Five Factors to Tax Effort in
201 School Districts (Multiple Regression
Analysis)

TABLE X Relationship of Four Factors to Tax Effort in
67 High Wealth Districts

TABLE XI Factor Correlation Matrix for 67 High Wealth
Districts


TABLE XII TABLE XIII TABLE XIV TABLE XV


Relationship of Five Factors to Tax Effort in 67 Low Wealth Districts

Factor Correlation Matrix for 67 Low Wealth Districts

Independent Variables Significantly Correlated with School District Tax Effort (In Descending Order of Correlation)

Regression Analysis of the Relationship of Local Tax Effort to Significant Independent Variables for All 201 School Districts









Page

TABLE XVI Regression Analysis of the Relationship of Local
Tax Effort to Significant Independent Variables
for School Districts with High Assessed Valuation
of Property per Pupil in Average Daily
Membership (N=67) 62

TABLE XVII Regression Analysis of the Relationship of Local
Tax Effort to Significant Independent Variables
for School Districts with Low Assessed Valuation
of Property per Pupil in Average Daily
Membership (N=67) 63

TABLE XVIII Independent Variables Correlating with Factors at
.40 or Higher 66








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



THE RELATIONSHIPS BETWEEN
SELECTED SOCIOECONOMIC VARIABLES AND LOCAL
TAX EFFORT TO SUPPORT PUBLIC SCHOOLS IN VERMONT

By

Edward Joseph Fabian

August, 1976


Chairman: Kern Alexander

Major Department: Educational Administration

This study sought to determine if certain socioeconomic characteristics" of school districts were related to tax effort in Vermont school districts during the 1972-1973 school year.

Tax effort was computed for Vermont's 201 school districts by

dividing local revenue per pupil in average daily membership by equalized grand list per pupil in average daily membership. The literature and research were reviewed to determine relevant variables. Twenty-five independent variables were selected.

The data were analyzed by means of stepwise multiple regression with tax effort as the dependent variable.

This statistical analysis was performed on all districts and then on districts whose equalized grand list valuation of property per pupil in average daily membership placed them in the top and bottom third of all districts in property wealth. Since it appeared that the stepwise regression equation masked many variables, a factor analysis was performed. Six factors were converted to factor scores and the scores were subjected to regression analysis.










This study concluded that no factors are significantly correlated to tax effort which would predict the type of districts making a substantial tax effort in the support of schools in Vermont. Even though the correlations were not high, the independent socioeconomic variables identified districts with high residential property values as the ones which make a greater effort in the support of schools in all districts and in the onethird of the school districts with the greatest wealth. Likewise, it appears, but is not conclusive, that the one-third low wealth districts, districts with high property farm value, make a greater effort in the support of education. The findings also suggest that districts which have a high concentration of economically disadvantaged children make less effort in the support of schools.

Even though the correlations were low and not significant, wealthy

residential areas appear to consistently tax themselves proportionately more than poor rural districts. A state-local finance system permitting optional local effort which operates to the benefit of residential districts tends to provide the greatest amount of educational resources per child in such districts.

Although the study does not fully explain the reasons for the wide

variations in tax effort for the support of schools in Vermont, wide variations do in fact exist. Because of this wide variation, the state should intervene and provide assistance which would reduce the disparities in expenditures for education.

Further studies are recommended to determine if geographic and political differences in school districts affect the level of the effort they make in the support of education in Vermont.


vii















CHAPTER I

INTRODUCTION

During the past twenty-five year period, between World War II and the beginning of the 1970's, America's school system experienced a tremendous increase in enrollment and an expansion in the number of years students were expected to stay in school. This growth, coupled with increases in teachers' salaries, greater offerings of school services, and inflation caused elementary and secondary costs to rise to a new high each year.

The total amount of school revenue from state
sources in current dollars . . . increased 77.3 percent between 1930 and 1940; the increase was
228.2 percent between 1940 and 1950; 164.2 percent
between 1950 and 1960; and 173.4 percent between
1960 and 1970.1

Expenditures in education have risen forty-three percent faster than increases in the economy as a whole. The result is that an increasing burden is being placed upon traditional school revenue sources.

School districts in all states, with the exception of Hawaii, utilize varying means of local taxation for the support of public schools. Local school districts in most states have legislative authority to raise revenue for the support of schools by school board action or voter approval. This practice has historically been associated with local control. It has been argued that local control increases local interest in schools. This view has raised serious questions regarding the equalization of educational opportunity.

R. Johns, K. Alexander, and K. F. Jordan, Financing Education - Fiscal and Legal Alternatives, Charles E. Merrill, Columbus, Ohio, 1972, p. 20.








2

Recent court decisions have been based on the fact that children in wealthy communities receive greater educational offerings than children growing up in poor communities.2 Two conclusions can be drawn from these decisions. The first is that state financing plans rely heavily on local property tax and cause substantial disparities among individual school districts in the amount of resources available per pupil for the district's educational program. The second conclusion which may be drawn is that as a result of the financing plans, taxpayers in less wealthy districts are required to pay a higher tax rate than taxpayers in many other school districts in order to obtain for their children the same or less educational opportunities afforded children in more wealthy school districts.

It is argued that economic and social conditions which exist within school districts bear a relationship to the fiscal efforts produced by the district in support of public education. It is reasonable to suspect that individuals with high incomes, and/or with a higher level of education, are more likely to support higher school tax rates than individuals with low educational levels and low incomes.

It is important to note that ninety-eight percent of all public

school districts in the United States rely on the local property tax for

4
financial support. Dependence on the local property tax effort as a

2Serrano v. Priest, 96 Cal. Rptr., 601, 487 P. 2d 1241 (1971).
Rodriguez v. San Antonio Independent School District, 337 F. Supp. 280 (1971) Reversed 41 "Law Week" 4407 March 21, 1973.
3R. Johns, K. Alexander, and D. Stollar, (Eds.) Status and Impact of Educational Finance Programs, Vol. 4, National Educational Finance Project, Gainesville, Florida, 1971.
4National Educational Finance Project, "Future Directions for School Financing," Gainesville, Florida, 1971.









3

method of obtaining future educational funds is seriously questioned in Vermont as well as throughout the nation. The above statements raise serious questions: Can we totally equalize education? How much effort are local districts willing to make in support of public schools? Will local control be eliminated? What type of foundation program should a state utilize? The answers to these questions may well require the ability to realistically measure local willingness to support public schools with taxes.

Statement of the Problem

The purpose of this study was to determine the relationship between selected socioeconomic variables and tax effort of school districts in Vermont.

Delimitations

1. This study is confined to school districts in the State
of Vermont.

Limitations

1. This study is ex post facto and possesses the weaknesses
inherent in this method.

2. Causal relationships cannot be determined from this type
of study.

3. Conclusions cannot be generalized beyond the districts
that are included in this study nor beyond the time for
which data are appropriate.

Justification for the Study

This study will add to our understanding of citizenry support of education. Its results should have practical value in light of the current status of financing school systems in Vermont and State Supreme Court decisions resulting from the Serrano v. Priest5 and Rodriguez v. San


5Serrano v. Priest, op. cit., 1971.








4

Antonio Independent School District actions.6

In view of the fact that many conventional arrangements and interests of school finance are being challenged in Vermont, it is timely that an investigation of this nature be conducted to determine whether state educational funds distributed on the basis of local optional effort are equitable. For the most part, the distribution of educational resources among school districts (on the basis of local optional effort) has favored wealthier districts.7

Campbell also supports this position.8 He states that locally raised revenue and state aid are still allocated in most school districts to favor the already socioeconomically advantaged schools.

This study should assist Vermont legislators in understanding the nature of local tax effort and it should have some practical value in future school finance decision making. No study of this nature has ever been undertaken in Vermont.

Assumptions

1. Equalized grand list valuations of property is a valid
measure of ability to finance education.

2. Equalized grand list valuation as computed is dependable
for all districts in the state.
3. Socioeconomic data secured from governmental sources are

reliable.

Definition of Terms

Ability.--A potential source of wealth, in this case equalized grand list valuation of property, which has the capacity to support education from the taxes based on it.

6Rodriguez v. San Antonio Independent School District, op. cit., (1971).

7R. Johns, K. Alexander, and D. Stollar, (Eds.), op. cit.

8Alan Campbell and Donna Shalala, "Resource Literature and Educational Revenue," Theory Into Practice, Ohio State University, Vol. XI,No. 2, April 1972, p. 77.









Average daily membership (A.D.M.).--The average enrollment of resident pupils of the district attending public schools for the first 30 days of a school year in which the school was actually in session; it is the quotient obtained by dividing by 30, the aggregate membership of resident pupils in the school district during the first 30 days in which the school was actually in session.9

Effort.--A ratio of school district revenue to some measure of
wealth. In this study it will be the ratio of local revenue per pupil in A.D.M. to equalized grand list per pupil in A.D.M.

Equalized grand list.--One percent of the fair market value of all
taxable property in a school district as established by the tax commissioner biennially plus the taxable polls.10

Local revenue.--Money collected at the local level primarily through the property tax for school purposes. This total represents all revenue receipts from local sources.

Percent of economically deprived children.--The percent of children eligible for Title I Elementary and Secondary Education Act benefits in proportion to total children of school age in the school district.

Poverty level.--A sliding scale of a family income in relation to marital status, age and family size, that is deemed inadequate or below poverty levels for United States Census purposes.11

State revenue.--The money obtained through state taxes and allocated to the local districts for school purposes through state aid.12

Socioeconomic variables.--Variables which are characteristic measures of the social and economic system. These do not include internal school characteristics such as teachers' salaries, number of teachers, pupilteacher ratios, etc.

Procedures

A stepwise multiple regression analysis was performed between the

measures of effort and selected socioeconomic variables. School districts


9Vermont Department of Education, "Vermont State Aid," Montpelier, Vermont, 1972, p. 2.

lOIbid., p. 2.

lUnited States Bureau of the Census, "Public Use Samples of Basic
Records from the 1970 Census," Washington, D. C., United States Department of Commerce, April 1972, p. 122.
12Vermont Department of Education, op. cit., p. 2.







6

were ranked according to the equalized grand list per pupil. Districts in the top and bottom third of this wealth spectrum were placed in two groups: a high wealth group and low wealth group. Variables found to be significantly related to effort in the two groups were compared to determine if high-effort, high-wealth districts have different socioeconomic variables related to their effort than high-effort, low-wealth districts.

A second step in the procedure was factor analysis. Factor analysis assisted in transformation of the independent variables into a set of factors which generally reveal important relationships which are difficult to discern among the variables in their original form. Collection of Data

Data were collected from the Vermont State Department of Education,

Vermont State Tax Department, the National Educational Finance Project and the United States Bureau of the Census. Analysis and Treatment of the Data

The statistical techniques used in this study were stepwise multiple regression and factor analysis. In stepwise regression, described by
13
Cooley and Lohnes and the BMD, Biomedical Computer Programs packet developed at the University of California,14 effort was the dependent variable and the socioeconomic factors were the independent variables. The stepwise procedure is a modification of multiple regression analysis that adds one variable at a time to the prediction equation. Variables are added or dropped in accordance with the significance of their contribution

13W. Cooley and P. Lohnes, Multivariate Procedures for the Behavioral Sciences, John Wiley and Sons, New York, 1962.
14W. T. Dixon, (Ed.) BMD, Biomedical Computer Programs, University of California Press, Berkeley, 1968.










7

to the prediction of the criterion variable. This results in development of intermediate regression equations as well as the complete equation.15

In factor analysis a correlation matrix is computed for the total data cluster. Then a line of best fit is computed for the cluster.16 Next, a series of perpendicular axes are computed to explain the maximum amount of variation remaining in the data cluster. These perpendicular (orthogonal) axes are then rotated so a minimum number of axes explain the variance of the data. Finally, the axes can be released from the requirement of orthogonality which permits them to conform more closely with individual data clusters. The resulting axes are the factors which will be extracted from the total data cluster.


15W. Cooley and P. Lohnes, op. cit., p. 35.

16Norman Nie, C. Hadley Hull and Dale H. Bent, Statistical Package for the Social Sciences, McGraw-Hill, New York, 1970, pp. 219-223.














CHAPTER II

REVIEW OF LITERATURE AND RELATED RESEARCH

Improvements can only come about if historical developments are

presented. History has been the basis on which values are improved, new ideas presented and mistakes corrected.

This chapter therefore, presents a national historical overview of

public education, use of real property tax for support of schools and the move toward educational equalization. Related studies reviewed are those that concern themselves with local financial ability, socioeconomic factors related to school expenditures and variables which influence tax effort.



National Historical Overview

The first public supported schools were established in Massachusetts and Rhode Island during the years of 1635-45.1 A European influence lies behind the formation of American Public Schools. Cubberley states that:

Schools with us, as with the older European countries
from which early settlers came, arose largely as children of the church . . . they brought with them their European
ideas as to religion and the training of children and hence,
a European backgr2und lies behind all the beginnings of
public education.

E. Cubberley, Public Education in the United States, Houghton Mifflin Co., Riverside Press, Cambridge, Massachusetts, 1934, pp. 11-12.

2Lbid., p. 12.











9

It was not long before individuals realized that voluntary support of public schools was not sufficient. Hence, two laws were enacted in the State of Massachusetts in 1642 and 1647. The law of 1642 was the first law passed by a state legislature making it compulsory that all children be taught to read. The law of 1647 established public supported schools. The significance of these two laws is presented by Cubberley:

Not only was a school system ordered established,
elementary for all towns and children, and secondary
for the youth in larger towns - but, for the first time among English speaking people, there was the assertion
of the right of the state to require communities to
establish and maintain schools, under a penalty of a fine
if they refused to do so.

It can be safely asserted that these two Massachusetts
laws of 1642 and 1647 represent not only new educational
ideas in the English speaking world, but that they,
together with the laws of 1634 and 1638, providing for
the equalized and compulsory taxation of all town charges,
also represent the very foundation stones upon which our
American public school systems have later been constructed.

Subsequently, after that three types of educational responsibilities emerged. The first was the support of a system of common schools, Latin schools and a college which served religious and civic ends. The second was the establishment of the parochial school concept, which stood for church control of all educational efforts and resented state interference. The third type conceived of public education as being intended chiefly for orphans and children of the poor and as a charity which the state was under
4
no obligation to support.


3Lbid., p. 17.
41bid., p. 25.









10

Cubberley5 states, that because of the Revolutionary War and these

three concepts of education, the Constitutional Convention did not evidently consider education to be important enough to be included in the Constitution and thus the responsibility for it was passed to the various states by the Tenth Amendment. The Tenth Amendment ratified in 1791 provided that, powerss not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the states respectively, or to the people."

Following the Revolutionary War, education of the young grew from

parental responsibility to church supported schools. Financial support in this early period was derived from lotteries, endorsements, licensing and commodities. The character of education during this early period and the lack of interest in it as described by Cubberley resulted from:

The simple agricultural life of the times, the homogenity
of the people, the absence of cities, the isolation and independence of the villages, the lack of full manhood suffrage in a number of states, the continuance of old
English laws, the want of any economic demand for education,
and the fact that no important political question calling
for settlement at the polls had as yet arisen, made the need
for schools and learning seem a relatively minor one.6

As living standards changed and the country moved from an agricultural to an industrial climate, citizens in newly formed states soon realized the need for education. During the Jackson administration education was expressed for the first time as political policy.7

During 1825 to 1830 in all the Northern states, the battle for direct,

local, county and state taxation for education was clearly evident, according


5Ibid., p. 86.
6Ibid., p. 110.

7P. R. Mort and W. C. Reusser, Public School Finance, McGraw-Hill Book Co., Inc., New York, 1941, p. 8.











to Cubberley.8 By 1825 it was recognized that the only safe reliance of a system of state schools lay in the general and direct taxation of all property for the support of education.

Massachusetts was the first state to take the lead and lay the

foundation for the state's common school fund in 1834 and it opened the first normal school in 1839. In New York an experimental free school law was in effect from 1795 to 1800. A permanent free school law dates from 1812. Also, in New York, a tax to raise the amount of money equal to that granted by the states was made compulsory.9

For the first time direct taxation for schools was likely to be felt by the taxpayer and the fight for and against the imposition of such taxation was on in earnest. In a general way, the progress of the conflict was as follows:10

1. Permission granted to communities so deserving to organize
a school taxing district, and to tax for school support the
property of those consenting and residing therein.

2. Taxation of all property in the taxing districts permitted.

3. State aid to such districts, at first from the income from
permanent endowment funds, and later from the proceeds of
a small state appropriation or a state or county tax.

4. Compulsory local taxation in the state or county grant.

With the beginnings of state aid the states were in a position to enforce definite requirements in many matters. As early as 1797 Vermont had required the towns to support their schools on penalty of forfeiting

8E. Cubberley, op. cit., p. 180.

9P. R. Mort and W. C. Reusser, op. cit., p. 9.

10E. Cubberley, op. cit., p. 180.







12

their share of state aid. The right to tax for support, and to compel local taxation, was the key to the whole state system of education.12 From this point on the process of evolving an adequate system of school support was different in each state. Equalization

Once taxation for free public schools was established, questions of equalization of educational opportunity were raised. Paul Mort indicated that revenue spent on education should go directly to where children are located and that these funds support a wide range of educational programs and services.13

Near the beginning of the twentieth century, Cubberley developed

the theory of state support. He saw the state's responsibility as being 14
concerned with equalization and reward for effort. Likewise, he saw the state as having the right to establish minimum educational standards which districts would be compelled to follow. School districts would 15
also be encouraged to extend their programs beyond the minimums.

American education was expanding rapidly and this concept of financial assistance to poorer districts had great influence. Updegraff, after a study of rural districts in New York, noted that granting of special financial assistance for individual functions ignored the economic weakness of vast sections of the states. He held that reward for effort on the basis of individual projects was too narrow in scope. He proposed


11Ibid., p. 188.

12Ibid., p. 189.

13P. R. Mort and W. C. Reusser, op. cit., p. 378.

14E. Cubberley, School Funds and Their Apportionment, Teachers College, Columbia University, New York, 1905, p. 16.
15Ibid., p. 17.










a sliding scale that provided increased amounts of state aid per teacher unit for each increase of one-half mill of school taxes levied ranging from three and one-half to nine mills, but he provided proportionately more state aid for a district with a low valuation per teacher unit.16

This marked the beginning of a state support program based upon the standard of more state aid to districts with low property valuation. In 1924 this concept of equalization was advanced further through a study completed by Strayer and Haig in New York State. Two pages of this study were devoted to a theoretical conceptualization of the equalization of educational opportunity which has had a major impact on the relationship of state and local agencies in exercising the responsibility of all people toward public education.

The concept, equalization of educational opportunity as described by Strayer and Haig, suggested that each state guarantee that minimal educational programs and facilities be available to every child within the states' borders and that the tax burden within the state be equal in relationship to each citizen's taxpaying ability.17

According to Mort18 the period following was highly productive in the further clarification of the equalization principle and in the development of various measuring devices and formulas for implementing these principles.

16R. Johns, K. Alexander and K. F. Jordan, Financing Education - Fiscal and Legal Alternatives, Charles Merrill Publishing Co., Columbus, Ohio, 1972, p. 7.
17G. Strayer and R. Haig, "The Financing of Education in the
State of New York," Report of the Educational Finance Inquiry Commission, Vol. I, Macmillan Co., New York, 1923, pp. 173-175.
18P. R. Mort and W. C. Reusser, op. cit., p. 384.











Mort defined a satisfactory equalization program as follows:

A satisfactory equalization program would demand that each
community have as many elementary and high school classroom or teacher units, or their equivalent, as is typical for communities having the same number of children to educate. It would
demand that each of these classrooms meet certain requirements
as to structure and physical environment. It would demand
that each of these classrooms be provided with a teacher,
course of study, equipment, supervision, and auxiliary activities
meeting certain minimum requirements. It would demand that
some communities furnish special facilities, such as transportation.19

He devised a foundation program using the concept of "weighting pupils" to compensate for the extra cost necessary for special programs.

Morrison,20 proposed a model of state support whereby all local

school districts are abolished and the state itself became both the unit for taxation for schools and for administration of public schools. He suggested that the most equitable form of tax for the state to use for the support of schools was the income tax. Summary

Programs which provide equalization of educational opportunity have been slow to develop in most states. The National Educational Finance Project classified the types of foundation programs in use in the United States into four basic categories. 21 Flat grants

Flat grants may be uniform or variable. They generally do not take into consideration the wealth of the school district and they are distributed on some basic unit of need, such as pupil or teacher unit. These grants may be of the general or special purpose type.

19R. Johns, K. Alexander, and K. F. Jordan, op. cit., p. 11.

20Ibid., p. 13

21R. Johns, K. Alexander and K. F. Jordan, op. cit., pp. 121-123.














Equalization grants

All equalization grants take into consideration variations in the taxpaying ability of the local school districts. However, not all equalization grants consider the variations of educational needs of the student population. Like flat grants, equalization grants may be general or special purpose.

Non-equalizing matching grants

Non-equalizing matching grants require local districts to match state funds on a dollar for dollar basis, or some proportion of a dollar without taking into consideration variations in the taxpaying ability of local school districts. These types of matching grants leave districts in the same relative status, and therefore, provide for little equalization. Complete state and federal support grants

All support for public education is from the state or federal

government. Local support is non-existent. Pupils' access to wealth is dependent upon the total wealth of the state. Revenue for the program can be obtained completely from the general funds of the state or from state taxes earmarked for this purpose.

Related Studies

A number of studies have shown that various socioeconomic factors affect decision making regarding school fiscal policy. The purpose of this section is to review relevant literature and research to identify socioeconomic characteristics related to local tax effort. This section will be divided into three subsections. The first subsection will identify variables related to voter attitudes toward education. Subsection two will consider socioeconomic variables correlated to school expenditures and the










final subsection will identify those variables which predict local district effort in support of education. A listing of all variables which have been identified as being correlated with local tax effort will be presented in the summary.

Public attitudes

In the Annual Gallup Poll22 of Public Attitudes Toward Education

conducted in 1972, fifty percent of the college graduates polled favored tax increases for schools while only twenty-seven percent of the people with only elementary education approved. White collar workers were much more favorable to increased school taxes than blue collar workers, while younger people were more favorable to increased school taxes than elderly people. People in higher income brackets were more favorable to tax increases than people on restricted incomes. Communities with populations between 25,000 to 50,000 were notably more approving of school tax increases than larger or smaller communities. Factors that were not significantly related to attitude on school taxes were region of the country, sex, people with no children in school, and private school patrons.

Likewise, the fifth and sixth Annual Gallup Polls of Public Attitude Toward Education conducted in 1973 and 1974, listed the lack of proper financial support as the third most important problem facing education for both years.
23
Meyers, in a study of urban communities, however, concluded that


22George Gallup, "Fourth Annual Gallup Poll of Public Attitudes Toward Education," Phi Delta Kappan, Vol. LIX, No. 1, September, 1972, pp. 33-46.
23Alfred V. Meyers, "The Financial Crisis in Urban Schools-Patterns of Support and Non-Support Among Organized Groups in Urban Communities," Doctoral Dissertation, Wayne State University, 1965, Dissertation Abstracts, Vol. 25, p. 5024.








17

parents of parochial and private school students tend to oppose increased financial support for public schools. He also found that organized groups within the community tend to influence the adequacy of financial support for public education. In still another study conducted by Carlson24 in three communities, he found that public and parochial school patrons had significantly different attitudes in federal aid, dual enrollment, shared facilities, the child benefit theory and on the extension of health and transportation services.

Patterson and Schoonhoven25 studied inconsistent voter behavior in

two Oregon public school districts. Despite demographic differences between the districts in level of schooling completed, occupation, income, period of residence, political affiliation and religious preference, reasons for voting positions were similar in both districts. In each district one-fourth of the voters attributed their failure to vote to forgetting or to being unaware of the election. A majority in both districts felt they had a legitimate reason for not voting previously and seemed determined to utilize their vote to affect the financial outcomes. Statistically significant differences were noted in both districts between positive voting and voters with children in school and more completed years of schooling. Significant findings in one district (not supported by the second district), related positive voting to persons under 45 years of age, level of occupational skill, spouses as income producers and family income over $5,000.

24DeVere Carlson, "Patron Attitudes Toward Selected Educational Issues in Communities with a Dual Educational System," Doctoral Dissertation, University of Pittsburgh, 1969, Dissertation Abstracts, Vol. 30, p. 3665A.

25Wade Patterson and John Schoonhoven, "A Comparative Study of
Inconsistent Voter Behavior in School Budget Elections," November, 1966, ERIC, ED-011135.








18

Voter behavior toward financial referendums was studied by Von Hatley.26 He found that family income appeared to have potential for predicting voter behavior in income based school referendums, but evidenced questionable predictive utility in property tax referendums. He also found that the voting behavior of middle income groups was more difficult to predict than the high and low income groups. Variables related to voting behavior were the number of children in the family, educational attainment, length of school district residency, the concepts of high versus low support for education and broad versus narrow conceptions of the value of education.

In a study by Witt27 of San Mateo voters, regarding a combined tax and bond proposal for additional Junior College funds, it was found that voters more than 50 years of age without children under 21 years of age were less in favor, particularly of the tax proposal, while white collar and professional people showed more favorable response than blue collar workers, housewives or retired people.

Petersen's28 study investigated the relationship between parent

attitudes toward school progress in a specific geographic area of a single school district and certain socioeconomic variables. Parents who had completed some high school education tended to be more approving of school programs than those who had some past high school training and those who had less than nine years of formal education. Non-white parents were found to be less approving

26R. F. Von Hatley, "Family Income Voting Behavior and Financial Referendums: Educational Finance and Politics in Albuquerque, 1968-69," Doctoral Dissertation, University of New Mexico, 1971, Dissertation Abstracts, Vol. 31, p. 5703A.
271rving Witt and Frank Pearce, "A Study of Voter Reaction to a Combination Bond-Tax Election on March 26, 1968," ERIC, ED 019945.

28Thor Petersen, "School Approval-Disapproval and Educational Enlightenment of Parents Based on Occupation, Educational Level, Age, Race, Geographic Area and Length of Residency," Doctoral Dissertation, Michigan State University, 1971, Dissertation Abstracts, Vol. 32., p. 2971A.








19

of school programs than white parents. Those parents who have lived in a particular school attendance area for more than five years were less approving than those who had resided there for less than one year. In this study, occupation and age were not significantly related to school approvals.

In a West Virginia study conducted by Photiadas and Zeller29 it was found that white collar, business, managerial and professional people favored raising taxes for the support of schools more so than did the unskilled or semi-skilled workers. A strong relationship existed between the level of education attained and the willingness to increase taxes. Income level was also positively correlated to a willingness to increase taxes in support of education.

Summary

The above studies indicate that the level of educational attainment, level of income, race, length in school district residency, occupations, non-public or parochial school attendance, age and voters with children in school can have some influence upon community attitudes toward public education.

Variables Related to School Expenditures

The primary goal of a study conducted by Ellis30 was to analyze the relationship between local expenditures on primary and secondary education and local income. The secondary aim of the study was to evaluate the significance of education relative to other factors as a source of difference

29J. Photiadas and F. Zeller, "Attitudes Toward State and Local Taxes in West Virginia: The Preliminary Results of a Survey," University Center for Appalachian Studies and Development, Morgantown, West Virginia, 1968.
30J. Ellis, Jr., "A Study of the Relationship Between Local
Expenditures on Education and Local Income," Doctoral Dissertation, University of Virginia, 1967, Dissertation Abstracts, Vol. 27, pp. 3177A-3178A.








20

in local income. Ellis concluded that average per student county educational expenditure was significantly related to levels of county per capita income. Attainment of a high school degree, urbanization, the non-white proportion of the population and occupational structure other than agriculture were related to labor income.

The purpose of MacDougall's31 study in Virginia was to analyze income flow to educational expenditures. The flow process was described by the measures of the following economic variables: income, property taxation, state and local charges, other state and local taxes, federal grants, federal revenue, federal grants to education, state and local expenditures. The economic variables and educational expenditures were analyzed by a multiple correlation and regression routine to establish a comprehensive set of relations describing the nature of the economic structure with educational expenditures. The variables, personal income and state and local charges, had a high correlation with educational expenditures.

The net result of a 1972 study conducted by Lazier32 in Utah also indicated that high personal income has a direct relationship to the amount of school tax a district will approve for school expenditures.

Likewise, Miner33 found in his study that the level of state per capita income was the most important determinant of total per capita expenditures. Median family income in local communities was inversely related to expenditures, but the proportion of families with incomes over


3M. A. MacDougall, "An Analysis of Income Flow to Educational Expenditures," Doctoral Dissertation, University of Virginia, 1964, Dissertation Abstracts, Vol. 24, p. 2738.
32Willard E. Lazier, "School Finance: A Determinant Model for
Eligibility," Doctoral Dissertation, University of Utah, Salt Lake City, 1972.

33Jerry Miner, Social and Economic Factors in Spending for Public Educations, Syracuse University Press, Syracuse, 1963.









21

$10,000 and the proportion of children in the population were positively related to expenditures. The proportion of children in private schools had a negative affect. Local expenditures were also strongly influenced by state aid formulas.

Harvey34 found that assessed property evaluations were most significant determinants of elementary educational expenditures. Residential, agricultural and commercial properties jointly were able to explain 89 percent of the variation in local expenditures while socioeconomic and voting characteristics explained 73 percent of the variation in current 35
expenditures. However, Fisher's study, which made extensive use of adjusted gross personal income per school district, demonstrated that there is justification for questioning property valuation as a valid measure of local fiscal ability.

In a study which attempted to find the relationship between community demand for education and the willingness of a population to finance a given amount of education and local financial support of public schools in Illinois, Metzcus36 found that median family income, population, percent of population non-white, proportion of the population in selected occupations correlated with local district current operating expense per pupil. He also found that the amount of property valuation per student had the highest positive correlation with local districts' current operating expenses per


34E. L. Harvey, "Property Tax Determinants of Educational Expenditures," Doctoral Dissertation, Stanford University, 1969, Dissertation Abstracts, Vol. 30, p. 1333A.
35Jack Fisher, "A Comparison Between Central Cities and Suburbs and Local Ability to Support Public Education," Doctoral Dissertation, University of Florida, Gainesville, 1972.
36Richard H. Metzcus, "Community Human Resources and Local Financial Support for Public Schools," Doctoral Dissertation, University of Illinois, 1969, Dissertation Abstracts, Vol. 30, p. 102A.








22

pupil and that districts with low property wealth tended to make a high tax effort in support of schools.

Districts with a high proportion of Industrial and Commercial property spend more money on schools for about the same tax rate as other districts.
37
This was the conclusion reached by Clune. He analyzed the effects of Industrial and Commercial property on wealth variations among school districts in Cook County, outside the City of Chicago.

Per pupil expenditures correlated positively with urbanization,

industrialization, income and the amount of education in a study conducted
38
by Dye. Effort, as was defined by expenditures relative to income, correlated negatively with urbanization, industrialization and income. He concluded that states with the greatest urbanization, industrialization and wealth made less tax effort, but maintained a high per pupil expenditure. Economic variables appear to be more influential than political - system characteristics in shaping educational fiscal outputs. Dye discovered, for example, that almost 70 percent of the total variation among the 50 states in per pupil expenditures was explained with median family income.
39
James and his colleagues, in an analysis of the determinants of

educational spending in 107 cities found that the socioeconomic variables; income, property value, adult education and race, were more highly correlated with expenditures than the characteristics of the school system. The percent


37William H. Clune, "Taxing and Spending for Public Schools; The Origin Descriptions and Effects of Non-School Taxes and Industrial and Commercial Property," Education and Government Division, Illinois Bureau of the Budget, Chicago, 1971.
38Thomas Dye, "Politics, Economics, and Educational Outcomes in the States," University of Georgia, Athens, 1967.
39Thomas James, J. Kelley and W. Garms, "Determinants of Educational Expenditures in Large Cities of the United States," School of Education, Stanford University, Palo Alto, California, 1966, pp. 95-134.







23

of labor force unemployed, median family income, percent of homeowners, median years of schooling, property valuation per pupil and the percent of pupils attending private schools were also all significantly correlated to school expenditures in this study.

Hendrix and Alkin40 found that 67 percent of the variation in school expenditures could be explained by individuals in certain age groups, and Farner41 found that the male labor force was positively related to expenditures for school districts as were housing standards.

In another study conducted for the Office of Education, James, Thomas and Dych42 found that property value was positively correlated to school expenditures in 15 sample states. Median family income was also positively related to all samples except in Nebraska. Likewise, percent of owner-occupied housing was negatively correlated to expenditures in all samples except Nebraska and Oregon. Median years of schooling was positively related to expenditures in two samples, negatively correlated to expenditures in three and not significant in ten. The percent of labor force unemployed was significant and negatively correlated to expenditures in three of the 15 samples. Percent of population non-white was significant in four of the 15 samples. Percent of farms population was negatively significant in four samples and the percent of elementary school children in

40V. Hendrix and M. Alkin, "Population Age Distribution and Public
Educational Expenditures." A paper, Educational Research Association, New York, February, 1967.
41Frank Farner, "Economic Sociological and Demographic Characteristics of Oregon School Districts and Their Relationship to District Financial Practices," University of Oregon, Bureau of Educational Research and Services, Eugene, April, 1966.
42H. James, A. Thomas and H. Dych, "Wealth Expenditures and Decision
Making for Education," USOE, Research Project No. 1241, Stanford University, Stanford, California, June, 1963.







24

private schools was significantly related to expenditures in only two samples. Property wealth, income and degree of home ownership had a significant relationship with school expenditures.

Alkin43 conducted a study which sought to examine the relationship

between religious composition of school districts and a measure of financial resources provided to the districts. He hypothesized that values implicit within the religious framework of belief, customs and practices and the apparent ethnic residues within religions provide differences in educational aspirations as reflected in expenditures for public education. Alkin studied 18 sample districts and the districts had to contain an enrollment of 300 A.D.A. or greater. He also categorized religious denominations into variables which could be examined statistically. His variables included percent of Catholic, percent Protestant Nurture, percent Protestant Borderline, percent Protestant Conversion, percent Jewish, percent Buddhist, percent Mormon, percent non-religion. Three variables were used for Protestant. This group was so large it could not be considered a single group because of loss of degrees of freedom in the multiple-regression equation used. Alkin found that public educational expenditures are related to the religious composition of communities and he concluded that studies utilizing socioeconomic characteristics as independent variables may do well to consider the effects of religious compositions. Summary

Variables identified as correlated to educational expenditures are: number of school years completed, percent of population non-white living in a district, per capita income, state and federal aid, equalized assessed

43Marvin Alkin, "Religious Correlates of School Expenditures," A
paper prepared for the American Research Association, Chicago, February 11, 1965, ERIC, ED 011143.








25

property valuation, percentage of property tax paid by residential, commercial and agricultural property owners, the percentage of pupils in a district attending non-public schools, the unemployment rate of the districts, the age of the voters and the number of children in the population as compared to the total population, white collar vs. blue collared occupations.

Variables related to tax effort

Effort is a measure of a district's willingness to pay for education.44 To measure effort, per capita personal income may be used as a measure, but a more practical, but by no means perfect measure, is to study the relationship between a school district's property value and the amount of income it raises locally. That is, measuring a district's taxable resources and the amount of money it raises against these resources.

Effort to support education depends on a community's willingness to

spend, based on its ability to spend. Ability to pay for education depends on the amount of wealth in the community and the amount of money coming into the community from non-local sources. Therefore, measures of local effort can be derived from: personal income level of the community, assessed valuation and true valuation, and the amount of revenue received from 45
state, federal and private sources.

Johns and Kimbrough46 supervised four studies which attempted to

determine factors associated with school district tax effort over a period

44"How to Evaluate Your District's Financial Effort," School Management, January, 1965, pp. 112-115.
45"Income v. Effort," School Management, January, 1969, pp. 62-69.

46R. Johns and R. Kimbrough, "The Relationship of Socioeconomic
Factors, Educational Leadership Patterns and Elements of Community Power Structures to Local School Fiscal Policy," USOE, Project No. 2842, May, 1968. ERIC ED 021336.











of time. School Districts with populations of 20,000 or more in four states were chosen. The authors concluded that through time, there is no combination of socioeconomic variables common to the four states that could explain much of the variations in effort of school districts. Independent variables were very unstable in their productive power and would not combine as the best predictors from one period of time to another. In general, however, the greater the income of the people of the districts included in the study, the greater the local effort in proportion to ability to support schools. The measures of per capita income explained more variance in local revenue receipts per pupil than all other socioeconomic variables combined.

As a part of the Johns and Kimbrough study, Adams47 analyzed socioeconomic factors associated with effort for education in Kentucky. Adams used 22 variables which he classified as socioeconomic. These variables included factors as public and private school enrollments, federal and state revenue, income of families, population, median school years completed, labor force employed and unemployed, owner-renter households and percent of population 65 years and older. Stepwise multiple regression was used to examine the relationship between the 22 variables and measures of local financial effort and elasticity of demand. He found that state revenue receipts per pupil in A.D.A., percent rural farm and median income of families were significant predictions of variability. Adams concluded that evidence cited in the study points out that socioeconomic variables leave a large part of local effort unexplained.


7perry R. Adams, "Socioeconomic Factors Associated with Patterns of School Fiscal Policy in Kentucky," Doctoral Dissertation, University of Florida, Gainesville, 1965.










Three other studies were conducted simultaneously with the Adams'

Kentucky study. King48 in his Georgia study, found that only one of the 22 variables was a significant predictor of local financial effort in 1950. The significant factor was percent of persons 65 years and over. In 1960 the significant predictor was state revenue receipts per pupil in A.D.A.

Hooper49 studied socioeconomic factors in Florida using the same 22 variables and procedures as were used in King's and Adams' studies. Using average effort for 1950 as the dependent variable, he found that percent of families with income of $10,000 or more, percent rural farm, percent in A.D.A. in public schools K-14 to total population, age 6-19, and per capita net effective buying power were reliable predictors of effort in Florida.

The fourth and final companion study was Quick's.50 He examined 28 districts in Illinois. Quick found that in 1950 there was significant relationship between effort and average daily attendance as a percent of total population. In 1960 effort correlated with A.D.A. as a percent of total population and state revenue receipts per pupil in A.D.A.

The conclusions reached in these four studies indicate that it is

impossible to generalize through time in any given state on the relationship with any particular set of economic variables to local school effort. Socioeconomic variables do not exclusively determine levels of local effort in school districts.

48Charles R. King, "Socioeconomic Factors Associated with Patterns of School Fiscal Policy in Georgia," Doctoral Dissertation, University of Florida, Gainesville, 1965.
49Harold Hooper, "Socioeconomic Factors Associated with Patterns of School Fiscal Policy in Florida," Doctoral Dissertation, University of Florida, Gainesville, 1965.
50W. J. Quick, "Socioeconomic Factors Associated with Patterns of School Fiscal Policy in Illinois," Doctoral Dissertation, University of Florida, Gainesville, 1965.









28
51
In a study conducted by Gentry, it was found that a high positive correlation existed between social climate, certain cultural conditions and local financial ability to support education. Gentry identified 13 variables clustered around six dimensions of a society. The six dimensions were identified as items of health, education, population composition, social stability, population distribution and economic distributions. Of the 13 identified variables, four were significantly correlated with local financial initiative. These were: median school years completed by persons 25 years of age and over, percent change in population, aggregate personal income per unit of educational load and percent of aggregate personal income in transfer payments. Gentry concluded that local initiative at higher ability levels results in a considerable increase in expenditures per unit of educational load.

The purpose of Martin's study52 was to evaluate certain social and economic characteristics of the school districts in Mississippi and to relate these characteristics to the local initiative of people in the support of schools. Local initiative was defined as tax effort beyond the minimum required by law for participation in the state foundation program. Local financial effort was defined as the total revenue effort of the districts and included financial effort required by law for participation in the foundation program.


51Gilbert Gentry, "The Relationship of Certain Cultural Factors to Initiative in the Local Support of Education in Florida," Doctoral Dissertation, University of Florida, Gainesville, 1959.

52Charles E. Martin, "The Relationship of Social and Economic
Characteristics to Local Initiative in the Financial Support of Public Schools in Mississippi," Doctoral Dissertation, University of Southern Mississippi, 1964, Dissertation Abstracts, Vol. 23, pp. 3730-3731.








29

The data presented indicated a total of fifteen variables significantly correlated with the criterion variable, local initiative. The identified variables were median school years completed 25 years and over, percentage of college graduates, percentage enrolled in college, percentage with income under $3,000, percentage with income $10,000 or more, per capita expenditures in A.D.A., percentage foreign born, percentage native population, population trend, percentage in manufacturing industry, percentage of white collar workers, percentage population 65 years and over, gross level of local initiative and local effort.

He concluded that three variables are the highest predictors of local initiative in the financial support of education in Mississippi. These were median income of population which accounted for twelve percent of the variance, population trends accounted for eleven percent of the variance and population with income $10,000 and over, accounted for ten percent.

A statistical analysis of the interrelationships of 36 factors selected as related to taxpaying ability for schools in Arkansas was 53
conducted by Garrison. An analysis of variance on each of the variables for the four classifications revealed a significant difference at the .05 level of significance in personal property assessments, assessments per A.D.A., median educational grade level, income per A.D.A. and local revenue per schools per $1,000 income.

The purpose of a study by Turck54 was to determine the relationship 53C. B. Garrison, "An analysis of the Interrelationships of Economic Index of Taxpaying Ability for Schools of Arkansas Counties," Doctoral Dissertation, University of Arkansas, 1965, Dissertation Abstracts, Vol. 26, p. 820.
54M. Turck, "A Study of the Relationships Among the Factors of Financial Need, Effort and Ability in 581 High School Districts in Michigan," Doctoral Dissertation, Michigan State University, 1966, Dissertation Abstracts. Vol. 21, p. 116.







30

between acceptable measures of need, effort and ability in Michigan Public High School districts. He found that a relationship exists between size of membership and taxable wealth. There is a tendency for a school district, as it increases in size of membership, to incur more effort for the support of educational programs and there appears to be no consistent relationship between the ability of a high school district and its effort.
55
An early study (1950) conducted by Myers, of certain phases of

local tax effort in relation to taxpaying ability in Florida's 67 county school districts, concluded that wealthy districts made greater effort in support of education even though receiving substantial increases in grants of state aid. He attributed this to the fact that adults in the wealthier counties had completed a greater number of years of schooling than adults in poor counties.

The findings of a Wisconsin study by Peterson56 revealed that personal income tax paid was the most adequate of the present measures of wealth available from public records and no single measure of wealth currently in use was adequate to describe fully the ability of a community to support public service including education. This study was conducted in 1963 and investigated problems of state support and local support for education.

Kay57 used variables related to categories of wealth and income, revenue sources, property tax assessments, social status and education, population

55H. 0. Myers, "A Study of Certain Phases of Local Tax Effort in
Relation to Taxpaying Ability in Florida," Masters Thesis, University of Florida, Gainesville, 1950.
56L. Peterson, R. Rossmiller, S. North and H. Wakefield, "Economic Impact of State Support Models on Education," University of Wisconsin, Madison, 1963.
57Harold B. Kay, "A Study of the Relationships Between Selected
Socioeconomic Variables and Local Tax Effort to Support Public Schools in Kentucky," Doctoral Dissertation, University of Florida, Gainesville, 1973.









31

and those related to district and student characteristics in his study. He concluded that the degree of urbanization, (that is, the degree to which there was a high proportion of business and residences--or no agricutlure--in a district) was the most important single factor in explaining the variance in local school tax effort in Kentucky. The second most important factor was the amount of wealth and income in the district.

Local variables related to differences in the amount of resources

raised locally appear to include median family income, proportion of families with incomes over $10,000 per year, proportion of students in secondary grades, the education level of the parents and the extent of home ownership as evidenced in a study conducted by Campbell.58 Summary

The studies and literature suggest the following variables related to tax effort: percent of families with incomes of $10,000 or more, per capita net effective buying income, percent rural farm, percent in A.D.A. in public schools, percent of students in non-public schools, number of school age children per family, population trends, state revenue per pupil, local revenue per pupil, federal revenue per pupil, income per A.D.A., cost of living, occupations, white collar vs. blue collar, percent population non-white, assessed valuation of district property, educational attainment, proportion of revenue derived from homeowners, proportion of revenue derived from commercial enterprise, proportion of students in secondary schools, length of school district residency, median family income, percentage of

58Alan Campbell, "The Socioeconomic, Political and Fiscal Environment of Educational Policy Making in Large Cities," Michael Kirst, (Ed), The Politics of Education, McCutchan Publishing Company, Berkeley, California, 1970.








32

income under $3000, percentage native population, percentage foreign born, percentage of college graduates, percentage of population 65 years and older and percentage of growth in school enrollments.

Summary

School tax effort is influenced by public attitudes toward education. Certain socioeconomic factors also appear to be related to school expenditures and may be influential in determining school district tax effort. It is apparent, from the literature reviewed, that less wealthy districts often made a greater tax effort than the wealthier districts to provide acceptable 59
educational programs. Income and wealth are repeatedly mentioned as being significantly related to effort and school expenditures.60

The following socioeconomic factors have been identified as being related to public attitude, school expenditures and local tax effort:

Variables related to wealth and income: equalized assessed
valuation of property, percent of economically deprived children, percent of labor force unemployed, and income
variables.

Variables related to revenue sources: percentage of revenue
derived from federal, state and local sources.

Variables related to property tax assessments: percent of
tax derived from residential, commercial and farm property.

Variables related to social status and education: educational attainment, religious composition of the school district, and
makeup of labor force.

Variables related to population: number of children in the
district, number of families with school age children, amount of time citizens have lived within the community, and ethnic
makeup of the school district.

59R. Johns, K. Alexander, and D. Stollar, (Eds.) Status and Impact of Educational Finance Programs, Vol. 4, National Educational Finance Project, Gainesville, Florida, 1971, p. 11.
60Adams, op. cit.; Dye, op. cit.; Gallup, op. cit.; Hendrix and Alkin, op" cit.; Johns and Kimbrough, op. cit.; MacDougall, op. cit.; Ketzcus, op. cit.; Petersen, op. cit.; Photiadas and Zeller, op. cit.














CHAPTER III

STATISTICAL PROCEDURES AND PRESENTATION OF THE DATA

The research and literature were examined to determine what socioeconomic variables are related to school district tax effort. This chapter will list the variables that were selected from the available data, identify the sources of those data, describe the statistical treatment of the data and will report the findings of the statistical analysis.

A stepwise multiple regression analysis was performed between the

measures of effort and selected socioeconomic variables; school districts were ranked according to the equalized grand list per pupil. School districts in the top and bottom third of this wealth spectrum were placed in two groups: a high wealth group and a low wealth group. Variables found to be significantly related to effort in the two groups were compared to determine if high wealth districts had different socioeconomic variables related to tax effort than the low wealth districts. A factor analysis was also performed. This was done to group the variables possessing correlation commonalities and therefore reduce the original set of variables into a smaller number which assisted in understanding the complex patterns of inter-relationships between the variables. Tax effort, as used in this study, is the ratio of local revenue per pupil in average daily membership to the equalized grand list of property per pupil in average daily membership.









34

The Variables

The following variables were chosen as a result of a search of the related literature and research and the availability of data:

XA school district tax effort (dependent variable)

X, average daily membership X2 total school expenditures

X3 total school current expenditures


total listed value of real estate


X5 percent of population


percent of percent of percent of percent of school percent of percent of percent of percent of percent of percent of and salary percent of percent of percent of retirement percent of percent of


population population population population population population population population population population population


18-65 years of age 65 and over 3-34 years of age in parochial schools in school ages 5-19 25 years old and over completing high


18-65 years with family with family with family with family with family


of age income income income income income


in the labor force $1000 to $5999 $6000 to $9999 $10,000 to $24,999 $25,000 and over derived by wage


with family income derived by non-farm


X6 X7 X8 X9 XIO XI1

X12 X13

X14 X15 X16

X17 X18 X19

X20


X21 percent of economically deprived children within each school
district


population with family income derived by farms family income derived from social sectiritv and family income det-ived from weI fare fami.ty income derived from olfher










X 22 percent of residential property value within each school district

X 23 percent of commercial property value within each school district

X24 percent of industrial property value within each school district

X25 percent of farm property value within each school district



Source of Data

Data for school district tax effort and for variables XI, X2, X3, and
i
X21 were obtained from the Vermont State Department of Education. Data

for variable X4 and variables X22 through X25 were obtained from the

Vermont State Tax Department.2

All remaining variables for municipalities and counties were available

through the 1970 census tract. These data were obtained through the fourth

count census tape.3

1For effort, average daily membership, equalized grand list and current expenditures were used which are listed in the Vermont State Department of Education publication, "State Aid to Education 1973" (Columns 2, 3, 7). Variable X21 was obtained from the Title I, E.S.E.A. office, Division of Federal Programs, Vermont State Department of Education from the list entitled, "Percentage of Economically Deprived Children, 1973".
2
Vermont State Tax Department, "1972 Real Estate and Personal Property Taxes Paid by Town and Category," Montpelier, Vt.
incorporated school districts, unified school districts and union school districts whose boundaries are not coterminous with municipal boundaries and school districts for which data was unavailable were not included. United States Bureau of the Census, 1970 "Census of Population: General, Social and Economic Characteristics of Vermont," Washington, D. C. United States Department of Commerce.
It was necessary to use the 1970 fourth count census tape because
eighteen of the two hundred and forty eight municipalities had population in excess of 2500 and were the only ones listed in the United States Bureau of the Census Report, "1970 General and Social Characteristics of Vermont," Washington, D. C., United States Department of Commerce.









36

With the exception of unified (K-12, 1-12), and union districts

(1-6, 1-8, 5-12, 7-12, 9-12) and some incorporated school districts, all municipal and school district boundaries are coterminous. Consequently, districts identified through the assistance of staff members from the State Department of Education consisted of six incorporated, four unified and twenty-eight union districts which were eliminated from the study due to the inability of matching their boundaries with census boundaries. Another twelve districts were eliminated because data was not available. Consequently, two hundred and one districts were included in the study. (Appendix A)

Statistical Procedure

After variables related to school district tax effort were identified in the review of the literature, they were subjected to the following statistical procedures. First, a correlation matrix of Pearson Correlation Coefficients was computed for all twenty-five variables. This demonstrated the statistical inter-relationships of all the data.

Next, a stepwise multiple regression analysis was used to determine

the relationships among the independent variables as they related to effort. In this analysis the independent variable having the highest simple correlation with the dependent variable first entered the equation. In the second and each subsequent step, the independent variable having the highest partial correlation with tax effort was entered; thus, at each step, the variable being brought into the computation was the one which made the greatest reduction in error in the analysis of variance based on the sum of squares deviation.









4
All stepwise multiple regression procedures utilized the SPSS program. This program brought new variables into the equation based on the normalized regression-coefficient value (3) the variable would have it brought into the equation. The significance of 3 is measured by the F statistic (the ratio of two variances to each other - the larger the ratio, the smaller is the probability that the differences between the variances was caused by chance.) If the F level is below .01, the variable is not admitted to the equation. A second factor utilized by the SPSS program prior to admitting a new variable to the equation is its tolerance. (The degree to which it covaries with the preceding variable.) If the tolerance is small - the covariance approaches unity (in this case .001) - then the variable is nearly a linear combination of variables already in the equation, so it will be excluded.

The stepwise regression was performed on all 201 school districts in Vermont, on the one-third with the highest assessed valuation of property per pupil, and upon the one-third with the lowest assessed valuation of property per pupil. It was suspected that some of the independent variables were masking others, if two independent variables were significantly related to each other and to the dependent variable, the one with the highest correlation with tax effort would be shown as accounting for most of its variance. The second independent variable might show only an insignificant increase in the explanation of variance, even though it might itself - if not masked by the stronger variable - be able to account for a substantial amount of the variance of the dependent variable. When the independent variables are substantially uncorrelated, they tend to measure different aspects of the criterion variable and so make a maximum

4Norman Nie, C. Hadlai Hull and Dale H. Bent, Statistical Package for the Social Sciences, McGraw-Hill, New York, 1970.








38

contribution to the explanation of variance. Using the statistical method of factor analysis, clusters of variables could be identified and analyzed.

The first step in factor analysis is to compute a correlation

matrix. Then a line of best fit is computed for the whole data cluster. This axis seeks to explain as much of the variance in the data as possible. Another axis at right angles to the first is computed to explain as much of the remaining variance in the data as possible. This process is continued until the remaining amount of unexplained variance in the data is considered insignificant. This process is called principal components analysis and it results in the extraction of the underlying factors from the data.

The next step is rotation of the axes. The axes are rotated while they remain at right angles (orthogonal) to each other so that a maximum amount of variance in the data can be explained with the fewest axes. The axes remain uncorrelated with each other. Rotation is necessary because although the first axis represented the "line of best fit" for the total data cluster, the succeeding axes - since they must remain at 900 to each other - do not necessarily fit well with subclusters of data. The rotation brings the whole framework of axes into better alignment with the total data cluster.

The last step used here was an oblique rotation of axes. This removed the requirement of orthogonality (axes being at 900 to each other) and allowed axes to be fitted more effectively to individual clusters of data. These axes can become correlated and the cosine of the angle between these factor axes represents the amount of correlation between the factors.5


51bid., pp. 219-223.








39

The program produced a factor analysis showing the loadings of each variable on each factor and a correlation matrix of the factors. These factors were identified by noting which variables loaded highly on them.

Factor scores were computed as if the six resultant factors were six individual variables. These factors were then subjected to stepwise regression with tax effort as the dependent variable. This was done for all school districts; for the one-third highest wealth districts and for the one-third lowest wealth districts. Correlation matrices of the factors were also computed for the high wealth and low wealth districts.

Presentation of the Data

Correlation between Variables

A correlation matrix of Pearson Correlation Coefficients was first
6
computed (Appendix B). Zero order correlations between the dependent variable and the independent variables and their level of significance as determined by their F ratio are presented in Table I.

Variable X22 (percent of residential property value in the school

district) correlated at .44 with the dependent variable effort at the .001 level of significance.

Variables X1 (Average daily membership), X2 (Total school expenditures), X3 *(Total school current expenditures), X13 (percent of population with family income between $10,000 - $24,999) were positively correlated with district tax effort at the .001 level of significance. Variable X (percent of economically deprived children), however, was negatively correlated at the .001 level of significance with district tax effort.


6Appendix B includes Pearson correlation coefficient for the total number of districts.










TABLE I

Zero Order Correlations Between the Dependent
Variable and the Independent Variables


Variable R Significance Level


X22 X2 xl
x3 X21 x8 X13 X25 X7 x9
x15 x4 X19 xll
x5 xlo
X24 X17
X12 X16
X23 X20 X6 X14 X18


.4400 .2992
.2957 .2881
-.2315 .1990 .1586 .1517 .1517
.1472 .1382 .1268 .1107
-.0788 .0759 .0747 .0712
-.0629 .0617
.0304 .0165
-.0126 .0063 .0032
-.0015


.001 .001 .001 .001 .001
.002 .012 .016
.016 .019 .025 .036 .059 .133
.142 . 146 .157 . 188
.192 .334 .408 .429 .465 .482 .491











TABLE II

Relationship of Tax Effort to 24 Socioeconomic
Variables in 201 School Districts

Multiple Regression Analysis


Variables Multiple R R2 R2 Change Simple R


X22 X25 X8 X2 X24 X20 X16 X19 X1

X14 X5
X15
X21 X6 XIO X9
X3 X18 X17
X12
X4 X7 xli
X13


0.43995 0.49859 0.53678 0.56013 0.57754 0.59367 0.60075
0.60740 0.61376 0.61779 0.62117 0.63215 0.63368 0.63487 0.63611 0.63702 0.63779 0.63850
0.63914 0.63959 0.63980
0.64001 0.64017 0.64029


0.19356 0.24859 0.28813 0.31374 0.33356
0.35244 0.36091 0.36893 0.37671 0.38166 0.38585 0.39961
0.40155 0.40306 0.40463 0.40580 0.40678 0.40768
0.40850 0.40908 0.40935 0.40961 0.40982 0.40998


0.19356 0.05503
0.03954 0.02561 0.01981 0.01888 0.00847 0.00803 0.00777
0.00495 0.00419 0.01376
0.00194 0.00151 0.00157 0.00116 0.00098 0.00090 0.00082 0.00057 0.00027 0.00026
0.00021 0.00015


0.43995 0.15167 0.19875 0.29917 0.07122
-0.01265 0.03036 0.11067 0.29567 0.00323 0.07586 0.13812
-0.23151 0.00629
0.07434 0.14717 0.28812
-0.00153
-0.06289 0.06165 0.12679 0.15168
-0.07888 0.15856








42

Variable X8 (percent population in school, ages 5-19) was positively correlated with tax effort at the .002 level of significance. Thus six of the twenty-five variables were correlated with district tax effort at the .05 level of significance or higher. Relationship of Tax Effort to Independent Variables

A stepwise multiple regression of tax effort with the twenty-five

independent variables for the 201 school districts was computed (Table II). This program provided the multiple correlation coefficient (Multiple R) which is the correlation between a dependent variable and the weighted sum of the independent variables.

The multiple regression program also provided the R2 statistic. This does not imply causation, but merely demonstrates the degree of covariation. Of the twenty-five variables, twenty-four had a combined Multiple R of .64029 explaining .40998 of the variance in tax effort. It will be noted that variable X23 (percent of comercial property value) dropped out of the regression equation because apparently the tolerance was too small. Variable X23 had a correlation of .18253 with X13 (percent of population with family income $10,000 to $24,999). However, X23 correlated with tax effort at .01648 and X13 at .15856. Thus, X13 entered the equation first and since both variables covarled so closely, although not significantly, X23 did not add sufficiently to the explanation of variance to be retained in the stepwise equation.

Variable X22 (percent of residential-property value) explained 19 percent of the variance in tax effort and was positively correlated to it. The next strongest variable was X25 (percent property value farms) which explained 5 percent of the variance in tax effort. Variable X8 (percent of population in school, ages 5-19) explained 4 percent, and X2









43

(total school expenditures) explained 3 percent of the variance in tax effort. Two other variables each explained 2 percent of the variance, X24 (percent property value industrial) and X20 (percent population family income other). These six variables cumulatively explained 35 percent of the variance in district tax effort. The remaining 18 variables together explained an additional 5 percent of the variance in tax effort.

The stepwise multiple regression equation was computed for the 67

school districts having the highest equalized assessed valuation of property per child - the highest third - using the twenty-five independent variables and district tax effort as the dependent variable (Table III). Sixteen of the twenty-four variables were dropped from the regression equation because the tolerance level of these variables was small (Table IV).

The eight variables in the regression equation had a combined

multiple R of .38573 and explained .14879 of the variance in tax effort. Variable X22 (percent residential property value) explained the highest variance, 4 percent in tax effort, while 1.6 (percent population family income non-farm) explained another 3 percent. The remaining variables explained 2 percent or less: X1 (average daily membership) negative 2 percent, and X8 (percent population in school, ages 5-19) 2 percent. X21

(percent economically deprived children), X5 (percent population 18-65 years of age), X9 (percent population 25 and over completing high school and X24 (percent of property value industrial) all explained 1 percent or less of the variance in local tax effort.

The stepwise multiple regression was computed for the 67 school

districts having the lowest equalized assessed valuation of property per child - the lowest third of all districts - using the twenty-four independent











TABLE III

Relationship of Tax Effort to 24 Socioeconomic
Variables in 67 High Wealth Districts

Multiple Regression Analysis


Variable Multiple R R2 R2 Change Simple R


x22 x16 x1
x21 x5 x8 x9
X24


0.21617 0.28049 0.30851
0.32067 0.33669 0.36098 0.37442 0.38573


0.04673 0.07867 0.09518 0.10283 0.11336 0.13030
0.14019 0.14879


0.04673
0.03194 0.01651 0.00765 0.01053 0.01694 0.00989 0.00860


0.21617
0.16361
-0.05390
-0.13198
-0.06090 0.00475 0.09880
-0.08453










TABLE IV

Variables Not In the Equation of 67 High Wealth Districts


Variable Beta In* Partial* Tolerance* F


X2 0.01495 0.00404 0.06222 0.001 X3 0.11545 0.02197 0.03083 0.023 X4 -0.00031 -0.00013 0.14657 0.000 X6 0.08172 0.06821 0.59299 0.271 X7 0.00749 0.00711 0.76651 0.003 X10 0.02718 0.01455 0.24391 0.012 XII -0.02511 -0.02347 0.74384 0.032 XI2 -0.03151 -0.02566 0.56471 0.038 X13 -0.06915 -0.04835 0.41617 0.136 X15 -0.08411 -0.05585 0.37530 0.181 X17 0.05709 0.05384 0.75698 0.169 X18 0.08629 0.08155 0.76018 0.388 X19 0.08804 0.08708 0.83271 0.443 X20 0.09452 0.08817 0.73908 0.454 X23 0.07741 0.05210 0.38551 0.158 X25 0.01598 0.01476 0.72593 0.013


*Beta In.--The normalized regression coefficient that the independent variable would have if it were brought into the equation on the next step. The significance of Beta is measured by the F statistic. If F is small, there is little reason to add the independent variable to the prediction equation.

*Partial.--The partial correlation which represents the correlation computed for a partial group selected on the basis of the one or more other variables that are held constant in the selection process.

*Tolerance.--The pivotal element which brings the variable into the
equation. A small tolerance indicates that the7variable is mearly a lineal combinaion of variable already in the equation.

7Norman Nie, C. Hadlai Hull and Dale H. Bent,,op. cit., pp. 179-181.








46

variables and district tax effort as the dependent variable. Like the top third, eight variables remained in the regression equation (Table V). Sixteen variables dropped out of the equation (Table VI). The eight variables had a combined multiple R of .64822 which explained .42019 of the variance in school district tax effort.

Variable X25 (percent property value farms) explained 13 percent of the variance in tax effort. Variables X1 (average daily membership) and X4 (total listed value of real estate) explained 10 percent, while X15 (percent population family income wage and salary) explained 4 percent of the variance in local tax effort. These variables, plus the combination of the remaining four, X17 (percent family income farm), X20 (percent family income other), X.6 (percent population family income nonfarm), and X23 (percent property value commercial) explained 42 percent of the variance in local tax effort.

Relationship of Independent Variables to Factors

As was noted previously, there appeared to be a high degree of

inter-relationship between independent variables. It further appeared that these related variables might form several clusters with each cluster representing an underlying reality or factor. The twenty-five independent variables were subjected to factor analysis.

The factor analysis extracted six factors from the independent

variables. The columns of figures under each factor show the correlation between a specific variable and that factor (Table VII). Factors are identified by noting the configuration of correlations between it and the independent variables.

In factor 1 variables X1 (average daily membership) and X3 (total school current expenditures) had positive correlations of .99 with the











TABLE V

Relationship of Tax Effort to 24 Socioeconomic
Variables in 67 Low Wealth Districts

Multiple Regression Analysis


Variable Multiple R R2 R2 Change Simple R


X25 X 1

X4 X15 X17
X20 X16
X23


0.36678 0.48572 0.57701 0.60963
0.61991 0.63275
0.64080 0.64822


0.13453 0.23593
0.33294 0.37165 0.38428
0.40037 0.41062 0.42019


0.13453 0.10140 0.09701 0.03871
0.01264 0.01609 0.01025 0.00957


0.36678 0.32449 0.14594 0.25277
0.31920 0.10095 0.17363 0.21962











TABLE VI

Variables Not in Equation of 67 Low Wealth Districts


Variable Beta In Partial Tolerance F


x2 x3 x5 x6 x7 x8 x9 xl0 x11
x12 x13
x14 x18 x19
x21 x24


0.3174 0.40063 0.01768
-0.07841 0.04708
-0.13252 0.00727
-0.17970 0.03394
-0.07817 0.01250
-0.04703
-0.06390
-0.02798
-0.02277
0.02094


0.09778 0.10576 0.01329
-0.08731 0.06072
-0.13131 0.00631
-0.11065 0.03180
-0.08395 0.01110
-0.05895
-0.06971
-0.03280
-0.02634 0.02424


0.05596 0.04041 0.32764 0.71890 0.96434 0.56924 0.43665 0.21983 0.50900 0.66868
0.45710 0.91077 0.69011 0.79684 0.77550
0.77697


0.541 0.633 0.010 0.430 0.207 0.983 0.002 0.694 0.057 0.397
0.007
0.195 0.273 0.060 0.039 0.033











TABLE VII

Rotated Factor Matrix Derived from the
Independent Variables


FACTORS

1 2 3 4 5 6


0.99430 0.98074 0.99471 0.82551 0.06363
0.01096 0.33747
0.00634 0.13224
0.04022
-0.26232
-0.15233
0.24989 0.08832
-0.01422
-0.10777
-0.22687
-0.12138 0.01190
0.07470
-0.23488 0.36991 0.48228 0.02761
-0.16679


0.07302
0.06949 0.06646
0.02994 0.84005 0.03109 0.06982 0.49856 0.51293 0.82737 0.22695
0.41410 0.51849 0.15105
0.87504 0.18201 0.13449 0.12970 0.18412 0.40622
0.03370 0.12460
-0.09165 0.19408
-0.17930


0.01928
0.02984 0.02207
-0.03745 0.00089 0.87783
-0.03574
-0.02125 0.34855
0.05943 0.45723 0.06816
0.14073
-0.11030 0.22917 0.16317
0.02824 0.87327
0.23427 0.41567
0.13444 0.07571
-0.10111 0.03727
-0.05953


-0.04975
-0.06136
-0.04694
-0.11102 0.06526
-0.05435
-0.05640 0.22924
-0.00185
-0.09335
0.40532
-0.08074
-0.30215 0.15161
-0.17877 0.01917
0.86644 0.00349
0.01575
-0.00406
0.19345
-0.29892
-0.26292
-0.03761
0.61694


0.00008
0.02066 0.02479 0.10636 0.19235 0.05111
-0.13239 0.02727
0.17524 0.34169
-0.51824 0.05254 0.41821 0.32553 0.09901
0.45974 0.08473 0.01723
-0.01352
0.46104
-0.08981
0.11628 0.07998
-0.07518 0.10115


0.00352
-0.00612
-0.00483
-0.07063 0.09909
0.06629
-0.02049 0.06155
-0.03579
0.05060
-0.10982 0.85818
-0.37577
-0.00580 0.12725 0.01769
-0.09057
-0.00327
0.19930
-0.04888
-0.05628
-0.04709
-0.08292 0.07493
0.03585


X1
X2 X3
X4 X5 X6 X7 X8
X9 X10 Xl
X12 X13
X14 X15
X16 X17 X 18

X19 x20 x21 X22 x23
24 X25










50

factor. Variable X2 (total school expenditures) had a positive correlation of .98 and variable X4 (total listed value of real estate) correlated at .82. Variable X23 (percent property value commercial) had a positive correlation of .48 while variables X22 (percent residential property value) and X7 (percent population 3-34 in parochial schools, K-12) had positive correlations of .36 and .33, respectively.

In factor 2 variable Xi5 (percent population family income wage and

salary) had a positive correlation of .87. Variable X5 (percent of population 18-65 years of age) and variable X10 (percent of population 18-64 in labor force) correlated with factor 2 positively at .84 and .82. A positive correlation of .51 with the factor was indicative of variables X9 (percent of population completing high school) and X13 (percent of population family income $10,000 to $24,999). Three variables X8 (percent of population in school age 5-19), X12 (percent of population family income $6000-$9999) and X20 (percent of family income other) had a positive correlation of .40 or above.

In factor 3 variable X6 (percent population 65 and over) and variable X18 (percent population family income social security and retirement) had a positive correlation of .87 with the factor. Variable X1 (percent population family income $1000-$5999) and variable X20 (percent population family income other) had positive correlations with factor 3 at .45 and .41, respectively.

Variable XI7 (percent family income farm) had a positive correlation of .86 with factor 4, while variable X25 (percent property value farm) had a correlation of .61 with the factor. One other variable, variable X11 (percent population family income $1000-$5999) correlated with the factor at .40.







51

In factor 5, two variables had a positive correlation of .46 and .45. Those were variables X20 (percent population family income other) and X16 (percent population family income non-farm). Variable XI3 correlated at .41. Variable X11 (percent population family income $1000-$5999) had a negative correlation of .51 with the factor.

In factor 6 only one variable had a significant positive correlation with the factor and that was variable X12 (percent of population family income $6000-$9999).

A factor correlation matrix was computed and the factors did not

correlate significantly (Table VIII). The highest correlation was between factor 2 and factor 5 at .07 at a significant level of .13. Relationship of Factors to Tax Effort

Factor scores were next computed for all 201 school districts and a stepwise multiple regression between each factor and school district tax effort was performed. The factors extracted from the 25 independent variables were treated as six independent variables and placed in the regression equation with tax effort as the dependent variable (Table IX).

Five factors were included in the multiple regression. The tolerance level of factor 3 was insufficient for computation, therefore that factor was eliminated from the equation. The five factors within the equation had a combined multiple R of .2923. Likewise the R2 explained .0854 of the variance in tax effort. Factor 1 explained 6 percent of the variance in tax effort while factor 2 accounted for 2 percent.

The factor scores for the 67 high property wealth school districts were then subjected to a stepwise multiple regression with tax effort as the dependent variable (Table X). Factors 2 and 4 were eliminated from the equation because the tolerance level was insufficient. The remaining











TABLE VIII

Factor Correlation Matrix for 201 School Districts


FACTORS


1
1.000


2
0.0028 1.000


3
0.0039 0.0366 1.000


4
-0.0044 0.0026 0.0016
1.000


5
-0.0021 0.0768
-0.0265
-0.0457 1.000


TABLE IX


Relationship of Five Factors to Tax Effort
in 201 School Districts


Multiple Regression Analysis


Multiple R
0.26440 0.28610
0.29150 0.29224 0.29233


Factors
1
2


6
0.0052 0.0290 0.0035
-0.0196
-0.0502 1.000


Factors
1
2
6
5
4


R 2

0.06991 0.08185 0.08497
0.08540 0.08545


2 Change

0.06991 0.01195 0.00312
0.00043 0.00005


Simple R 0.26440 0.11003 0.06034 0.02565
0.00413









5.3

4 factors, factors 5, 3, 1 and 6 had a combined multiple R of .0771. The R2 explained only .0071 of the variance in tax effort. Factor 5 explained

4 percent of the variance in tax effort.

A factor correlation matrix was computed (Table XI) for the factor scores in the 67 high wealth districts, Factors 1 and 6 had a negative correlation of -.2813 significant at the .01 level. Factor 2, likewise, had a negative correlation of -.2738 with factor 3 at a significant level of .01. Factor I had a negative correlation of -.1828 significant at the .06 level with factor 4 and a positive correlation of .1626 with factor 5 at a significant level of .09.

The factor scores for the 67 low property wealth districts were

subjected to a multiple regression stepwise procedure using tax effort as the dependent variable (Table XII).

Factor 3 again was left out of the equation because of a low tolerance level so was not included in the computation. The five remaining variables had a combined R of .4981. The R2 explained .2481 of the variance in tax effort. In this analysis factor 4 accounted for approximately 10 percent of the variance in tax effort, while factor 2 accounted for 6 percent and factor 1, 5 percent of the variance.











TABLE X

Relationship of Four Factors to Tax Effort
in 67 High Wealth Districts


Factors Multiple R R2 R2 Change Simple R 5 0.21790 0.04748 0.04748 0.21790 3 0.24423 0.05965 0.01217 0.07495 1 0.26461 0.07002 0.01037 -0.06118 6 0.27777 0.07716 0.00714 -0.06098









TABLE XI

Factor Correlation Matrix for 67 High Wealth Districts



FACTORS

Factors 1 2 3 4 5 6


1.0000 0.0917
1.0000


0.0088
-0.2738
1.0000


-0. 1828
-0.0987
-0.0836 1.0000


0.1626 0.0369
-0.1560 0.0135 1.0000


-0.2813
0.1270 0.1342 0.0969
-0.1002 1.0000










TABLE XII

Relationship of Five Factors to Tax Effort
in 67 Low Wealth Districts


Factors Multiple R R2 R2 Change Simple R


0.32414 0.40287 0.46864 0.49280 0.49818


0.10507 0.16231 0.21962
0.24285 0.24819


0.10507 0.05724 0.05732 0.02323
0.00534


0.32414 0.29084 0.12984
-0.08486
-0.08752


A factor correlation matrix was computed (Table XIII) for the factor scores in the 67 low wealth districts.

Factor 1 had a negative correlation of -.4053 that was significant at the .001 level with factor 2. Factor 2 had a positive correlation with factor 3 of .2445 at a significant level of .02.



TABLE XIII

Factor Correlation Matrix for 67 Low Wealth Districts



FACTORS

Factors 1 2 3 4 5 6


1.000 -0.4053
1.0000


-0.1447 0.2445 1.0000


0.0386 0.1699 0.1818
1.0000


0.1708 0.1625
-0.0440
-0.1875 1.0000


-0.1042 0.0683
-0.0812 0.0094 0.0738 1.0000















CHAPTER IV

FINDINGS AND ANALYSIS

This study sought to determine if certain socioeconomic characteristics of school districts were related to tax effort in Vermont school districts during the 1972-1973 school year. Tax effort for each school district was computed by dividing local revenue per pupil in average daily membership by the equalized grand list of property valuation per pupil in average daily membership. The literature and research were reviewed to determine what variables might be correlated to tax effort. Twenty-five independent variables were selected for the study.

Data were collected from the Vermont State Department of Education,

the 1970 United States Bureau of the Census, the Vermont State Tax Department and the National Educational Finance Project. The data were analyzed by means of stepwise multiple regression. This statistical analysis was first performed on all districts and then on districts whose equalized grand

- list valuation of property per pupil in average daily membership placed them in the top and bottom third of all districts in property wealth.

Data were then subjected to factor analysis. The resultant six factors were identified and factor scores computed.

The factors were then subjected to a stepwise multiple regression

with tax effort as the dependent variable. This statistical analysis was performed on all 201 districts and on the 67 school districts whose equalized grand list of property valuation per pupil in average daily 56








57

membership placed them in the top one-third of all districts in property wealth and on the 67 school districts whose equalized grand list valuation of property placed them in the lowest one-third of all districts in property wealth.

This section will summarize and analyze the findings. Initially,

the simple correlations of independent variables with tax effort will be discussed. Next, the results of the regression equation using the independent variables will be presented. The results of the factor analysis will likewise be presented and finally the regression equations using factor scores will be analyzed.

The independent variables with significant simple correlations with tax effort are shown in Table XIV.

The variable with the highest simple correlation with effort was one relating to the main source of property taxes - residential property. Districts deriving much of their local taxes from residences (X22) tended to make high effort. This could be expected since the majority of school districts are rural and most property is residential or agricultural in nature in Vermont.

School districts with high expenditures (variables X2 and X3) tended

to make high effort as did districts with high average daily membership (X). In school districts characterized by a high percentage of economically deprived children (X21) there tended to be low effort (-.23). Low effort in these districts might be related to poverty levels, the availability of alternative funding sources and the low educational aspiration levels of individuals at poverty levels.

Districts which have a high percentage of population in school,

ages 5-19 (X.) make a high effort as do districts characterized by a high










TABLE XIV

Independent Variables Significantly Correlated
with School District Tax Effort

(In Descending Order of Correlation)


Variables


percent of residential property value total school expenditures average daily membership total school current expenditures percentage of economically deprived children percentage of population in school age 5-19 percentage of family income $10,000-$24,999 percent property value farms percent population 3-34 in parochial school K-12 percent population 25-over completing high school percent population family income wage and salary total listed value real estate

















significant at .01 level significant at .05 level


.4400* .2992* .2957* .2881*
-.2315* .1990** .1586** .1517** .1517**
.1472** .1382** .1268**


X22 X2 X I x3
X21 x8 X13 X25 X7 X9 X15
x4









59

percentage of families with incomes between $10,000 to $24,999. It is interesting to note that a simple R of .1517, significant at the .05 level, was characterized for variable X25 (percent property value farms) and variable X7 (percent population 3-34 in parochial schools K-12). This is probably due to the fact that most of the remaining parochial schools in Vermont are located in the wealthier urban areas, where income levels are somewhat higher and income (X13) is positively correlated with tax effort.

Districts which have a high percentage of population 25 and over completing high school (X9) and districts which have a high percent of population deriving family income from wage and salary (X 5) make substantially higher effort in supporting their schools. Finally, districts with a higher total listed value of real estate also made high effort. Residential property tax source, income and education all had higher correlation with effort than total value of real estate. Individual Variables in Regression Equation with Effort - 201 districts

In the stepwise regression equation showing the cumulative relationship between the independent variables and tax effort for all school districts, six variables accounted for 35 percent of the variation in tax effort as shown in Table XV.

Percent of residential property value accounted for 19 percent of the variance in tax effort. Five percent of the variance was attributable to farm property values. For all school districts in Vermont, 25 percent of the variance in school district tax effort could be explained by these two variables alone. Therefore, in districts where the local school revenue was derived from taxes mostly on residences and farms, the school district tax effort tended to be high.









60

TABLE XV Regression Analysis of the Relationship of Local Tax Effort to Significant Independent Variables for All 201 School Districts


Variables R R Change Simple R X22 percent residential property value .19356 .19356 .43995 X25 percent property value farms .24859 .05503 .15167 X8 percent population in school .28813 .03954 .19875
age 5-19
X2 total school expenditures .31374 .02561 .29917 X24 percent property value industrial .33356 .01981 .07122 X20 percent population family income .35244 .01888 -.01265
other


Variables X8, X2, X24, and X20 accounted for an additional 10 percent in the variance in tax effort. Districts which have a high percentage of children ages 5-19 in school explained 4 percent of the variance in tax effort while 3 percent of the variance was explained by total school expenditures. Income was correlated negatively with effort which could indicate that it (income) was not a determiner of effort in the 201 school districts.

It should be noted that variable X23 (percent property value

commercial) dropped out of the regression equation. The reason seems to be that it was insignificantly correlated with the other variable and consequently did not provide enough extra explanation of the variance in tax effort to be retained in the equation. Individual Variables in Regression Equation with Effort in 67 High Wealth Districts

In the stepwise regression equation for the high property wealth








61

districts, the cumulative relationship between tax effort and the independent variables, eight variables accounted for 14 percent of the variance in effort as shown in Table XVI.

The percent of residential property value accounted for 4 percent of the variance in tax effort and the percent of population with family income fron non-farm sources accounted for 3 percent.

In relatively wealthy districts the value of residential property positively related to high effort. Likewise, the non-farm income factor would suggest that the individuals in wealthy districts may be white collar workers deriving their income from other than farm sources. Likewise, an additional conclusion to be drawn may be that the high effort districts in the wealthier areas are not farming communities.

Average daily membership, the percent of economically deprived children and the percent of population 18-65 years of age, each explained less than

2 percent of the variance in tax effort for the support of schools. Each of these variables correlated negatively with tax effort which would indicate that the number of children in the district is not an influential factor in the tax effort they made. Likewise, these districts more than likely have a small number of economically disadvantaged children.

With the exception of percent of residential property value, none of the variables in the equation had a simple R above .2161. Individual Variables in Regression Equation With Effort in 67 Low Wealth Districts

In the stepwise regression of the low property weniith districts showing

the cumulative relationship between effort niid the independent variables, eight variables accounted for 42 percent of the variance in tax effort as shown in Table XVII.












TABLE XVI

Regression Analysis of the Relationship of Local Tax Effort to Significant Independent Variables for School Districts with High
Assessed Valuation of Property per Pupil in Average Daily Membership (N=67)



Variables R2 R2 Change Simple R


X22 percent residential property value .04673 .04673 .21617 X16 percent population family income non-farm .07867 .03194 .16361 X1 average daily membership .09518 .01651 -.05390 X21 percent economically deprived children .10283 .00765 -.13198 X5 percent population 18-65 years of age .11336 .01053 -.06090 X8 percent of population in school ages .13030 .01694 .00475
5-19
X9 percent population 25 and over completing .14019 .00989 .09880
high school
X24 percent property value industrial .14879 .00860 -.08453











TABLE XVII

Regression Analysis of the Relationship of Local Tax Effort to
Significant Independent Variables for School Districts with Low
Assessed Valuation of Property per Pupil in Average Daily Membership (N=67)



Variables R2 R2 Change Simple R

X25 percent property value farms .13453 .13453 .36678 X1 average daily membership .23593 .10140 .32449 X4 total listed value real estate .33294 .09701 .14594 X15 percent of population family income .37165 .03871 .25277
wage and salary
X17 percent of population family income farm .38428 .01264 .31920 X20 percent of population family income other .40037 .01609 .10095 X16 percent of population family income .41062 .01025 .17363
non-farm
X23 percent of property value commercial .42019 .00957 .21962








64

The percent of property value farms explained a little more than onequarter of the variance in tax effort for these districts. Average daily membership accounted for 10 percent of the variance and the total listed value of real estate 9 percent. The variable percent of population family income wage and salary explained 3 percent while percent of population family income farm, percent of population family income other and percent of population family income non-farm, each explained 1 percent of the variance in tax effort.

Although relationship is not high enough to be conclusive, what is suggested from these results is that districts with a high percentage of farm property make a high effort in the support of education. Likewise, districts with expanding enrollments or high enrollments made a significant tax effort as do districts which have a high total listed value of real estate. The income variables, i.e., percent of population family income wage and salary, percent of population family income farm, percent population family income other and percent of population family income non-farm had little significance in explaining tax effort in the 67 less wealthy districts.

While the variable (X23) percent property value commercial explained less than one percent of the variance in effort, it did have a Simple R of .2196.

The findings indicate that the most significant variable for all

districts and the 67 wealthier districts was the percent of all property which is residential. For the 67 low wealth districts the most significant variable was the percent of property value which was farm property. Districts which had high residential and farm property were evidently usually willing to make greater tax effort in the support of schools.








65

Since few income variables were included in the regression equation, it might be concluded that income should not be considered a determiner of tax effort.

Factor Analysis of Variables

As noted in Chapter III, it was suggested that many of the independent variables might be highly inter-correlated and that in the regression equation some variables might "mask" the significance of others. The independent variables were therefore subjected to factor analysis.

The factor analysis does not itself identify the factors that are extracted from the data. There is, however, a correlation coefficient between each variable and each factor. Thus, relationships between factors and variables provide clues regarding the nature of the factors. In the following summary table (Table XVIII) all correlations of .40 or higher between the twenty-five independent variables and the six factors are reported. The .40 cutoff was chosen after consultation with statisticians who suggested the lowest limit generally considered as aiding in the description between a variable and a factor as being .30. The .40 level was chosen as being rather conservative.

In factor 1, total school expenditures, correlated at .98, while

average daily membership and total school current expenditures correlated at .99. Variable X4 (total listed value real estate) correlated at .82 and percent property value commercial correlated at .48. So factor 1 related strongly to expenditures, students, and property values, and was given the label of "wealth and expenditures."

In factor 2, three variables correlated at .82 or higher. These three variables were percent population 18-65 years of age, percent of

ILeonard Tashman and David Bee, private discussion at University of Vermont, February and March, 1975.







66

TABLE XVIII Independent Variables Correlating with Factors at .40 or Higher


Factors
Variable 1 2 3 4 5 6

X Average daily membership .99 X2 Total school expenditures .98 X" Total school current expenditures .99


X4 Total listed value X5 Percent population X6 Percent population X7 Percent population
K-12
X8 Percent population X9 Percent population
high school
X10 Percent population X 1 Percent population
$5999
XI2 Percent population
$9999
X13 Percent population
$10,000-$24,999
XI4 Percent population
$25,000-over
X15 Percent population
and salary


of real estate 18-65 years of age 65 and over in parochial schools in school, age 5-19 25 and over completing 18-64 in labor force family income $1000family income $6000family income family income family income wage


.45 .40 -.51


population family population family population family and retirement population family population family


income income income

income income


non-farm farm social

welfare other


economically deprived children residential property value property value commercial property value industrial Property value farms


X 16 X17 X18

X19
X20 X21 X22 X23 X24 X25


Percent Percent Percent security Percent Percent Percent Percent Percent Percent Percent


.40 .41


Y








67

population 18-64 in labor force, and percent population with family income from wage and salaries. Percent of population 25 and over completing high school, and percent of population with family income $10,000-$24,999 correlated at .51. Three additional variables correlated between .40 and .49. These variables were percent population family income other .40, percent family income $6000-$9999, .41, and percent of population in school ages 5-19, .49.

Factor 2 suggests some form of urbanization, since high levels of education and income are related and a high percentage of people in the labor force drawing their income from wage and salaries tend to be located in most urban areas. This factor, then, was labeled "urbanization."

In factor 3, percent population 65 and over and percent population family income social security and retirement correlated at .87 with the factor. Percent population family income $1000-$5999 correlated at .45 and percent population family income other at .41.

Factor 3, therefore, would seem to predominate in areas where there is a relatively older population with limited incomes. These could be rural districts where the young people are moving out of. This factor was designated "elderly limited income."

Percent population family income farm correlated at .86 with factor 4 and percent property value farms correlated at .61. Percent population family income $1000-$5999 correlated at .40. The label provided this factor was "farming district."

In factor 5 percent population family income non-farm and family

income other correlated at .45 and .46, respectively. Percent population family income $1000-$5999 had a negative correlation of -.51 with the factor, while percent population family income $10,000-$24,999 had a








68

positive correlation of .41. This factor would suggest a rural middle class "bedroom" or residential community. The factor designation, therefore, was "residential middle class district."

In factor 6 the only variable with a correlation above .40 was percent of population family income $6000-$9999. This variable correlated with factor 6 at .85. This would indicate a "low middle class district" and the factor was so labeled.

A factor correlation matrix was completed to see if any of the factors correlated significantly with each other. (See Table VIII) The study indicated that no one factor was significantly correlated with another. As a matter of fact, factor 2, designated urbanization, had the highest correlation of .0768 with factor 5, residential middle class district. This suggests that no one factor could be identified as being related to tax effort. Regression Analysis of Factor Scores

As indicated in Chapter III, factor scores for each district were

computed and the factors placed in a stepwise multiple regression equation with tax effort as the dependent variable. This was done for all 201 school districts and the five factors which remained in the equation explained a little more than 8 percent of the variance in effort. Wealth and expenditures explained 6 percent of the variance, while urbanization explained approximately 2 percent. The other three factors, low middle class, residential middle class and farming districts explained less than one-half percent of the variance in tax effort.

When all independent variables were placed in the regression equation, they explained 40 percent of the effort instead of the 8 percent explained by the factor scores. The reason for this could be that the factors represented the averaging of a number of related variables. Thus, a greater








69

proportion of the variance in effort would be explained by individual variables since they were not comingled with groups of other variables which might reduce their individual correlation with effort.

The high wealth districts were next examined. In the factor correlation matrix the simple correlation with the exception of two, all increased, but not significantly. The correlations between the factors wealth and expenditures and residential middle class decreased, as did the correlation between farming districts and residential middle class. The highest correlations were between urbanization and elderly limited income, a negative -.27; and wealth and expenditures and low middle class, a negative -.28.

Even though these factors are not significantly correlated, it would tend to indicate that the urbanized high wealth areas do not have the element of elderly with limited income. Likewise, the high wealth districts with a substantial number of low income residents do not have high expenditures.

In the stepwise regression of the high wealth districts, using factor scores, 7 percent of the variance in effort was atrributable to 4 factors. The factors identified as urbanization and farming were eliminated from the equation. As a matter of fact, factor 5, residential middle class explained

5 percent of the variance, while the remaining three factors together explained 2 percent of the variance in tax effort. Since the variance in tax effort of high wealth districts as identified through the stepwise regression equation is so slight, the conclusion reached is that high wealth districts cannot be predicted through factors of socioeconomic variables.

The low wealth districts were then examined. In the factor correlation matrix the simple correlations between expenditures and wealth and urbanization had a negative correlation of -.40, a substantial increase over the .09 correlation between the same factors in high wealth districts.









70

This indicated that the low wealth urbanized areas do not make high

expenditures or have the property wealth to do so. The urbanized low wealth districts would also tend to have a higher percentage of "elderly with limited income" residing in them as evidenced by the positive simple correlation of .24.

In the stepwise regression for the low wealth districts, all five

factors combined explained 25 percent of the variance in tax effort. The factor "elderly with limited income" was dropped from the equation. Farming districts explained 11 percent of the variance in tax effort, while urbanization explained 5 percent, and wealth and expenditures explained 6 percent, while residential middle class was responsible for 4 percent. This would tend to indicate that many of the low wealth districts are farming communities and they make the greatest effort in the support of schools in these low wealth districts. Urbanized and residential middle class districts made about the same amount of effort.

Finally, it should be noted that a greater amount of variance in tax effort was explained by the independent variables and factors in the low wealth districts than in all districts and in high wealth districts. In the regression equation using the independent variables, the total amount of variance in effort explained for all 201 school districts was 41 percent. In the high wealth districts it was 7 percent; but in the low wealth districts, it was 42 percent. Similarly when the factors were used in the regression equation for all districts, they explained but 8 percent of the variance in effort. In the high wealth districts 7 percent of the variance in effort was explained and in the low wealth districts a total of 25 percent of the variance in effort was explained.









71

The findings suggest that for all 201 school districts, as well as for the 67 high wealth and 67 low wealth districts, variables other than demographic ones may be more important in determining the amount of local tax effort a community makes in the support of education. Some of these non-demographic variables might be a type of local power structure, local tradition regarding education or the attitudes of community leaders toward schooling.

Since Vermont is a small rural state organized into many small sparcely populated school districts, the accuracy of the fourth count census data may be questioned. Complete census surveys were conducted in approximately 8% of the districts used in this study. The remaining districts responded to a 15% sampling and thus the socioeconomic variables may have lacked precision of scale to make reliable predictions.














CHAPTER V

CONCLUSIONS AND RECOMMENDATIONS

This study found no significant factors correlated to tax effort which would predict the type of districts that make a substantial tax effort for the support of schools in Vermont. However, a study of the independent socioeconomic variables identifies districts with high residential property values as the ones which make a greater tax effort in the support of schools in all districts and high wealth districts.

Likewise, it appears, but is not conclusive, that in low wealth districts, districts with high farm value make a greater effort in the support of education than do other types of communities. This study also suggests, but once again is not statistically significant, that in districts with high concentrations of economically disadvantaged children, less tax effort is made for the support of education. It has been found in at least one other Vermont studyI that low income individuals generally have low educational aspiration levels and, therefore, do not consider education important enough to make a significant effort in its support.

Most studies have generally found one or more socioeconomic variables reflecting demands for education to be significant. However, there has not been agreement on which variables are crucial. Even where identical models have been used for several states, the importance and even the sign of such variables are not always consistent. As was indicated earlier, the method

1Steven F. Hochschild, "Postsecondary Education Access Study," State of Vermont, Commission on Higher Education Facilities, Montpelier, Vt., 1972.

72









73

of collecting socioeconomic data and the lack of precision of scale of these variables may have added to the source of measurement error in this particular study. Another factor which could have contributed to the low correlations between the socioeconomic variables and the dependent variable tax effort, aside from those already brought forth is the optional local effort which permits school districts to decide on the amount of money they elect to raise for the support of schools and methods used in property tax administration.

It appears, even though the correlations were low, that the wealthy residential areas consistently tend to tax themselves proportionately more than poor rural districts. A state local finance system permitting optional local effort which operates to the benefit of these residential districts tends to provide the greatest amount of educational resources per child in those districts. This study, as many others, did not highly correlate property valuation and family income. These two variables measure different aspects of fiscal ability, that is to say that commercial and business properties are not included or reflected in personal income; yet taxes which individuals pay, regardless of the base, are paid from current income.

Although the study does not fully explain the reasons for the wide

variations in tax effort for the support of schools in Vermont, wide variations do in fact exist. Because of this wide variation, the state should intervene and provide assistance which would reduce the disparities in expenditures for education.

Further studies are recommended to determine if geographic and political differences in school districts affect the level of the effort they make in the support of education in Vermont.












APPENDIX A


RANK ORDER OF VERMONT SCHOOL DISTRICTS BY TAX EFFORT INDEX FROM HIGH TO LOW 1973


School District


Index Rank


School District


i 1


Bristol East Montpelier Hinesburg Middlesex Barnet

Hardwick Bradford I.D. Whiting Calais Monkton

South Burlington Montpelier Shelburne Williston Charlotte

Chester Waterbury Berlin Cornwall Johnson

Cabot New Haven Duxbury Milton I.D. Proctor

Rockingham St. Johnsbury Franklin Jericho Swanton


Danby Weybridge Brookfield Brattleboro Fair Haven


3.06 2.82 2.80 2.70
2.61

2.55 2.51 2.45 2.43 2.35

2.34 2.31 2.31 2.31 2.30

2.24 2.24 2.23
2.21
2.19

2.16 2.15 2.13
2.10
2.10

2.07
2.05 2.03 2.03 2.00

1.99 1.98 1.97 1.95 1.95


36. 37. 38. 39.
40.

41. 42. 43. 44. 45.

46. 47. 48. 49. 50.

51. 52.
53. 54. 55.

56.
57. 58. 59. 60.

61. 62. 63.
64. 65.

66. 67. 68. 69.
70.


Westford East Haven Vergennes I.D. Moretown Pittsford

Benson Colchester Randolph Castleton Newport City

Orwell Richmond Huntington Sunderland Waterford

Hartford Hyde Park Putney Middlebury I.D. Readsboro

Derby Springfield Winooski Orange Arlington

Richford Clarendon Ferrishurg St. Albans City
Watervi le

Newark Worcester
Lowell Williamstown West Rutland


Rank


Index


1.94 1.90 1.90 1.89
1.85

1.84 1.84 1.84 1.83 1.83

1.82 1.82 1.81
1.81 1.80

1.79 1.78
1.78 1.77 1.77

1.76 1.76 1.75 1.74 1.73

1.73
1.72 1.72
1.72 1.71

1.69 1.69 1.68 1.68 1.67








Appendix A - Continued


School District


71. 72. 73. 74. 75.

76. 77. 78. 79. 80.

81. 82. 83. 84. 85.

86. 87. 88. 89. 90.

91. 92. 93. 94. 95.

96. 97. 98. 99. 100.

101. 102. 103. 104. 105.

106. 107. 108. 109. 110.


Index Rank


Albany Ludlow Rutland City Grand Isle Troy

Weathersfield West Fairlee Bridport Shoreham Alburg

Burlington Hartland Leicester Westminster Charleston

Fairfield Kirby Norton St. Albans Town Georgia

Glover Highgate Barre City Roxbury Northfield

Peacham Danville Concord Irasburg Lincoln

Rutland Town Cambridge Manchester Windsor Addison

Newport Town Barre Town Wallingford Chelsea Ira


i i


1.66 1.66 1.66
1.65 1.64

1.64 1.64 1.62 1.62
1.61

1.61 1.61
1.61 1.60 1.59

1.58 1.58 1.58 1.57 1.55

1.55 1.55
1.54 1.54 1.52

1.52 1.51 1.50 1.50 1.50

1.50
1.49 1.48 1.48 1.47

1.47 1.46 1.46 1.46
1.46


School District


Whitingham Coventry Washington Braintree Craftsbury

Fletcher Sutton Newfane Stamford Middletown Springs

Shaftsbury Berkshire Bridgewater Pownal Athens


I11.
112. 113. 114. 115.

116. 117. 118. 119. 120.

121. 122. 123.
124. 125.

126. 127.
128. 129. 130.

131. 132.
133. 134. 135.

136. 137. 138. 139.
140.

141. 142. 143. 144.
145.

146. 147. 148. 149. 150.


Rank


Bolton Reading Windham Lyndon Sheldon

Wolcott Bethel Woodstock Canaan Enosburg Falls

Fairlee Starksboro West Haven Woodbury Poultney

Salisbury Brownington Rochester Baltimore Rupert

Tunbridge Vershire Isle LaMotte Waitsfield Eden


Index


1.46 1.45 1.45 1.44 1.43

1.43 1.43 1.41 1.41 1.40

1.40 1.39
1.39 1.39 1.36

1.36 1.36 1.36 1.35 1.34

1.34 1.33 1.32 1.31 1.31

1.31 1.31 1.31 1.30
1.29

1.29 1.28 1.28 1.27 1.27

1.26 1.26 1.25 1.22 1.21








Appendix A - Continued


Rehonl District


Thetford Westfield Burke Tinmouth Strafford


151.
152. 153. 154. 155.

156. 157. 158.
159. 160.

161. 162. 163. 164. 165.

166. 167. 168. 169. 170.

171. 172. 173. 174. 175.

176. 177. 178. 179. 180.

181. 182. 183. 184. 185.

186. 187. 188. 189. 190.


Index Rank


'R ~~ -q n cho isrc


1.19 1.19 1.18
1.18 1.17

1.17 1.16 1.16 1.15 1.14

1.13 1.11
1.10 1.09 1.09

1.09
1.06 1.05
1.02 1.01

1.01 .96 .96 .96 .92


191.
192. 193.
194. 195.

196. 197. 198. 199.
200.

201.


School District


Vernon Warren Marlboro Dover Wells

Goshen Peru Fayston Landgrove Plymouth

Winhall


Index


Thank


West Windsor Morristown Stockbridge Chittenden Wilmington

Pomfret Bakersfield Stowe Elmore Walden

Woodford Granville South Hero Belvidere Jamaica

Pawlet Londonderry Montgomery Pittsfield Dorset

Lunenburg Halifax Barnard Greensboro Guildhall

Grafton Jay
Hubbardton Mt. Holly Ripton

Sherburne Hancock Morgan Shrewsbury North Hero




AFPENDIX B
CORRELATION COEFFICIENTS

X1 X2 X3 X4 X5 X6 X7 X8 X9 X1O
X1 1.00000 0.96167 0.96654 0.57306 0.10043 0.12622 0.06241 0.11728 0.14669 0.11340 X2 0.96167 1.0000 0.97721 0.65477 0.07288 0.11325 0.02987 0.08025 0.12748 0.10348 X3 0.96654 0.97721 1.00000 0.66881 0.06853 0.11012 0.01783 0.07169 0.13222 0.10600 X4 0.57306 0.65477 0.66881 1.00000 -0.06794 0.02695 -0.05718 -0.06850 -0.02146 -0.02124
X5 0.10043 0.07288 0.06853 -0.06794 1.00000 0.20654 -0.08764 0.63200 0.59031 0.87564
X6 0.12622 0.11325 0.11012 0.02695 0.20654 1.00000 -0.02823 0.24518 0.44497 0.24165
X7 0.06241 0.02987 0.01783 -0.05718 -0.08764 -0.02823 1.00000 0.08155 0.02498 -0.12333
X8 0.11728 0.08025 0.07169 -0.06850 0.63200 0.24518 0.08155 1.00000 0.51320 0.53693
X 0.14669 0.12748 0.13222 -0.02146 0.59031 0.44497 0.02498 0.51320 1.00000 0.55523
X 0.11340 0.10348 0.10600 -0.02124 0.87564 0.24165 -0.12333 0.53693 0.55523 1.00000
X -0.11377 -0.17602 -0.17309 -0.30998 0.29348 0.29348 0.01147 0.34526 0.36225 0.16177
X12 -0.10036 -0.08638 -0.11407 -0.18572 0.52499 0.13417 -0.10191 0.28259 0.18704 0.52721
XI3 0.20905 0.19856 0.22921 0.20764 0.47782 0.28374 0.02377 0.37575 0.60884 0.60625 X14 0.19073 0.22323 0.20730 0.19590 0.24355 0.14776 0.20548 0.46042 0.06954 0.24309 xi5 0.06314 0.02011 0.02371 -0.09864 0.78631 0.38327 -0.04063 0.53751 0.61888 0.84405
X16 0.23356 0.23528 0.22866 0.12590 0.41733 0.15098 0.01190 0.24429 0.30949 0.52487 X17 -0.07492 -0.11833 -0.10579 -0.18620 0.28918 0.22815 0.03392 0.35792 0.42363 0.19781
X18 0.06007 0.02244 0.03788 -0.03700 0.26515 0.82992 0.00248 0.23406 0.47882 0.31383
X19 -0.09322 -0.12963 -0.11864 -0.12526 0.27765 0.11730 -0.09042 0.23226 0.22978 0.24416
X20 0.17171 0.13418 0.16187 0.09317 0.59595 0.40331 -0.08249 0.45733 0.60715 0.69926
X21 -0.31251 -0.28809 -0.28361 -0.21275 0.08758 0.21257 -0.08432 -0.06807 0.18174 0.02140
X22 0.43852 0.37160 0.37305 -0.06948 0.24250 0.25300 0.04128 0.15459 0.21144 0.27853

X23 0.14043 0.20613 0.23094 0.24090 -0.30977 -0.39380 0.07246 -0.24375 -0.20718 -0.23403
X24 0.11625 0.09160 0.06001 0.21685 0.23780 0.14520 0.10861 0.17389 0.20887 0.24305 X-_ 0.01664 -0.04275 -0.05652 -0.23453 -0.03164 0.06332 0.05389 -0.04142 -0.00889 -0.05912




Appendix B - Continued

Xii X12 X13 X14 X15 X16 X17 X18 X19 X20
X -0.11377 -0.10036 0.20905 0.19073 0.06314 0.23356 -0.07492 0.06007 -0.09322 0.17171
X2 -0.17602 -0.08638 0.19856 0.22323 0.02011 0.23528 -0.11833 0.02244 -0.12963 0.13418
X3 -0.17309 -0.11407 0.22921 0.20730 0.02371 0.22866 -0.10579 0.03788 -0.11864 0.16187
X4 -0.30998 -0.18572 0.20764 0.19590 -0.09864 0.12590 -0.18620 ,-0.03700 -0.12526 0.09317
X5 0.29348 0.52499 0.47782 0.24355 0.78631 0.41733 0.28918 0.26515 0.27765 0.59595 X6 0.45911 0.13417 0.28374 0.14776 0.38327 0.15098 0.22815 0.82992 0.11730 0.40331 X7 0.01147 -0.10191 0.02377 0.20548 -0.04063 0.01190 0.03392 0.00248 -0.09042 -0.08249
X8 0.34526 0.28259 0.37575 0.46042 0.53751 0.24429 0.35792 0.23406 0.23226 0.45733 X9 0.36225 0.18704 0.60884 0.06954 0.61888 0.30949 0.42363 0.47882 0.22978 0.60715 X10 0.16177 0.52721 0.60625 0.24309 0.84405 0.52487 0.19781 0.31383 0.24416 0.69926 X 1.00000 -0.02397 0.04427 -0.13605 0.37408 -0.08979 0.54446 0.48935 0.21492 0.13437
X12 -0.02397 1.00000 -0.08917 0.10401 0.52939 0.18607 0.03164 0.13651 0.39863 0.37459
X13 0.04427 -0.08917 1.00000 0.06538 0.64483 0.43695 0.07728 0.37254 0.04756 0.55109
X14 -0.13605 0.10401 0.06538 1.00000 0.10017 0.19864 0.02420 0.02365 0.10262 0.18579
X15 0.37408 0.52939 0.64483 0.10017 1.00000 0.37474 0.20131 0.46465 0.34584 0.65206 X16 -0.08979 0.18607 0.43695 0.19864 0.37474 1.00000 0.00607 0.11848 -0.03616 0.48295
X17 0.54446 0.03164 0.07728 0.02420 0.20131 0.00607 1.00000 0.26528 -0.00581 0.28185
X18 0.48935 0.13651 0.37254 0.02365 0.46465 0.11848 0.26528 1.00000 0.19123 0.46921 X19 0.21492 0.39863 0.04756 0.10262 0.34584 -0.03616 -0.00581 0.19123 1.00000 0.25414
X20 0.13437 0.37459 0.55109 0.18579 0.65206 0.48295 0.28185 0.46921 0.25414 1.00000 X21 0.33293 -0.03362 0.02573 -0.15305 0.09696 -0.04364 0.32559 0.30557 -0.13322 0.13782
X22 0.05075 0.10340 0.17075 0.06871 0.21998 0.16748 -0.08992 0.13442 -0.05036 0.29955
X23 -0.40928 -0.16949 -0.06851 -0.00047 -0.32764 0.05851 -0.31815 -0.27237 -0.25495 -0.27546
X24 0.01670 0.12117 0.28446 0.01261 0.33808 0.00646 0.06277 0.26179 0.13262 0.26106 X 0.31028 -0.02644 -0.21653 -0.08689 -0.05164 -0.07476 0.57197 -0.01194 -0.05764 -0.07975




Appendix B - Continued


X21 X22 X23 X24 X25 X1 -0.31251 0.43852 0.14043 0.11625 0.01664 X2 -0.28809 0.37160 0.20613 0.09160 -0.04275 X3 -0.28361 0.37305 0.23094 0.06001 -0.05652 X4 -0.21275 -0.06948 0.24090 0.21685 -0.23453 X5 0.08758 0.24250 -0.30977 0.23780 -0.03164 X6 0.21257 0.25300 -0.39380 0.14520 0.06332 X7 -0.08432 0.04128 0.07246 0.10861 0.05389 X8 -0.06807 0.15459 -0.24375 0.17389 -0.04142 X9 0.18174 0.21144 -0.20718 0.20887 -0.00889 X 0.02140 0.27853 -0.23403 0.24305 -0.05912 X 0.33293 0.05075 -0.40928 0.01670 0.31028 x -0.03362 0.10340 -0.16949 0.12117 -0.02644 X13 0.02573 0.17075 -0.06851 0.28446 -0.21653 X14 -0.15305 0.06871 -0.00047 0.01261 -0.08689 X15 0.09696 0.21998 -0.32764 0.33808 -0.05164 X16 -0.04364 0.16748 0.05851 0.00646 -0.07476 XI7 0.32559 -0.08992 -0.31815 0.06277 0.57197 X18 0.30557 0.13442 -0.27237 0.26179 -0.01194 XI9 -0.13322 -0.05036 -0.25495 0.13262 -0.05764 x20 0.13782 0.29955 -0.27546 0.26106 -0.07975 X21 1.00000 -0.28936 -0.03764 0.15643 0.13277 X22 -0.28936 1.00000 -0.26148 -0.16012 -0.05501 X23 -0.03764 -0.26148 1.00000 -0.18155 -0.30106 X24 0.15643 -0.16012 -0.18155 1.00000 -0.03959 X25 0.13277 -0.05501 -0.30106 -0.03959 1.00000














BIBLIOGRAPHY


Adams, Perry R., "Socioeconomic Factors Associated with Patterns of School
Fiscal Policy in Kentucky," Doctoral Dissertation, University of
Florida, Gainesville, 1965.

Alkin, Marvin, "Religious Correlates of School Expenditures," A paper
prepared for the American Research Association, Chicago, February 11,
1965, ERIC, ED 011143.

Campbell, Alan, "The Socioeconomic, Political and Fiscal Environment of
Educational Policy Making in Large Cities," Michael Kirst, (Ed), The
Politics of Education, McCutchan Publishing Company, Berkeley,
California, 1970.

Campbell, Alan and Shalala, Donna, "Resource Literature and Educational
Revenue," Theory Into Practice, Ohio State University, Vol. XI, No. 2,
April 1972.

Carlson, DeVere, "Patron Attitudes Toward Selected Educational Issues in
Communities with a Dual Educational System," Doctoral Dissertation,
University of Pittsburgh, 1969, Dissertation Abstracts, Vol. 30.

Clune, William H., "Taxing and Spending for Public Schools; The Origin
Descriptions and Effects of Non-School Taxes and Industrial and
Commercial Property," Education and Government Division, Illinois
Bureau of the Budget, Chicago, 1971.

Cooley, W. and Lohnes, P., Multivariate Procedures for the Behavioral
Sciences, John Wiley and Sons, New York, 1962.

Cubberley, E., Public Education in the United States, Houghton Mifflin Co.,
Riverside Press, Cambridge, Massachusetts, 1934.

Cubberley, E., School Funds and Their Apportionment, Teachers College,
Columbia University, New York, 1905.

Dixon, W. T., (Ed.) BMD, Biomedical Computer Programs, University of
California Press, Berkeley, 1968.

Dye, Thomas, "Politics, Economics, and Educational Outcomes in the States,"
University of Georgia, Athens, 1967.

Ellis, J., Jr. "A Study of the Relationship Between Local Expenditures on
Education and Local Income," Doctoral Dissertation, University of
Virginia, 1967, Dissertation Abstracts, Vol. 27.

Farner, Frank, "Economic Sociological and Demographic Characteristics of
Oregon School Districts and Their Relationship to District Financial Practices," University of Oregon, Bureau of Educational Research and
Services, Eugene, April 1966.
80









Fisher, Jack, "A Comparison Between Central Cities and Suburbs and Local
Ability to Support Public Education," Doctoral Dissertation,
University of Florida, Gainesville, 1972.

Gallup, George, "Fourth Annual Gallup Poll of Public Attitudes Toward
Education," Phi Delta Kappan, Vol. LIX, No. 1, September, 1972.

Garrison, C. B., "An analysis of the Interrelationships of Economic Index
of Taxpaying Ability for Schools of Arkansas Counties," Doctoral
Dissertation, University of Arkansas, 1965, Dissertation Abstracts,
Vol. 26.

Gentry, Gilbert, "The Relationship of Certain Cultural Factors to Initiative
in the Local Support of Education in Florida," Doctoral Dissertation,
University of Florida, Gainesville, 1959.

Harvey, E. L., "Property Tax Determinants of Educational Expenditures,"
Doctoral Dissertation, Stanford University, 1969, Dissertation
Abstracts, Vol. 30.

Hendrix, V. and Alkin, M., "Population Age Distribution and Public Educational
Expenditures." A paper, Educational Research Association, New York,
February 1967.

Hochschild, Steven F., "Postsecondary Education Access Study," State of
Vermont, Commission on Higher Education Facilities, Montpelier, Vt., 1972.

Hooper, Harold, "Socioeconomic Factors Associated with Patterns of School
Fiscal Policy in Florida," Doctoral Dissertation, University of
Florida, Gainesville, 1965.
"How to Evaluate Your District's Financial Effort," School Management,

January 1965.

"Income v. Effort," School Management, January, 1969.

James, Thomas; Kelley, J.; and Garms, W., "Determinants of Educational
Expenditures in Large Cities of the United States," School of Education,
Stanford University, Palo Alto, California, 1966.

James, H.; Thomas, A.; and Dych, H., "Wealth Expenditures and Decision
Making for Education," USOE, Research Project No. 1241, Stanford
University, Stanford, California, June 1963.

Johns, R.; Alexander, K.; and Jordan, K. F., Financing Education - Fiscal
and Legal Alternatives, Charles E. Merrill, Columbus, Ohio, 1972.

Johns, R.; Alexander, K.; and Stollar, D.; (Eds.) Status and Impact of
Educational Finance Programs, Vol. 4, National Educational Finance
Project, Gainesville, Florida, 1971.

Johns, R. and Kimbrough, R., "The Relationship of Socioeconomic Factors,
Educational Leadership Patterns and Elements of Community Power
Structures to Local School Fiscal Policy," USOE, Project No. 2842,
May 1968, ERIC ED 021336.









Kay, Harold B., "A Study of the Relationships Between Selected Socioeconomic
Variables and Local Tax Effort to Support Public Schools in Kentucky,"
Doctoral Dissertation, University of Florida, Gainesville, 1973.

King, Charles R., "Socioeconomic Factors Associated with Patterns of
School Fiscal Policy in Georgia," Doctoral Dissertation, University of
Fiorida, Gainesville, 1965.

Lazier, Willard E., "School Finance: A Determinant Model for Eligibility,"
Doctoral Dissertation, University of Utah, Salt Lake City, 1972.

MacDougall, M. A., "An Analysis of Income Flow to Educational Expenditures,"
Doctoral Dissertation, University of Virginia, 1964, Dissertation
Abstracts, Vol. 24.

Martin, Charles E., "The Relationship of Social and Economic Characteristics
to Local Initiative in the Financial Support of Public Schools in
Mississippi," Doctoral Dissertation, University of Southern Mississippi,
1964, Dissertation Abstracts, Vol. 23.

Metzcus, Richard H., "Community Human Resources and Local Financial Support
for Public Schools," Doctoral Dissertation, University of Illinois,
1969, Dissertation Abstracts, Vol. 30.

Meyers, Alfred V., "The Financial Crisis in Urban Schools-Patterns of Support
and Non-Support Among Organized Groups in Urban Communities,"
Doctoral Dissertation, Wayne State University, 1965, Dissertation
Abstracts, Vol. 25.

Miner, Jerry, Social and Economic Factors in Spending for Public Educations,
Syracuse University Press, Syracuse, 1963.

Mort, P. R., and Reusser, W. C., Public School Finance, McGraw-Hill Book
Co., Inc., New York, 1941.

Myers, H. 0., "A Study of Certain Phases of Local Tax Effort in Relation
to Taxpaying Ability in Florida," Masters Thesis, University of
Florida, Gainesville, 1950.

National Educational Finance Project, "Future Directions for School
Financing," Gainesville, Florida, 1971.

Nie, Norman; Hull, C. Hadley; and Bent, Dale H., Statistical Package for
the Social Sciences, McGraw-Hill, New York, 1970.

Patterson, Wade, and Schoonhoven, John, "A Comparative Study of Inconsistent
Voter Behavior in School Budget Elections," November 1966, ERIC,
ED 011135.

Petersen, Thor, "School Approval-Disapproval and Educational Enlightenment
of Parents Based on Occupation, Educational Level, Age, Race, Geographic
Area and Length of Residency," Doctoral Dissertation, Michigan State
University, 1971, Dissertation Abstracts, Vol. 32.










Peterson, L., Rossmiller, R., North, S., and Wakefield, H., "Economic
Impact of State Support Models on Education," University of Wisconsin,
Madison, 1963.

Photiadas, J., and Zeller, F., "Attitudes Toward State and Local Taxes in
West Virginia: The Preliminary Results of a Survey," University Center
for Appalachian Studies and Development, Morgantown, West Virginia, 1968.

Quick, W. J. "Socioeconomic Factors Associated with Patterns of School
Fiscal Policy in Illinois," Doctoral Dissertation, University of Florida,
Gainesville, 1965.

Rodriguez v. San Antonio Independent School District, 337 F. Supp. 280 (1971)
Reversed 41 "Law Week" 4407 March 21, 1973.

Serrano v. Priest, 96 Cal. Rptr., 601, 487 P. 2d 1241 (1971).

Strayer, G., and Haig, R., "The Financing of Education in the State of New
York," Report of the Educational Finance Inquiry Commission, Vol. I,
Macmillan Co., New York, 1923.

Turck, M., "A Study of the Relationships Among the Factors of Financial
Need, Effort and Ability in 581 High School Districts in Michigan,"
Doctoral Dissertation, Michigan State University, 1966, Dissertation
Abstracts, Vol. 21.

United States Bureau of the Census, "Public Use Samples of Basic Records
from the 1970 Census," Washington, D. C., United States Department
of Commerce, April 1972.

Vermont Department of Education, "Vermont State Aid," Montpelier, Vermont
1972.

Von Hatley, R. F., "Family Income Voting Behavior and Financial Referendums:
Educational Finance and Politics in Albuquerque, 1968-69," Doctoral
Dissertation, University of New Mexico, 1971, Dissertation Abstracts,
Vol. 31.

Witt, Irving, and Pearce, Frank, "A Study of Voter Reaction to a Combination
Bond-Tax Election on March 26, 1968," ERIC, ED 019945.
















BIOGRAPHICAL SKETCH

Edward Joseph Fabian was born October 7, 1937 in Middle Granville, New York. He graduated with a B.S. in Education from Castleton State College in 1961, from St. Michael's College in 1966 with a M.Ed. in Guidance and Administration, from the University of Vermont in 1973 with a C.A.G.S. in Educational Planning and Administration, and a Ph.D. in Educational Administration from the University of Florida in 1976.

Mr. Fabian worked as an elementary and junior high school teacher in Orwell, Vermont and as an elementary principal in Orwell, Vermont. He moved to the State Department of Education in 1966 as Chief of Education Field Services and currently serves as Deputy Commissioner of Education.

Mr. Fabian has been included in Outstanding Young Men of America (1967), Who's Who in State Government (1975) and is a member of Phi Delta Kappa, Vermont Education Association, American Association of School Administrators, as well as many other organizations. He is married to the former Martha Anne Ryan and has five children: four boys, Charles, Kevin, Christopher, Edward, Jr. and one girl, Julianne.









I certify that I have read this study and that, in my opinion, it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.




Kerh Alexander, Ohairman
Professor of Educational Administration





I certify that I have read this study and that, in my opinion, it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.




' William Al xander, Professor"of Curricu!Am and Instrt tion





I certify that I have read this study and that, in my opinion, it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.




Arthur J. Lewis, Professor of Curriculum and Instruction





I certify that I have read this study and that, in my opinion, it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.




Michael Nunner y Professor of Educat nal Administration













This dissertation was submitted to the Dean of the College of Education and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy.


August 1976


D an, Co id e cho ol



Dean, Graduate School




Full Text

PAGE 1

THE RELATIONSHIPS BETWEEN SELECTED SOCIOECONOMIC VARIABLES AND LOCAL TAX EFFORT TO SUPPORT PUBLIC SCHOOLS IN VERMONT by EDWARD JOSEPH FABIAN A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1976

PAGE 2

ACKNOWLEDGMENTS This writer wishes to extend his appreciation to Dr. Kern Alexander, Chairman of his Supervisory Committee, for his counsel, assistance and support throughout this study. Thanks are also expressed to other members of his committee for their assistance. A debt of gratitude is owed Mr. Earl Blekking of the University of Florida for his guidance throughout the array of statistical procedure and to Mrs. Connie Hebert of the Vermont State Department of Education for the technical assistance she provided. The researcher expresses the deepest gratitude to his wife, Martha, and children, Charles, Kevin, Chris, Edward, Jr. and Julianne for their many sacrifices and encouragements throughout the writing of this study. ii

PAGE 3

TABLE OF CONTENTS Page ACKNOWLEDGMENTS li LIST OF TABLES iv ABSTRACT vi CHAPTER I. INTRODUCTION 1 Statement of the Problem 3 Justification for the Study 3 Assumptions 4 Definition of Terms 4 Procedures 5 .II. REVIEW OF LITERATURE AND RELATED RESEARCH 8 National Historical Overview 8 Related Studies 15 Summary 32 III. STATISTICAL PROCEDURES AND PRESENTATION OF THE DATA .. 33 The Variables 34 Statistical Procedure 36 Presentation of the Data 39 IV. FINDINGS AND ANALYSIS 56 V. CONCLUSIONS AND RECOMMENDATIONS 72 APPENDIX A. RANK ORDER OF VERMONT SCHOOL DISTRICTS BY TAX EFFORT INDEX FROM HIGH TO LOW B. CORRELATION COEFFICIENTS BIBLIOGRAPHY BIOGRAPHICAL SKETCH ill

PAGE 4

LIST OF TABLES Page TABLE I Zero Order Correlations Between the Dependent Variable and the Independent Variables TABLE II Relationship of Tax Effort to 2A Socioeconomic Variables in 201 School Districts (Multiple Regression Analysis) TABLE III Relationship of Tax Effort to 24 Socioeconomic Variables in 67 High Wealth Districts (Multiple Regression Analysis) TABLE IV Variables Not In the Equation of 67 High Wealth Districts TABLE V Relationship of Tax Effort to 24 Socioeconomic Variables in 67 Low Wealth Districts (Multiple Regression Analysis) TABLE VI Variables Not in Equation of 67 Low Wealth Districts 40 41 44 45 47 48 TABLE VII Rotated Factor Matrix Derived from the Independent Variables TABLE VIII Factor Correlation Matrix for 201 School Districts 49 52 TABLE IX Relationship of Five Factors to Tax Effort in 201 School Districts (Multiple Regression Analysis) 52 TABLE X Relationship of Four Factors to Tax Effort in 67 High Wealth Districts 54 TABLE XI Factor Correlation Matrix for 67 High Wealth Districts 54 TABLE XII Relationship of Five Factors to Tax Effort in 67 Low Wealth Districts 55 TABLE XIII Factor Correlation Matrix for 67 Low Wealth Districts 55 TABLE XIV Independent Variables Significantly Correlated with School District Tax Effort (In Descending Order of Correlation) 58 TABLE XV Regression Analysis of the Relationship of Local Tax Effort to Significant Independent Variables for All 201 School Districts 60 iv

PAGE 5

TABLE XVI Regression Analysis of the Relationship of Local Tax Effort to Significant Independent Variables for School Districts with High Assessed Valuation of Property per Pupil in Average Daily Membership (N=67) Page 62 TABLE XVII Regression Analysis of the Relationship of Local Tax Effort to Significant Independent Variables for School Districts with Low Assessed Valuation of Property per Pupil in Average Daily Membership (N=67) 63 TABLE XVIII Independent Variables Correlating with Factors at .40 or Higher 66

PAGE 6

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE RELATIONSHIPS BETWEEN SELECTED SOCIOECONCMIC VARIABLES AND LOCAL TAX EFFORT TO SUPPORT PUBLIC SCHOOLS IN VERMONT By Edward Joseph Fabian August, 1976 Chairman: Kern Alexander Major Department: Educational Administration This study sought to determine if certain socioeconomic characteristics of school districts were related to tax effort in Vermont school districts during the 1972-1973 school year. Tax effort was computed for Vermont's 201 school districts by . dividing local revenue per pupil in average daily membership by equalized grand list per pupil in average daily membership. The literature and research were reviewed to determine relevant variables. Twenty-five independent variables were selected. The data were analyzed by means of stepwise multiple regression with tax effort as the dependent variable. This statistical analysis was performed on all districts and then on districts whose equalized grand list valuation of property per pupil in average daily membership placed them in the top and bottom third of all districts in property wealth. Since it appeared that the stepwise regression equation masked many variables, a factor analysis was performed. Six factors were converted to factor scores and the scores were subjected to regression analysis. vi

PAGE 7

This study concluded that no factors are significantly correlated to tax effort which would predict the type of districts making a substantial tax effort in the support of schools in Vermont. Even though the correlations were not high, the independent socioeconomic variables identified districts with high residential property values as the ones which make a greater effort in the support of schools in all districts and in the onethird of the school districts with the greatest wealth. Likewise, it appears, but is not conclusive, that the one-third low wealth districts, districts with high property farm value, make a greater effort in the support of education. The findings also suggest that districts which have a high concentration of economically disadvantaged children make less effort in the support of schools. Even though the correlations were low and not significant, wealthy residential areas appear to consistently tax themselves proportionately more than poor rural districts. A state-local finance system permitting optional local effort which operates to the benefit of residential districts tends to provide the greatest amount of educational resources per child in such districts. Although the study does not fully explain the reasons for the wide variations in tax effort for the support of schools in Vermont, wide variations do in fact exist. Because of this wide variation, the state should intervene and provide assistance which would reduce the disparities in expenditures for education. Further studies are recommended to determine if geographic and political differences in school districts affect the level of the effort they make in the support of education in Vermont. vii

PAGE 8

CHAPTER I INTRODUCTION During the past twenty-five year period, between World War II and the beginning of the 1970' s, America's school system experienced a tremendous increase in enrollment and an expansion in the number of years students were expected to stay in school. This growth, coupled with increases in teachers' salaries, greater offerings of school services, and inflation caused elementary and secondary costs to rise to a new high each year. The total amount of school revenue from state sources in current dollars . . . Increased 77.3 percent between 1930 and 1940; the increase was 228.2 percent between 1940 and 1950; 164.2 percent between 1950 and 1960; and 173.4 percent between 1960 and 1970.1 Expenditures in education have risen forty-three percent faster than increases in the economy as a whole. The result is that an increasing burden is being placed upon traditional school revenue sources. School districts in all states, with the exception of Hawaii, utilize varying means of local taxation for the support of public schools. Local school districts in most states have legislative authority to raise revenue for the support of schools by school board action or voter approval. This practice has historically been associated with local control. It has been argued that local control increases local interest in schools. This view has raised serious questions regarding the equalization of educational opportunity. ""R. Johns, K. Alexander, and K. F. Jordan, Financing Educatio n Fiscal and Legal Alternatives . Charles E. Merrill, Columbus, Ohio, 1972, p. 20.

PAGE 9

2 Recent court decisions have been based on the fact that children in wealthy communities receive greater educational offerings than children 2 growing up in poor communities. Two conclusions can be drawn from these decisions. The first is that state financing plans rely heavily on local property tax and cause substantial disparities among individual school districts in the amount of resources available per pupil for the district's educational program. The second conclusion which may be drawn is that as a result of the financing plans, taxpayers in less wealthy districts are required to pay a higher tax rate than taxpayers in many other school districts in order to obtain for their children the same or less educational opportunities afforded children in more wealthy 3 • school districts. It is argued that economic and social conditions which exist within school districts bear a relationship to the fiscal efforts produced by the district in support of public education. It is reasonable to suspect that individuals with high incomes, and/or with a higher level of education, are more likely to support higher school tax rates than individuals with low educational levels and low incomes. It is important to note that ninety-eight percent of all public school districts in the United States rely on the local property tax for 4 financial support. Dependence on the local property tax effort as a 2 Serrano v. Priest, 96 Cal. Rptr. . 601, 487 P. 2d 1241 (1971). Rodriguez v. San Antonio Independent School District, 337 F. Supp. 280 (1971) Reversed 41 "Law Week" 4407 March 21, 1973. 3 R. Johns, K. Alexander, and D. Stollar, (Eds.) Status and Impact of Educational Finance Programs . Vol. 4, National Educational Finance Project, Gainesville, Florida, 1971. 4 National Educational Finance Project, "Future Directions for School Financing," Gainesville, Florida, 1971.

PAGE 10

3 method of obtaining future educational funds is seriously questioned in Vermont as well as throughout the nation. The above statements raise serious questions: Can we totally equalize education? How much effort are local districts willing to make in support of public schools? Will local control be eliminated? What type of foundation program should a state utilize? The answers to these questions may well require the ability to realistically measure local willingness to support public schools with taxes. Statement of the Problem The purpose of this study was to determine the relationship between selected socioeconomic variables and tax effort of school districts in Vermont. Delimitations — 4 • 1. This study is confined to school districts in the State of Vermont. Limitations 1. This study is ex post facto and possesses the weaknesses inherent in this method. 2. Causal relationships cannot be determined from this type of study. 3. Conclusions cannot be generalized beyond the districts that are included in this study nor beyond the time for which data are appropriate. Justification for the Study This study will add to our understanding of citizenry support of education. Its results should have practical value in light of the current status of financing school systems in Vermont and State Supreme Court decisions resulting from the Serrano v. Priest^ and Rodriguez v. San ^Serrano v. Priest, op. cit .. 1971.

PAGE 11

Antonio Independent School District actions. In view of the fact that many conventional arrangements and interests of school finance are being challenged in Vermont, it is timely that an investigation of this nature be conducted to determine whether state educational funds distributed on the basis of local optional effort are equitable. For the most part, the distribution of educational resources among school districts (on the basis of local optional effort) has favored wealthier districts.^ g Campbell also supports this position. He states that locally raised revenue and state aid are still allocated in most school districts to favor the already socioeconomically advantaged schools. This study should assist Vermont legislators in understanding the nature of local tax effort and it should have some practical value in future school finance decision making. No study of this nature has ever been undertaken in Vermont. Assumptions 1. Equalized grand list valuations of property is a valid measure of ability to finance education. 2. Equalized grand list valuation as computed is dependable for all districts in the state. 3. Socioeconomic data secured from governmental sources are reliable. Definition of Terms Ability . — A potential source of wealth, in this case equalized grand list valuation of property, which has the capacity to support education from the taxes based on it. ^Rodriguez v. San Antonio Independent School District, op. cit ., (1971). ^R. Johns, K. Alexander, and D. Stollar, (Eds.), op. cit . g Alan Campbell and Donna Shalala, "Resource Literature and Educational Revenue," Theory Into Practice . Ohio State University, Vol. XI, No. 2, April 1972, p. 77.

PAGE 12

Average daily membership (A.D.M. ) .--The average enrollment of resident pupils of the district attending public schools for the first 30 days of a school year in which the school was actually in session; it is the quotient obtained by dividing by 30, the aggregate membership of resident pupils in the school district during the first 30 days in which the school was actually in session. ^ Effort. — A ratio of school district revenue to some measure of wealth. In this study it will be the ratio of local revenue per pupil in A.D.M. to equalized grand list per pupil in A.D.M, Equalized grand list . — One percent of the fair market value of all taxable property in a school district as established by the tax commissioner biennially plus the taxable polls. '^ Local revenue . — Money collected at the local level primarily through the property tax for school purposes. This total represents all revenue receipts from local sources. Percent of economically deprived children . — The percent of children eligible for Title I Elementary and Secondary Education Act benefits in proportion to total children of school age in the school district. Poverty level . — A sliding scale of a family income in relation to marital status, age and family size, that is deemed inadequate or below poverty levels for United States Census purposes . State revenue . — The money obtained through state taxes and allocated to the local districts for school purposes through state aid.^^ Socioeconomic variables . — Variables which are characteristic measures of the social and economic system. These do not include internal school characteristics such as teachers' salaries, number of teachers, pupilteacher ratios, etc. Procedures A stepwise multiple regression analysis was performed between the measures of effort and selected socioeconomic variables. School districts ^Vermont Department of Education, "Vermont State Aid," Montpelier, Vermont, 1972, p. 2. ^°Ibid. , p. 2. '-'United States Bureau of the Census, "Public Use Samples of Basic Records from the 1970 Census," Washington, D. C, United States Department of Commerce, April 1972, p. 122. 12 Vermont Department of Education, op. cit ., p. 2.

PAGE 13

6 were ranked according to the equalized grand list per pupil. Districts in the top and bottom third of this wealth spectrum were placed in two groups: a high wealth group and low wealth group. Variables found to be significantly related to effort in the two groups were compared to determine if high-effort, high-wealth districts have different socioeconomic variables related to their effort than high-effort, low-wealth districts. A second step in the procedure was factor analysis. Factor analysis assisted in transformation of the independent variables into a set of factors which generally reveal important relationships which are difficult to discern among the variables in their original form. Collection of Data Data were collected from the Vermont State Department of Education, Vermont State Tax Department, the National Educational Finance Project and the United States Bureau of the Census. Analysis and Treatment of the Data The statistical techniques used in this study were stepwise multiple regression and factor analysis. In stepwise regression, described by 13 Cooley and Lohnes and the BMD, Biomedical Computer Programs packet developed at the University of California,"'"^ effort was the dependent variable and the socioeconomic factors were the independent variables. The stepwise procedure is a modification of multiple regression analysis that adds one variable at a time to the prediction equation. Variables are added or dropped in accordance with the significance of their contribution 13 W. Cooley and P. Lohnes, Multivariate Procedures for the Behavioral Sciences , John Wiley and Sons, New York, 1962. 14 W. T. Dixon, (Ed.) BMD, Biomedical Computer Programs , University of California Press, Berkeley, 1968.

PAGE 14

7 to the prediction of the criterion variable. This results in development of intermediate regression equations as well as the complete equation. '^^ In factor analysis a correlation matrix is computed for the total data cluster. Then a line of best fit is computed for the cluster. Next, a series of perpendicular axes are computed to explain the maximum amount of variation remaining in the data cluster. These perpendicular (orthogonal) axes are then rotated so a minimum number of axes explain the variance of the data. Finally, the axes can be released from the requirement of orthogonality which permits them to conform more closely with individual data clusters. The resulting axes are the factors which will be extracted from the total data cluster. "'^W. Cooley and P. Lohnes, op. cit ., p. 35. "'^^Norman Nie, C. Hadley Hull and Dale H. Bent, Statistical Package for the Social Sciences , McGraw-Hill, New York, 1970, pp. 219-223.

PAGE 15

CHAPTER II REVIEW OF LITERATURE AND RELATED RESEARCH Improvements can only come about if historical developments are presented. History has been the basis on which values are improved, new ideas presented and mistakes corrected. This chapter therefore, presents a national historical overview of public education, use of real property tax for support of schools and the move toward educational equalization. Related studies reviewed are those that concern themselves with local financial ability, socioeconomic factors related to school expenditures and variables which influence tax effort. National Historical Overview The first public supported schools were established in Massachusetts and Rhode Island during the years of 1635-45. A European influence lies behind the formation of American Public Schools. Cubberley states that: Schools with us, as with the older European countries from which early settlers came, arose largely as children of the church . . . they brought with them their European ideas as to religion and the training of children and hence, a European background lies behind all the beginnings of public education. E. Cubberley, Public Education in the United States . Houghton Mifflin Co., Riverside Press, Cambridge, Massachusetts, 1934, pp. 11-12. ^Ibid . , p. 12. 8

PAGE 16

9 It was not long before individuals realized that voluntary support of public schools was not sufficient. Hence, two laws were enacted in the State of Massachusetts in 16A2 and 1647. The law of 1642 was the first law passed by a state legislature making it compulsory that all children be taught to read. The law of 1647 established public supported schools. The significance of these two laws is presented by Cubberley: Not only was a school system ordered established, elementary for all towns and children, and secondary for the youth in larger towns but, for the first time among English speaking people, there was the assertion of the right of the state to require communities to establish and maintain schools, under a penalty of a fine if they refused to do so. It can be safely asserted that these two Massachusetts laws of 1642 and 1647 represent not only new educational ideas in the English speaking world, but that they, together with the laws of 1634 and 1638, providing for the equalized and compulsory taxation of all town charges, also represent the very foundation stones upon which our American public school systems have later been constructed. Subsequently, after that three types of educational responsibilities emerged. The first was the support of a system of common schools, Latin schools and a college which served religious and civic ends. The second was the establishment of the parochial school concept, which stood for church control of all educational efforts and resented state interference. The third type conceived of public education as being intended chiefly for orphans and children of the poor and as a charity which the state was under 4 no obligation to support. 3 Ibid . , p. 17. ^Ibld. , p. 25.

PAGE 17

10 Cubberley^ states, that because of the Revolutionary War and these three concepts of education, the Constitutional Convention did not evidently consider education to be important enough to be included in the Constitution and thus the responsibility for it was passed to the various states by the Tenth Amendment. The Tenth Amendment ratified in 1791 provided that, "powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the states respectively, or to the people." Following the Revolutionary War, education of the young grew from parental responsibility to church supported schools. Financial support in this early period was derived from lotteries, endorsements, licensing and commodities. The character of education during this early period and the lack of interest in it as described by Cubberley resulted from: The simple agricultural life of the times, the homogenity of the people, the absence of cities, the isolation and independence of the villages, the lack of full manhood suffrage in a number of states, the continuance of old English laws, the want of any economic demand for education, and the fact that no important political question calling for settlement at the polls had as yet arisen, made the need for schools and learning seem a relatively minor one. 6 As living standards changed and the country moved from an agricultural to an industrial climate, citizens in newly formed states soon realized the need for education. During the Jackson administration education was expressed for the first time as political policy.^ During 1825 to 1830 in all the Northern states, the battle for direct, local, county and state taxation for education was clearly evident, according ^Ibid . , p. 86. ^ Ibid . , p. 110. ^P. R. Mort and W. C. Reusser, Public School Finance . McGraw-Hill Book Co., Inc., New York, 1941, p. 8.

PAGE 18

11 8 to Cubberley. By 1825 it was recognized that the only safe reliance of a system of state schools lay in the general and direct taxation of all property for the support of education. Massachusetts was the first state to take the lead and lay the foundation for the state's common school fund in 1834 and it opened the first normal school in 1839. In New York an experimental free school law was in effect from 1795 to 1800. A permanent free school law dates from 1812. Also, in New York, a tax to raise the amount of money equal 9 to that granted by the states was made compulsory. For the first time direct taxation for schools was likely to be felt by the taxpayer and the fight for and against the imposition of such taxation was on in earnest. In a general way, the progress of the conflict was as follows:''"^ 1. Permission granted to communities so deserving to organize a school taxing district, and to tax for school support the property of those consenting and residing therein. 2. Taxation of all property in the taxing districts permitted. 3. State aid to such districts, at first from the income from permanent endowment funds, and later from the proceeds of a small state appropriation or a state or county tax. 4. Compulsory local taxation in the state or county grant. With the beginnings of state aid the states were in a position to enforce definite requirements in many matters. As early as 1797 Vermont had required the towns to support their schools on penalty of forfeiting g E. Cubberley, op. cit ., p. 180. 9 P. R. Mort and W. C. Reusser, op. cit ., p. 9. "'"^E. Cubberley, op. cit ., p. 180.

PAGE 19

12 their share of state aid.'^"'' The right to tax for support, and to compel local taxation, was the key to the whole state system of education. From this point on the process of evolving an adequate system of school support was different in each state. Equalization Once taxation for free public schools was established, questions of equalization of educational opportunity were raised. Paul Mort indicated that revenue spent on education should go directly to where children are located and that these funds support a wide range of educational programs , . 13 and services. Near the beginning of the twentieth century, Cubberley developed the theory of state support. He saw the state's responsibility as being 14 concerned with equalization and reward for effort. Likewise, he saw the state as having the right to establish minimum educational standards which districts would be compelled to follow. School districts would also be encouraged to extend their programs beyond the minimums.'^^ American education was expanding rapidly and this concept of financial assistance to poorer districts had great influence. Updegraff, after a study of rural districts in New York, noted that granting of special financial assistance for individual functions Ignored the economic weakness of vast sections of the states. He held that reward for effort on the basis of individual projects was too narrow In scope. He proposed ""^Ibid. , p. 188. "•^Ibid. , p. 189. 13 P. R. Mort and W. C. Reusser, op. cit ., p. 378. 14 E. Cubberley, School Funds and Their Apportionment , Teachers College Columbia University, New York, 1905, p. 16. "*^^Ibid. , p. 17.

PAGE 20

13 a sliding scale that provided increased amounts of state aid per teacher unit for each increase of one-half mill of school taxes levied ranging from three and one-half to nine mills, but he provided proportionately more state aid for a district with a low valuation per teacher unit.^^ This marked the beginning of a state support program based upon the standard of more state aid to districts with low property valuation. In 1924 this concept of equalization was advanced further through a study completed by Strayer and Haig in New York State. Two pages of this study were devoted to a theoretical conceptualization of the equalization of educational opportunity which has had a major impact on the relationship of state and local agencies in exercising the responsibility of all people toward public education. The concept, equalization of educational opportunity as described by Strayer and Haig, suggested that each state guarantee that minimal educational programs and facilities be available to every child within the states' borders and that the tax burden within the state be equal in relationship to each citizen's taxpaying ability.''"^ 18 According to Mort the period following was highly productive in the further clarification of the equalization principle and in the development of various measuring devices and formulas for implementing these principles. 16 R. Johns, K. Alexander and K. F. Jordan, Financing Education Fiscal and Legal Alternatives . Charles Merrill Publishing Co., Columbus. Ohio. 1972, p. 7. "^G. Strayer and R. Haig, "The Financing of Education in the State of New York," Report of the Educational Finance Inquiry Commission, Vol. I, Macmillan Co., New York, 1923, pp. 173-175. 18 P. R. Mort and W. C. Reusser, op. cit ., p. 384.

PAGE 21

14 Mort defined a satisfactory equalization program as follows: A satisfactory equalization program would demand that each community have as many elementary and high school classroom or teacher units, or their equivalent, as is typical for communities having the same number of children to educate. It would demand that each of these classrooms meet certain requirements as to structure and physical environment. It would demand that each of these classrooms be provided with a teacher, course of study, equipment, supervision, and auxiliary activities meeting certain minimum requirements. It would demand that some communities furnish special facilities, such as transportation. 19 He devised a foundation program using the concept of "weighting pupils" to compensate for the extra cost necessary for special programs. 20 Morrison, proposed a model of state support whereby all local school districts are abolished and the state itself became both the unit for taxation for schools and for administration of public schools. He suggested that the most equitable form of tax for the state to use for the support of schools was the income tax. Summary Programs which provide equalization of educational opportunity have been slow to develop in most states. The National Educational Finance Project classified the types of foundation programs in use in the United 21 States into four basic categories. Flat grants Flat grants may be uniform or variable. They generally do not take into consideration the wealth of the school district and they are distributed on some basic unit of need, such as pupil or teacher unit. These grants may be of the general or special purpose type. 19 R. Johns, K. Alexander, and K. F. Jordan, op. cit ., p. 11. Ibid . , p. 13 21 R. Johns, K. Alexander and K. F. Jordan, op. cit ., pp. 121-123.

PAGE 22

15 Equalization grants All equalization grants take into consideration variations in the taxpaying ability of the local school districts. However, not all equalization grants consider the variations of educational needs of the student population. Like flat grants, equalization grants may be general or special purpose. Non-equalizing matching grants Non-equalizing matching grants require local districts to match state funds on a dollar for dollar basis, or some proportion of a dollar without taking into consideration variations in the taxpaying ability of local school districts. These types of matching grants leave districts in the same relative status, and therefore, provide for little equalization. Complete state and federal support grants All support for public education is from the state or federal government. Local support is non-existent. Pupils' access to wealth is dependent upon the total wealth of the state. Revenue for the program can be obtained completely from the general funds of the state or from state taxes earmarked for this purpose. Related Studies A number of studies have shown that various socioeconomic factors affect decision making regarding school fiscal policy. The purpose of this section is to review relevant literature and research to identify socioeconomic characteristics related to local tax effort. This section will be divided into three subsections. The first subsection will identify variables related to voter attitudes toward education. Subsection two will consider socioeconomic variables correlated to school expenditures and the

PAGE 23

16 final subsection will identify those variables which predict local district effort in support of education. A listing of all variables which have been identified as being correlated with local tax effort will be presented In the sunmary. Public attitudes 22 In the Annual Gallup Poll of Public Attitudes Toward Education conducted in 1972, fifty percent of the college graduates polled favored tax increases for schools while only twenty-seven percent of the people with only elementary education approved. White collar workers were much more favorable to increased school taxes than blue collar workers, while younger people were more favorable to increased school taxes than elderly people. People in higher income brackets were more favorable to tax increases than people on restricted incomes. Communities with populations between 25,000 to 50,000 were notably more approving of school tax increases than larger or smaller communities. Factors that were not significantly related to attitude on school taxes were region of the country, sex, people with no children in school, and private school patrons. Likewise, the fifth and sixth Annual Gallup Polls of Public Attitude Toward Education conducted in 1973 and 1974, listed the lack of proper financial support as the third most important problem facing education for both years. 23 Meyers, in a study of urban communities, however, concluded that 22 George Gallup, "Fourth Annual Gallup Poll of Public Attitudes Toward Education," Phi Delta Kappan , Vol. LIX, No. 1, September, 1972, pp. 33-46. 23 Alfred V. Meyers, "The Financial Crisis in Urban Schools-Patterns of Support and Non-Support Among Organized Groups in Urban Communities," Doctoral Dissertation, Wayne State University, 1965, Dissertation Abstracts, Vol. 25, p. 5024.

PAGE 24

17 parents of parochial and private school students tend to oppose increased financial support for public schools. He also found that organized groups within the community tend to influence the adequacy of financial support for public education. In still another study conducted by Carlson^^ in three communities, he found that public and parochial school patrons had significantly different attitudes in federal aid, dual enrollment, shared facilities, the child benefit theory and on the extension of health and transportation services. 25 Patterson and Schoonhoven studied inconsistent voter behavior in two Oregon public school districts. Despite demographic differences between the districts in level of schooling completed, occupation, income, period of residence, political affiliation and religious preference, reasons for voting positions were similar in both districts. In each district one-fourth of the voters attributed their failure to vote to forgetting or to being unaware of the election. A majority in both districts felt they had a legitimate reason for not voting previously and seemed determined to utilize their vote to affect the financial outcomes. Statistically significant differences were noted in both districts between positive voting and voters with children in school and more completed years of schooling. Significant findings in one district (not supported by the second district), related positive voting to persons under 45 years of age, level of occupational skill, spouses as income producers and family income over $5,000. 24 DeVere Carlson, "Patron Attitudes Toward Selected Educational Issues in Communities with a Dual Educational System," Doctoral Dissertation, University of Pittsburgh, 1969, Dissertation Abstracts, Vol, 30, p. 3665A. 25 Wade Patterson and John Schoonhoven, "A Comparative Study of Inconsistent Voter Behavior in School Budget Elections," November, 1966 ERIC, ED-011135.

PAGE 25

18 26 Voter behavior toward financial referendums was studied by Von Hatley. He found that family income appeared to have potential for predicting voter behavior in income based school referendums, but evidenced questionable predictive utility in property tax referendums. He also found that the voting behavior of middle income groups was more difficult to predict than the high and low income groups. Variables related to voting behavior were the number of children in the family, educational attainment, length of school district residency, the concepts of high versus low support for education and broad versus narrow conceptions of the value of education. 27 In a study by Witt of San Mateo voters, regarding a combined tax and bond proposal for additional Junior College funds, it was found that voters more than 50 years of age without children under 21 years of age were less in favor, particularly of the tax proposal, while white collar and professional people showed more favorable response than blue collar workers, housewives or retired people. 28 Petersen's study investigated the relationship between parent attitudes toward school progress in a specific geographic area of a single school district and certain socioeconomic variables. Parents who had completed some high school education tended to be more approving of school programs than those who had some past high school training and those who had less than nine years of formal education. Non-white parents were found to be less approving 26 R. F. Von Hatley, "Family Income Voting Behavior and Financial Referendums: Educational Finance and Politics in Albuquerque, 1968-69," Doctoral Dissertation, University of New Mexico, 1971, Dissertation Abstracts, Vol. 31, p. 5703A. 27 Irving Witt and Frank Pearce, "A Study of Voter Reaction to a Combination Bond-Tax Election on March 26, 1968," ERIC, ED 019945. 28 Thor Petersen, "School Approval-Disapproval and Educational Enlightenment of Parents Based on Occupation, Educational Level, Age, Race, Geographic Area and Length of Residency," Doctoral Dissertation, Michigan State University, 1971, Dissertation Abstracts, Vol. 32., p. 2971A.

PAGE 26

19 of school programs than white parents. Those parents who have lived in a particular school attendance area for more than five years were less approving than those who had resided there for less than one year. In this study, occupation and age were not significantly related to school . approvals. 29 In a West Virginia study conducted by Photiadas and Zeller it was found that white collar, business, managerial and professional people favored raising taxes for the support of schools more so than did the unskilled or semi-skilled workers. A strong relationship existed between the level of education attained and the willingness to increase taxes. Income level was also positively correlated to a willingness to increase taxes in support of education. Summary The above studies indicate that the level of educational attainment, level of income, race, length in school district residency, occupations, non-public or parochial school attendance, age and voters with children in school can have some influence upon community attitudes toward public education. Variables Related to School Expenditures 30 The primary goal of a study conducted by Ellis was to analyze the relationship between local expenditures on primary and secondary education and local income. The secondary aim of the study was to evaluate the significance of education relative to other factors as a source of difference 29 J. Photiadas and F. Zeller, "Attitudes Toward State and Local Taxes in West Virginia: The Preliminary Results of a Survey," University Center for Appalachian Studies and Development, Morgantown, West Virginia, 1968. 30 J. Ellis, Jr., "A Study of the Relationship Between Local Expenditures on Education and Local Income," Doctoral Dissertation, University of Virginia, 1967, Dissertation Abstracts, Vol. 27, pp. 3177A-3178A.

PAGE 27

20 in local income. Ellis concluded that average per student county educational expenditure was significantly related to levels of county per capita income. Attainment of a high school degree, urbanization, the non-white proportion of the population and occupational structure other than agriculture were related to labor income. 31 The purpose of MacDougall's study in Virginia was to analyze income flow to educational expenditures. The flow process was described by the measures of the following economic variables: income, property taxation, state and local charges, other state and local taxes, federal grants, federal revenue, federal grants to education, state and local expenditures. The economic variables and educational expenditures were analyzed by a multiple correlation and regression routine to establish a comprehensive set of relations describing the nature of the economic structure with educational expenditures. The variables, personal income and state and local charges, had a high correlation with educational expenditures. 32 The net result of a 1972 study conducted by Lazier in Utah also indicated that high personal income has a direct relationship to the amount of school tax a district will approve for school expenditures. 33 Likewise, Miner found in his study that the level of state per capita income was the most important determinant of total per capita expenditures. Median family income in local communities was inversely related to expenditures, but the proportion of families with incomes over hi. A. MacDougall, "An Analysis of Income Flow to Educational Expenditures," Doctoral Dissertation, University of Virginia, 1964, Dissertation Abstracts, Vol. 24, p. 2738. 32 Willard E. Lazier, "School Finance: A Determinant Model for Eligibility," Doctoral Dissertation, University of Utah, Salt Lake City, 1972. 33 Jerry Miner, Social and Economic Factors in Spending for Public Educations , Syracuse University Press, Syracuse, 1963.

PAGE 28

21 $10,000 and the proportion of children in the population were positively related to expenditures. The proportion of children in private schools had a negative affect. Local expenditures were also strongly influenced by state aid formulas. 34 Harvey found that assessed property evaluations were most significant determinants of elementary educational expenditures. Residential, agricultural and commercial properties jointly were able to explain 89 percent of the variation in local expenditures while socioeconomic and voting characteristics explained 73 percent of the variation in current 35 expenditures. However, Fisher's study, which made extensive use of adjusted gross personal income per school district, demonstrated that there is justification for questioning property valuation as a valid measure of local fiscal ability. In a study which attempted to find the relationship between community demand for education and the willingness of a population to finance a given amount of education and local financial support of public schools in 36 Illinois, Metzcus found that median family income, population, percent of population non-white, proportion of the population in selected occupations correlated with local district current operating expense per pupil. He also found that the amount of property valuation per student had the highest positive correlation with local districts' current operating expenses per 34 E. L. Harvey, "Property Tax Determinants of Educational Expenditures," Doctoral Dissertation, Stanford University, 1969, Dissertation Abstracts, Vol. 30, p. 1333A. 35 Jack Fisher, "A Comparison Between Central Cities and Suburbs and Local Ability to Support Public Education," Doctoral Dissertation, University of Florida, Gainesville, 1972. 36 Richard H. Metzcus, "Community Human Resources and Local Financial Support for Public Schools," Doctoral Dissertation, University of Illinois, 1969, Dissertation Abstracts, Vol. 30, p. 102A.

PAGE 29

22 pupil and that districts with low property wealth tended to make a high tax effort in support of schools. Districts with a high proportion of Industrial and Commercial property spend more money on schools for about the same tax rate as other districts. 37 This was the conclusion reached by Clune. He analyzed the effects of Industrial and Commercial property on wealth variations among school districts in Cook County, outside the City of Chicago. Per pupil expenditures correlated positively with urbanization, industrialization, income and the amount of education in a study conducted 38 by Dye. Effort, as was defined by expenditures relative to income, correlated negatively with urbanization, industrialization and income. He concluded that states with the greatest urbanization, industrialization and wealth made less tax effort, but maintained a high per pupil expenditure. Economic variables appear to be more influential than political system characteristics in shaping educational fiscal outputs. Dye discovered, for example, that almost 70 percent of the total variation among the 50 states in per pupil expenditures was explained with median family income. 39 James and his colleagues, in an analysis of the determinants of educational spending in 107 cities found that the socioeconomic variables; income, property value, adult education and race, were more highly correlated with expenditures than the characteristics of the school system. The percent 37 William H. Clune, "Taxing and Spending for Public Schools; The Origin Descriptions and Effects of Non-School Taxes and Industrial and Commercial Property," Education and Government Division, Illinois Bureau of the Budget, Chicago, 1971. 38 Thomas Dye, "Politics, Economics, and Educational Outcomes in the States," University of Georgia, Athens, 1967. 39 Thomas James, J. Kelley and W. Garms, "Determinants of Educational Expenditures in Large Cities of the United States," School of Education, Stanford University, Palo Alto, California, 1966, pp. 95-134.

PAGE 30

23 of labor force unemployed, median family income, percent of homeowners, median years of schooling, property valuation per pupil and the percent of pupils attending private schools were also all significantly correlated to school expenditures in this study. 40 Hendrix and Alkin found that 67 percent of the variation in school expenditures could be explained by individuals in certain age groups, and 41 Farner found that the male labor force was positively related to expenditures for school districts as were housing standards. In another study conducted for the Office of Education, James, Thomas 42 and Dych found that property value was positively correlated to school expenditures in 15 sample states. Median family income was also positively related to all samples except in Nebraska. Likewise, percent of owner-occupied housing was negatively correlated to expenditures in all samples except Nebraska and Oregon. Median years of schooling was positively related to expenditures in two samples, negatively correlated to expenditures in three and not significant in ten. The percent of labor force unemployed was significant and negatively correlated to expenditures in three of the 15 samples. Percent of population non-white was significant in four of the 15 samples. Percent of farms population was negatively significant in four samples and the percent of elementary school children in 40 V. Hendrix and M. Alkin, "Population Age Distribution and Public Educational Expenditures." A paper. Educational Research Association, New York, February, 1967. 41 Frank Farner, "Economic Sociological and Demographic Characteristics of Oregon School Districts and Their Relationship to District Financial Practices," University of Oregon, Bureau of Educational Research and Services Eugene, April, 1966. ' 42„ ^ . H. James, A. Thomas and H. Dych, "Wealth Expenditures and Decision Making for Education," USOE, Research Project No. 1241, Stanford University Stanford, California, June, 1963.

PAGE 31

24 private schools was significantly related to expenditures in only two samples. Property wealth, income and degree of home ownership had a significant relationship with school expenditures. 43 Alkin conducted a study which sought to examine the relationship between religious composition of school districts and a measure of financial resources provided to the districts. He hypothesized that values implicit within the religious framework of belief, customs and practices and the apparent ethnic residues within religions provide differences in educational aspirations as reflected in expenditures for public education. Alkin studied 18 sample districts and the districts had to contain an enrollment of 300 A.D.A. or greater. He also categorized religious denominations into variables which could be examined statistically. His variables included percent of Catholic, percent Protestant Nurture, percent Protestant Borderline, percent Protestant Conversion, percent Jewish, percent Buddhist, percent Mormon, percent non-religion. Three variables were used for Protestant. This group was so large it could not be considered a single group because of loss of degrees of freedom in the multiple-regression equation used. Alkin found that public educational expenditures are related to the religious composition of communities and he concluded that studies utilizing socioeconomic characteristics as independent variables may do well to consider the effects of religious compositions. Summary Variables identified as correlated to educational expenditures are: number of school years completed, percent of population non-white living in a district, per capita income, state and federal aid, equalized assessed 43 Marvin Alkin, "Religious Correlates of School Expenditures," A paper prepared for the American Research Association, Chicago, February 11 1965, ERIC, ED 011143. ' S , eoruary xi..

PAGE 32

25 property valuation, percentage of property tax paid by residential, commercial and agricultural property owners, the percentage of pupils in a district attending non-public schools, the unemployment rate of the districts, the age of the voters and the number of children in the population as compared to the total population, white collar vs. blue collared occupations. Variables related to tax effort Effort is a measure of a district's willingness to pay for education. To measure effort, per capita personal income may be used as a measure, but a more practical, but by no means perfect measure, is to study the relationship between a school district's property value and the amount of income it raises locally. That is, measuring a district's taxable resources and the amount of money it raises against these resources. Effort to support education depends on a community's willingness to spend, based on its ability to spend. Ability to pay for education depends on the amount of wealth in the community and the amount of money coming into the community from non-local sources. Therefore, measures of local effort can be derived from: personal income level of the community, assessed valuation and true valuation, and the amount of revenue received from state, federal and private sources. Johns and Kimbrough^^ supervised four studies which attempted to determine factors associated with school district tax effort over a period 44 "How to Evaluate Your District's Financial Effort," School Management January, 1965, pp. 112-115. ^ 45 "Income v. Effort," School Management . January, 1969, pp. 62-69. 46 R. Johns and R. Kimbrough, "The Relationship of Socioeconomic Factors, Educational Leadership Patterns and Elements of Community Power mrEn21336^°''^^ ^""^""^ ^^^""^^ Policy." USOE, Project No. 2842, May, 1968.

PAGE 33

26 of time. School Districts with populations of 20,000 or more in four states were chosen. The authors concluded that through time, there is no combination of socioeconomic variables common to the four states that could explain much of the variations in effort of school districts. Independent variables were very unstable in their productive power and would not combine as the best predictors from one period of time to another. In general, however, the greater the income of the people of the districts included in the study, the greater the local effort in proportion to ability to support schools. The measures of per capita income explained more variance in local revenue receipts per pupil than all other socioeconomic variables combined. As a part of the Johns and Kimbrough study, Adams^^ analyzed socioeconomic factors associated with effort for education in Kentucky. Adams used 22 variables which he classified as socioeconomic. These variables included factors as public and private school enrollments, federal and state revenue, income of families, population, median school years completed, labor force employed and unemployed, owner-renter households and percent of population 65 years and older. Stepwise multiple regression was used to examine the relationship between the 22 variables and measures of local financial effort and elasticity of demand. He found that state revenue receipts per pupil in A.D.A., percent rural farm and median income of families were significant predictions of variability. Adams concluded that evidence cited in the study points out that socioeconomic varinhles leave a large part of local effort unexplained. 47 Perry R, Adams, "Socioeconomic Factors Associated with Patterns of School Fiscal Policy in Kentucky," Doctoral Dissertation, University of Florida, Gainesville, 1965.

PAGE 34

27 Three other studies were conducted simultaneously with the Adams' 48 Kentucky study. King in his Georgia study, found that only one of the 22 variables was a significant predictor of local financial effort in 1950. The significant factor was percent of persons 65 years and over. In 1960 the significant predictor was state revenue receipts per pupil in A.D.A. 49 Hooper studied socioeconomic factors in Florida using the same 22 variables and procedures as were used in King's and Adams' studies. Using average effort for 1950 as the dependent variable, he found that percent of families with income of $10,000 or more, percent rural farm, percent in A.D.A. in public schools K-14 to total population, age 6-19, and per capita net effective buying power were reliable predictors of effort in Florida. The fourth and final companion study was Quick's. He examined 28 districts in Illinois. Quick found that in 1950 there was significant relationship between effort and average daily attendance as a percent of total population. In 1960 effort correlated with A.D.A. as a percent of total population and state revenue receipts per pupil in A.D.A. The conclusions reached in these four studies indicate that it is impossible to generalize through time in any given state on the relationship with any particular set of economic variables to local school effort. Socio economic variables do not exclusively determine levels of local effort in school districts. 48 Charles R. King, "Socioeconomic Factors Associ«Tted with Patterns of School Fiscal Policy in Georgia," Doctoral Dissertation, University of Florida, Gainesville, 1965, 49 Harold Hooper, "Socioeconomic Factors Associated with Patterns of School Fiscal Policy in Florida," Doctoral Dissertation, University of Florida, Gainesville, 1965. W, J. Quick, "Socioeconomic Factors Associated with Patterns of School Fiscal Policy in Illinois," Doctoral Dissertation, University of Florida, Gainesville, 1965.

PAGE 35

28 In a study conducted by Gentry, ^'^ it was found that a high positive correlation existed between social climate, certain cultural conditions and local financial ability to support education. Gentry identified 13 variables clustered around six dimensions of a society. The six dimensions were identified as items of health, education, population composition, social stability, population distribution and economic distributions. Of the 13 identified variables, four were significantly correlated with local financial initiative. These were: median school years completed by persons 25 years of age and over, percent change in population, aggregate personal income per unit of educational load and percent of aggregate personal income in transfer payments. Gentry concluded that local initiative at higher ability levels results in a considerable Increase in expenditures per unit of educational load. 52 The purpose of Martin's study was to evaluate certain social and economic characteristics of the school districts in Mississippi and to relate these characteristics to the local initiative of people in the support of schools. Local initiative was defined as tax effort beyond the minimum required by law for participation in the state foundation program. Local financial effort was defined as the total revenue effort of the districts and included financial effort required by law for participation in the foundation program. ^""Gilbert Gentry, "The Relationship of Certain Cultural Factors to Initiative in the Local Support of Education in Florida," Doctoral Dissertation, University of Florida, Gainesville, 1959. 52 Charles E. Martin, "The Relationship of Social and Economic Characteristics to Local Initiative in the Financial Support of Public Schools in Mississippi," Doctoral Dissertation, University of Southern Mississippi, 1964, Dissertation Abstracts, Vol. 23, pp. 3730-3731.

PAGE 36

29 The data presented indicated a total of fifteen variables significantly correlated with the criterion variable, local initiative. The identified variables were median school years completed 25 years and over, percentage of college graduates, percentage enrolled in college, percentage with income under $3,000, percentage with income $10,000 or more, per capita expenditures in A.D.A., percentage foreign born, percentage native population, population trend, percentage in manufacturing industry, percentage of white collar workers, percentage population 65 years and over, gross level of local initiative and local effort. He concluded that three variables are the highest predictors of local initiative in the financial support of education in Mississippi. These were median income of population which accounted for twelve percent of the variance, population trends accounted for eleven percent of the variance and population with income $10,000 and over, accounted for ten percent. A statistical analysis of the interrelationships of 36 factors selected as related to taxpaying ability for schools in Arkansas was 53 conducted by Garrison. An analysis of variance on each of the variables for the four classifications revealed a significant difference at the .05 level of significance in personal property assessments, assessments per A.D.A., median educational grade level, income per A.D.A. and local revenue per schools per $1,000 income. 54 The purpose of a study by Turck was to determine the relationship 53 C. B. Garrison, "An analysis of the Interrelationships of Economic Index of Taxpaying Ability for Schools of Arkansas Counties," Doctoral Dissertation, University of Arkansas, 1965, Dissertation Abstracts, Vol. 26, p. 820. 54 M. Turck, "A Study of the Relationships Among the Factors of Financial Need, Effort and Ability in 581 High School Districts in Michigan," Doctoral Dissertation, Michigan State University, 1966, Dissertation Abstracts. Vol. 21, p. 116.

PAGE 37

30 between acceptable measures of need, effort and ability in Michigan Public High School districts. He found that a relationship exists between size of membership and taxable wealth. There is a tendency for a school district, as it increases in size of membership, to incur more effort for the support of educational programs and there appears to be no consistent relationship between the ability of a high school district and its effort. An early study (1950) conducted by Myers, of certain phases of local tax effort in relation to taxpaying ability in Florida's 67 county school districts, concluded that wealthy districts made greater effort in support of education even though receiving substantial increases in grants of state aid. He attributed this to the fact that adults in the wealthier counties had completed a greater number of years of schooling than adults in poor counties. The findings of a Wisconsin study by Peterson^^ revealed that personal income tax paid was the most adequate of the present measures of wealth available from public records and no single measure of wealth currently in use was adequate to describe fully the ability of a community to support public service including education. This study was conducted in 1963 and investigated problems of state support and local support for education. Kay^^ used variables related to categories of wealth and income, revenue sources, property tax assessments, social status and education, population ^^H. 0. Myers, "A Study of Certain Phases of Local Tax Effort in Relation to Taxpaying Ability in Florida," Masters Thesis, University of Florida, Gainesville, 1950. 56 L. Peterson, R. Rossmiller, S. North and H. Wakefield, "Economic Impact of State Support Models on Education," University of Wisconsin, Madison, 1963. ^^Harold B. Kay, "A Study of the Relationships Between Selected Socioeconomic Variables and Local Tax Effort to Support Public Schools in Kentucky," Doctoral Dissertation, University of Florida, Gainesville, 1973.

PAGE 38

31 and those related to district and student characteristics in his study. He concluded that the degree of urbanization, (that is, the degree to which there was a high proportion of business and residences — or no agricutlure — in a district) was the most important single factor in explaining the variance in local school tax effort in Kentucky. The second most important factor was the amount of wealth and income in the district. Local variables related to differences in the amount of resources raised locally appear to include median family income, proportion of families with incomes over $10,000 per year, proportion of students in secondary grades the education level of the parents and the extent of home ownership as 58 evidenced in a study conducted by Campbell. Summary The studies and literature suggest the following variables related to tax effort: percent of families with incomes of $10,000 or more, per capita net effective buying income, percent rural farm, percent in A.D.A. in public schools, percent of students in non-public schools, number of school age children per family, population trends, state revenue per pupil, local revenue per pupil, federal revenue per pupil, income per A.D.A. , cost of living, occupations, white collar vs. blue collar, percent population non-white, assessed valuation of district property, educational attainment, proportion of revenue derived from homeowners, proportion of revenue derived from commercial enterprise, proportion of students in secondary schools, length of school district residency, median family income, percentage of 58 Alan Campbell, "The Socioeconomic, Political and Fiscal Environment of Educational Policy Making in Large Cities," Michael Kirst, (Ed), The Politics of Education , McCutchan Publishing Company, Berkeley, California, 1970.

PAGE 39

32 income under $3000, percentage native population, percentage foreign born, percentage of college graduates, percentage of population 65 years and older and percentage of growth in school enrollments. Summary School tax effort is influenced by public attitudes toward education. Certain socioeconomic factors also appear to be related to school expenditures and may be influential in determining school district tax effort. It is apparent, from the literature reviewed, that less wealthy districts often made a greater tax effort than the wealthier districts to provide acceptable 59 educational programs. Income and wealth are repeatedly mentioned as being significantly related to effort and school expenditures.^^ The following socioeconomic factors have been identified as being related to public attitude, school expenditures and local tax effort: Variables related to wealth and income: equalized assessed valuation of property, percent of economically deprived children, percent of labor force unemployed, and income variables. Variables related to revenue sources: percentage of revenue derived from federal, state and local sources. Variables related to property tax assessments: percent of tax derived from residential, commercial and farm property. Variables related to social status and education: educational attainment, religious composition of the school district, and makeup of labor force. Variables related to population: number of children in the district, number of families with school age children, amount of time citizens have lived within the community, and ethnic makeup of the school district. 59 R. Johns, K. Alexander, and D. Stollar, (Eds.) Status and Impact of Education al Finance Programs . Vol. 4, National Educational Finance Project, Gainesville, Florida, 1971, p. 111. Adams, op. cit .; Dye, op. cit .; Gallup, op. clt .; Hendrix and Alkin, o£^^_cit. ; Johns and Kimbrough, op. cit .; MacDougall, op. cit .; Ketzcus, 0£j_cit. ; Petersen, op. cit .; Photiadas and Zeller, op. cit .

PAGE 40

CHAPTER III STATISTICAL PROCEDURES AND PRESENTATION OF THE DATA The research and literature were examined to determine what socioeconomic variables are related to school district tax effort. This chapter will list the variables that were selected from the available data, identify the sources of those data, describe the statistical treatment of the data and will report the findings of the statistical analysis. A stepwise multiple regression analysis was performed between the measures of effort and selected socioeconomic variables; school districts were ranked according to the equalized grand list per pupil. School districts in the top and bottom third of this wealth spectrum were placed in two groups: a high wealth group and a low wealth group. Variables found to be significantly related to effort in the two groups were compared to determine if high wealth districts had different socioeconomic variables related to tax effort than the low wealth districts. A factor analysis was also performed. This was done to group the variables possessing correlation commonalities and therefore reduce the original set of variables into a smaller number which assisted in understanding the complex patterns of inter-relationships between the variables. Tax effort, as used in this study, is the ratio of local revenue per pupil in average daily membership to the equalized grand list of property per pupil in average daily membership. 33

PAGE 41

34 The Variables The following variables were chosen as a result of a search of the related literature and research and the availability of data: school district tax effort (dependent variable) average daily membership X2 total school expenditures X3 total school current expenditures X^ total listed value of real estate X3 percent of population 18-65 years of age Xg percent of population 65 and over Xy percent of population 3-34 years of age in parochial schools Xg percent of population in school ages 5-19 Xg percent of population 25 years old and over completing high school Xj^Q percent of population 18-65 years of age in the labor force ^11 percent of population with family income $1000 to $5999 X]^2 percent of population with family income $6000 to $9999 X23 percent of population with family income $10,000 to $24,999 percent of population with family income $25,000 and over ^15 percent of population with family income derived by wage and salary ^16 percent of population with family income derived by non-farm ^17 percent of population with family income derived by farms ^18 percent of family income derived from s
PAGE 42

35 percent of residential property value within each school district percent of commercial property value within each school district X^^ percent of industrial property value within each school district X^^ percent of farm property value within each school district Source of Data Data for school district tax effort and for variables X^, X^, X^, and ^21 obtained from the Vermont State Department of Education.^ Data for variable X^ and variables X^^ through were obtained from the 2 Vermont State Tax Department. All remaining variables for municipalities and counties were available through the 1970 census tract. These data were obtained through the fourth 3 count census tape. For effort, average daily membership, equalized grand list and current expenditures were used which are listed in the Vermont State Department of Education publication, "State Aid to Education 1973" (Columns 2, 3, 7). Variable X21 was obtained from the Title I, E.S.E.A. office. Division of Federal Programs, Vermont State Department of Education from the list entitled, "Percentage of Economically Deprived Children, 1973". ' *^ 2 Vermont State Tax Department, "1972 Real Estate and Personal Property Taxes Paid by Town and Category," Montpelier, Vt. 3 Incorporated school districts, unified school districts and union school districts whose boundaries are not coterminous with municipal boundaries and school districts for which data was unavailable were not included. United States Bureau of the Census, 1970 "Census of Population: General, Social and Economic Characteristics of Vermont," Washington, D. C. United States Department of Commerce. It was necessary to use the 1970 fourth count census tape because eighteen of the two hundred and forty eight municipalities had population in excess of 2500 and were the only ones listed in the United States Bureau of the Census Report, "1970 General and Social Characteristics of Vermont." Washington, D. C. , United States Department of Commerce.

PAGE 43

36 With the exception of unified (K-12, 1-12), and union districts (1-6, 1-8, 5-12, 7-12, 9-12) and some incorporated school districts, all municipal and school district boundaries are coterminous. Consequently, districts identified through the assistance of staff members from the State Department of Education consisted of six incorporated, four unified and twenty-eight union districts which were eliminated from the study due to the inability of matching their boundaries with census boundaries. Another twelve districts were eliminated because data was not available. Consequently, two hundred and one districts were included in the study. (Appendix A) Statistical Procedure After variables related to school district tax effort were identified in the review of the literature, they were subjected to the following statistical procedures. First, a correlation matrix of Pearson Correlation Coefficients was computed for all twenty-five variables. This demonstrated the statistical inter-relationships of all the data. Next, a stepwise multiple regression analysis was used to determine the relationships among the independent variables as they related to effort In this analysis the independent variable having the highest simple correla tion with the dependent variable first entered the equation. In the second and each subsequent step, the independent variable having the highest partial correlation with tax effort was entered; thus, at each step, the variable being brought into the computation was the one which made the greatest reduction in error in the analysis of variance based on the sum of squares deviation.

PAGE 44

37 4 All stepwise multiple regression procedures utilized the SPSS program. This program brought new variables into the equation based on the normalized regression-coefficient value (3) the variable would have it brought into the equation. The significance of 3 is measured by the F statistic (the ratio of two variances to each other the larger the ratio, the smaller is the probability that the differences between the variances was caused by chance.) If the F level is below .01, the variable is not admitted to the equation. A second factor utilized by the SPSS program prior to admitting a new variable to the equation is its tolerance. (The degree to which it covaries with the preceding variable.) If the tolerance is small the covariance approaches unity (in this case .001) then the variable is nearly a linear combination of variables already in the equation, so it will be excluded. The stepwise regression was performed on all 201 school districts in Vermont, on the one-third with the highest assessed valuation of property per pupil, and upon the one-third with the lowest assessed valuation of property per pupil. It was suspected that some of the independent variables were masking others, if two independent variables were significantly related to each other and to the dependent variable, the one with the highest correlation with tax effort would be shown as accounting for most of its variance. The second independent variable might show only an insignificant increase in the explanation of variance, even though it might itself if not masked by the stronger variable be able to account for a substantial amount of the variance of the dependent variable. When the independent variables are substantially uncorrelated, they tend to measure different aspects of the criterion variable and so make a maximum 4 Norman Nie, C. Hadlai Hull and Dale H. Bent, Statistical Pa ckage for the Social Sciences . McGraw-Hill, New York, 1970. ~

PAGE 45

38 . contribution to the explanation of variance. Using the statistical method of factor analysis, clusters of variables could be identified and analyzed. The first step in factor analysis is to compute a correlation matrix. Then a line of best fit is computed for the whole data cluster. This axis seeks to explain as much of the variance in the data as possible. Another axis at right angles to the first is computed to explain as much of the remaining variance in the data as possible. This process is continued until the remaining amount of unexplained variance in the data is considered insignificant. This process is called principal components analysis and it results in the extraction of the underlying factors from the data. The next step is rotation of the axes. The axes are rotated while they remain at right angles (orthogonal) to each other so that a maximum amount of variance in the data can be explained with the fewest axes. The axes remain uncorrelated with each other. Rotation is necessary because although the first axis represented the "line of best fit" for the total data cluster, the succeeding axes since they must remain at 90° to each other do not necessarily fit well with subclusters of data. The rotation brings the whole framework of axes into better alignment with the total data cluster. The last step used here was an oblique rotation of axes. This removed the requirement of orthogonality (axes being at 90° to each other) and allowed axes to be fitted more effectively to individual clusters of data. These axes can become correlated and the cosine of the angle between these factor axes represents the amount of correlation between the factors.^ ^Ibid., pp. 219-223.

PAGE 46

The program produced a factor analysis showing the loadings of each variable on each factor and a correlation matrix of the factors. These factors were identified by noting which variables loaded highly on them. Factor scores were computed as if the six resultant factors were six individual variables. These factors were then subjected to stepwise regression with tax effort as the dependent variable. This was done for all school districts; for the one-third highest wealth districts and for the one-third lowest wealth districts. Correlation matrices of the factors were also computed for the high wealth and low wealth districts. Presentation of the Data Correlation between Variables A correlation matrix of Pearson Correlation Coefficients was first computed (Appendix B) . ^ Zero order correlations between the dependent variable and the independent variables and their level of significance as determined by their F ratio are presented In Table I. Variable (peJ^cent of residential property value in the school district) correlated at .44 with the dependent variable effort at the .001 level of significance. Variables X^^ (Average dally membership), (Total school expenditures), X^ (Total school current expenditures), X^^^ (percent of population with family Income between $10,000 $24,999) were positively correlated with district tax effort at the .001 level of significance. Variable X^^ (percent of economically deprived children), however, was negatively correlated at the .001 level of significance with district tax effort. ^Appendix B includes Pearson correlation coefficient for the total number of districts.

PAGE 47

40 TABLE I Zero Order Correlations Between the Dependent Variable and the Independent Variables Variable R Significance Level .4400 .001 i. .2992 .001 Xi .2957 .001 Xo i .2881 .001 X 0 1 -.2315 .001 Xq o .1990 .002 X 1 o 1 J .1586 .012 X 0 c Zj .1517 .016 X7 .1517 .016 Xq .1472 .019 .1382 .025 X/. .1268 .036 X T Q .1107 .059 •^11 0788 ^5 .0759 .142 ^10 .0747 .146 ^24 .0712 .157 Xl7 -.0629 .188 ^12 .0617 .192 ^16 .0304 .334 ^23 .0165 .408 X20 -.0126 .429 ^6 .0063 .465 Xi4 .0032 .482 ^18 -.0015 .491

PAGE 48

41 TABLE II Relationship of Tax Effort to 24 Socioeconomic Variables in 201 School Districts Multiple Regression Analysis Variables Multiple R r2 2 R Change Simple R 0.A3995 0.19356 0.19356 0 43995 Xo c 0.49859 0.24859 0.05503 0 15167 Xfi o 0.53678 0.28813 0.03954 0 lQft7S Xo L 0.56013 0.31374 0.02561 0 2QQ17 XoA 0.57754 0.33356 0.01981 0 07122 Xon 0.59367 0. 35244 0.01888 -0 01265 Xl6 0.60075 0. 36091 0.00847 0 030*^6 Xl Q 0.60740 0. 36893 0.00803 0 11067 0.61376 0.37671 0-00777 0 2Q';A7 0.61779 0.38166 0 00495 0.62117 0 3858S V* J\J Jt U. U/ j{5o Xl5 0.63215 0.39961 0.01376 0.13812 X21 0.63368 0.40155 0.00194 -0.23151 H 0.63487 0.40306 0.00151 0.00629 0.63611 0. 40463 0.00157 0.07434 X9 0.63702 0. 40580 0.00116 0.14717 X3 0.63779 0.40678 0.00098 0.28812 ^18 . 0.63850 0.40768 0.00090 -0.00153 0.63914 0.40850 0. 00082 -0.06289 X12 0.63959 0.40908 0.00057 0.06165 0.63980 0.40935 0.00027 0.12679 0.64001 0.40961 0.00026 0.15168 0.64017 0.40982 0.00021 -0.07888 Xi3 0. 64029 0.40998 0.00015 0. 15856

PAGE 49

42 Variable Xg (percent population In school, ages 5-19) was positively correlated with tax effort at the .002 level of significance. Thus six of the twenty-five variables were correlated with district tax effort at the .05 level of significance or higher. Relationship of Tax Effort to Independent Variables A stepwise multiple regression of tax effort with the twenty-five independent variables for the 201 school districts was computed (Table II). This program provided the multiple correlation coefficient (Multiple R) which is the correlation between a dependent variable and the weighted sum of the independent variables. 2 The multiple regression program also provided the R statistic. This does not imply causation, but merely demonstrates the degree of covariation. Of the twenty-five variables, twenty-four had a combined Multiple R of .64029 explaining .40998 of the variance in tax effort. It will be noted that variable X^^ (percent of commercial property value) dropped out of the regression equation because apparently the tolerance was too small. Variable had a correlation of .18253 with X^^^ (percent of population with family income $10,000 to $24,999). However, X^^ correlated with tax effort at .01648 and X^^^ at .15856. Thus, X^^^ entered the equation first and since both variables covaried so closely, although not significantly, X23 did not add sufficiently to the explanation of variance to be retained in the stepwise equation. Variable X^^ (percent of residential property value) explained 19 percent of the variance in tax effort and was positively correlated to it. The next strongest variable was X23 (percent property value farms) which explained 5 percent of the variance In tax effort. Variable X 8 (percent of population in school, ages 5-19) explained 4 percent, and X

PAGE 50

43 . (total school expenditures) explained 3 percent of the variance in tax effort. Two other variables each explained 2 percent of the variance, (percent property value industrial) and X^q (percent population family income other). These six variables cumulatively explained 35 percent of the variance in district tax effort. The reoaining 18 variables together explained an additional 5 percent of the variance in tax effort. The stepwise multiple regression equation was computed for the 67 school districts having the highest equalized assessed valuation of property per child the highest third using the twenty-five independent variables and district tax effort as the dependent variable (Table III). Sixteen of the twenty-four variables were dropped from the regression equation because the tolerance level of these variables was small (Table IV). The eight variables in the regression equation had a combined multiple R of .38573 and explained .14879 of the variance in tax effort. Variable (percent residential property value) explained the highest variance, 4 percent in tax effort, while X^^ (percent population family income non-farm) explained another 3 percent. The remaining variables explained 2 percent or less: X^^ (average daily membership) negative 2 percent, and Xg (percent population in school, ages 5-19) 2 percent. X2^ (percent economically deprived children), (percent population 18-65 years of age), X^ (percent population 25 and over completing high school and X2^ (percent of property value Industrial) all explained 1 percent or less of the variance in local tax effort. The stepwise multiple regression was computed for the 67 school districts having the lowest equalised assessed valuation of property per child the lowest third of all districts using the twenty-four independent

PAGE 51

44 TABLE III Relationship of Tax Effort to 24 Socioeconomic Variables in 67 High Wealth Districts Multiple Regression Analysis Variable Multiple R r2 2 R Change Simple R ^22 0.21617 0.04673 0.04673 0.21617 ^16 0.28049 0.07867 0.03194 0.16361 ^1 0.30851 0.09518 0.01651 -0.05390 ^21 0.32067 0.10283 0.00765 -0.13198 h 0.33669 0.11336 0.01053 -0.06090 h 0.36098 0.13030 0.01694 0.00475 0.37442 0.14019 0.00989 0.09880 0.38573 0.14879 0.00860 -0.08453

PAGE 52

45 TABLE IV Variables Not In the Equation of 67 High Wealth Districts Variable Beta In* Partial* Tolerance* F X2 0.01495 0.00404 0.06222 0.001 X3 0.11545 0.02197 0.03083 0.023 X4 -0.00031 -0.00013 0.14657 0.000 ^6 0.08172 0.06821 0.59299 0.271 ^7 0.00749 0.00711 0.76651 0.003 ^10 0.02718 0.01455 0.24391 0.012 ^11 -0.02511 -0.02347 0.74384 0.032 ^12 -0.03151 -0.02566 0.56471 0.038 ^13 -0.06915 -0.04835 0.41617 0.136 ^15 -0.08411 -0.05585 0.37530 0.181 ^17 0.05709 0.05384 0.75698 0.169 ^18 0.08629 0.08155 0. 76018 0.388 ^19 0.08804 0.08708 0.83271 0.443 ^20 0.09452 0.08817 0.73908 0.454 ^23 0.07741 0.05210 0.38551 0.158 X25 0.01598 0.01476 0.72593 0.013 *Beta In . — The normalized regression coefficient that the Independent variable would have if it were brought Into the equation on the next step. The significance of Beta is measured by the F statistic. If F is small, there is little reason to add the Independent variable to the prediction equation. *Partial . — The partial correlation which represents the correlation computed for a partial group selected on the basis of the one or more other variables that are held constant in the selection process. *Tolerance .— The pivotal element which brings the variable into the equation. A small tolerance indicates that the^varlable is mearly a lineal combinaion of variable already in the equation. Norman Nie, C. Hadlai Hull and Dale H. Bent, op. clt . , pp. 179-181.

PAGE 53

46 variables and district tax effort as the dependent variable. Like the top third, eight variables remained in the regression equation (Table V). Sixteen variables dropped out of the equation (Table VI) . The eight variables had a combined multiple R of .64822 which explained .42019 of the variance in school district tax effort. Variable (percent property value farms) explained 13 percent of the variance in tax effort. Variables X^^ (average daily membership) and (total listed value of real estate) explained 10 percent, while X^^ (percent population family income wage and salary) explained 4 percent of the variance in local tax effort. These variables, plus the combination of the remaining four, X^^ (percent family income farm), X2Q (percent family income other), X^^ (percent population family income nonfarm), and X22 (percent property value connercial) explained 42 percent of the variance in local tax effort. Relationship of Independent Variables to Factors As was noted previously, there appeared to be a high degree of inter-relationship between independent variables. It further appeared that these related variables might form several clusters with each cluster representing an underlying reality or factor. The twenty-five independent variables were subjected to factor analysis. The factor analysis extracted six factors from the independent variables. The columns of figures under each factor show the correlation between a specific variable and that factor (Table VII). Factors are identified by noting the configuration of correlations between it and the independent variables. In factor 1 variables X^^ (average daily membership) and X^ (total school current expenditures) had positive correlations of .99 with the

PAGE 54

47 TABLE V Relationship of Tax Effort to 24 Socioeconomic Variables in 67 Low Wealth Districts Multiple Regression Analysis Variable Multiple R 2 R Change Simple R ^25 0.36678 0.13453 0.13453 0.36678 ^1 0.48572 0.23593 0. 10140 0.32449 ^4 0.57701 0.33294 0.09701 0.14594 ^5 0.60963 0.37165 0.03871 0.25277 ^17 0.61991 0.38428 0.01264 0.31920 ^20 0.63275 0.40037 0.01609 0.10095 ^16 0.64080 0.41062 0.01025 0.17363 ^23 0.64822 0.42019 0.00957 0.21962

PAGE 55

A8 TABLE VI Variables Not in Equation of 67 Low Wealth Districts Variable Beta In Partial Tolerance F X2 0.3174 0.09778 0.05596 0.541 ^3 0. AOOoJ 0. 10576 A A A A A 1 0. 04041 A £ 0 0 0. 633 ^5 0. 01768 0. 01329 0.32764 0. 010 ^6 -0. 07841 -0.08731 0. 71890 0.430 V ^7 O.O47O0 0. 06072 0. 96434 0.207 ^8 -0.13252 -0.13131 0. 56924 0.983 ^9 0.00727 0.00631 0.43665 0.002 ^10 -0.17970 . -0.11065 0.21983 0.694 ^11 0.03394 0.03180 0.50900 0.057 ^12 -0.07817 -0.08395 0.66868 0.397 ^13 0.01250 0.01110 0.45710 0.007 ^14 -0.04703 -0.05895 0.91077 0.195 ^18 -0.06390 -0.06971 0.69011 0.273 ^19 -0.02798 -0.03280 0. 79684 0.060 ^21 -0.02277 -0.02634 0.77550 0.039 ^24 0.02094 0.02424 0. 77697 0.033

PAGE 56

49 TABLE VII Rotated Factor Matrix Derived from the Independent Variables FACTORS 1 2 3 4 5 6 Y Q Q /. "3 n U. U/iOZ A A/ AT C -0. 04975 A AAAAO 0.00008 0. 00352 V ^2 n 0 Q n "7 /. A AO AO / 0.02984 A A£ 10^ -0. 06136 0.02066 -0. 00612 Y X3 0.02207 -0.04694 0.02479 -0.00483 V ^4 r\ 0 0 c c 1 0.02994 -0.03745 -0. 11102 0.10636 -0.07063 Y ^5 U. Uojo J A 0 /. AA C 0. 84005 0.00089 0.06526 0.19235 0.09909 Y ^6 u. uiuyo A An AA 0. 03109 A 0 T T 0 1 0. 87783 -0.05435 0.05111 0.06629 Y ^7 U. Jj/'f / 0. 0o9o2 A AO C •? / -0.03574 -0.05640 -0. 13239 -0.02049 Y ^8 U. UU0J«* A A a 0 C £ U. 49o5o A AO IOC -0.02125 0. 22924 0.02727 0.06155 Y Xg A 1 ^ 0 0 /. A CI OflO 0. 51293 A 0 i 0 er e 0. 34855 /\ A/\ top -0.00185 0.17524 -0.03579 Y ^10 A Q 0 7 0 7 0. o2/37 A AC A / 0 0.05943 -0.09335 0. 34169 0.05060 Y — U. ZDZ^Z A / C T 0 0 0.45723 A / A p 0 0 0.40532 -0.51824 -0.10982 Y ^12 A 1 COOT A A 1 /. 1 A U. 41410 0.06816 -0.08074 0.05254 0.85818 Y n 9/, QflQ A CI 0 /. A A 1 /. A "7 0 0. 14073 ^0.30215 0.41821 -0.37577 ^14 n AQQOO U. UooJ/ A 1 C1 AC 0. 15105 A ^ 1 A 0 /\ -0. 11030 0.15161 0.32553 -0.00580 ^15 -0.01422 0.87504 0.22917 -0.17877 0. 09901 0.12725 ^16 -0.10777 0.18201 0.16317 0.01917 0.45974 0.01769 ^17 -0.22687 0.13449 0.02824 0.86644 0.08473 -0.09057 ^18 -0.12138 0.12970 0.87327 0.00349 0.01723 -0.00327 X-L9 0.01190 0.18412 0.23427 0.01575 -0.01352 0.19930 ^20 0.07470 0.40622 0.41567 -0.00406 0.46104 -0.04888 ^21 -0. 23488 0.03370 0. 13444 0.19345 -0.08981 -0.05628 0.36991 0.12460 0.07571 -0.29892 0.11628 -0.04709 0.48228 -0.09165 -o.iom -0.26292 0.07998 -0.08292 ^24 0.02761 0.19408 0.03727 -0.03761 -0.07518 0.07493 ^25 -0.16679 -0.17930 -0.05953 0.61694 0.10115 0.03585

PAGE 57

50 factor. Variable X2 (total school expenditures) had a positive correlation of .98 and variable (total listed value of real estate) correlated at .82. Variable X^^ (percent property value comnerclal) had a positive correlation of .48 while variables (percent residential property value) and (percent population 3-34 in parochial schools, K-12) had positive correlations of .36 and .33, respectively. In factor 2 variable X^^^ (percent population family income wage and salary) had a positive correlation of .87. Variable X^ (percent of population 18-65 years of age) and variable X^^^ (percent of population 18-64 in labor force) correlated with factor 2 positively at .84 and .82. A positive correlation of .51 with the factor was indicative of variables X^ (percent of population completing high school) and Xj^^ (percent of population family income $10,000 to $24,999). Three variables X(percent o of population in school age 5-19), X^^^ (percent of population family income $6000-$9999) and X2Q (percent of family income other) had a positive correlation of .40 or above. In factor 3 variable X^ (percent population 65 and over) and variable ^18 (P^^'^®'^*^ population family income social security and retirement) had a positive correlation of .87 with the factor. Variable X^^^^ (percent population family income $1000-$5999) and variable X^^ (percent population family income other) had positive correlations with factor 3 at .45 and .41, respectively. Variable X^^^ (percent family Income farm) had a positive correlation of .86 with factor 4, while variable X2^ (percent property value farm) had a correlation of .61 with the factor. One other variable, variable X^^^^ (percent population family income $1000-$5999) correlated with the factor at .40.

PAGE 58

51 In factor 5, two variables had a positive correlation of .46 and .45. Those were variables X^^ (percent population family income other) and X^^^ (percent population family income non-f arm) . Variable X^^^ correlated at .41 Variable X^^^^ (percent population family income $1000-$5999) had a negative correlation of .51 with the factor. In factor 6 only one variable had a significant positive correlation with the factor and that was variable (percent of population family income $6000-$9999) . A factor correlation matrix was computed and the factors did not correlate significantly (Table VIII). The highest correlation was between factor 2 and factor 5 at .07 at a significant level of .13. Relationship of Factors to Tax Effort Factor scores were next computed for all 201 school districts and a stepwise multiple regression between each factor and school district tax effort was performed. The factors extracted from the 25 independent variables were treated as six independent variables and placed in the regression equation with tax effort as the dependent variable (Table IX). Five factors were included in the multiple regression. The tolerance level of factor 3 was insufficient for computation, therefore that factor was eliminated from the equation. The five factors within the equation had a combined multiple R of .2923. Likewise the R^ explained .0854 of the variance in tax effort. Factor 1 explained 6 percent of the variance in tax effort while factor 2 accounted for 2 percent. The factor scores for the 67 high property wealth school districts were then subjected to a stepwise multiple regression with tax effort as the dependent variable (Table X). Factors 2 and 4 were eliminated from the equation because the tolerance level was insufficient. The remaining

PAGE 59

52 TABLE VIII Factor Correlation Matrix for 201 School Districts FACTORS Factors 1 2 1 4 5 6_ 1 1.000 0.0028 0.0039 -0.0044 -0.0021 0.0052 2 1.000 0.0366 0.0026 0.0768 0.0290 3 1.000 0.0016 -0.0265 0.0035 4 . ^ 1.000 -0.0457 -0.0196 5 1.000 -0.0502 6 1.000 TABLE IX Relationship of Five Factors to Tax Effort in 201 School Districts Multiple Regression Analysis Factors Multiple R R R Change Simple R 1 0.26440 0.06991 0.06991 0.26440 2 0.28610 0.08185 0.01195 0.11003 6 0.29150 0.08497 0.00312 0.06034 5 0.29224 0.08540 0.00043 0.02565 ^ 0.29233 0.08545 0.00005 0.00413

PAGE 60

53 4 factors, factors 5, 3, 1 and 6 had a combined multiple R of .0771. The 2 R explained only .0071 of the variance in tax effort. Factor 5 explained 4 percent of the variance in tax effort. A factor correlation matrix was computed (Table XI) for the factor scores in the 67 high wealth districts. Factors 1 and 6 had a negative correlation of -.2813 significant at the .01 level. Factor 2, likewise, had a negative correlation of -.2738 with factor 3 at a significant level of .01. Factor 1 had a negative correlation of -.1828 significant at the .06 level with factor 4 and a positive correlation of .1626 with factor 5 at a significant level of .09. The factor scores for the 67 low property wealth districts were subjected to a multiple regression stepwise procedure using tax effort as the dependent variable (Table XII). Factor 3 again was left out of the equation because of a low tolerance level so was not included in the computation. The five remaining variables 2 had a combined R of .4981. The R explained .2481 of the variance in tax effort. In this analysis factor 4 accounted for approximately 10 percent of the variance in tax effort, while factor 2 accounted for 6 percent and factor 1, 5 percent of the variance.

PAGE 61

54 TABLE X Relationship of Four Factors to Tax Effort in 67 High Wealth Districts Factors Multiple R 2 R Change Simple R 5 0.21790 0.04748 0.04748 0.21790 :3 .; . 0.24423 0.05965 0.01217 0.07495 1 0.26461 0.07002 0.01037 -0.06118 6 0.27777 0.07716 0.00714 -0.06098 TABLE XI Factor Correlation Matrix for 67 High Wealth Districts FACTORS Factors 1 2 3 4 . . 5 . ,. 6 1 1.0000 0.0917 0.0088 -0. 1828 0.1626 -0.2813 2 1.0000 -0.2738 -0.0987 0.0369 0.1270 3 1.0000 -0. 0836 -0.1560 0.1342 4 1.0000 0.0135 0.0969 5 1.0000 -0.1002 6 1.0000

PAGE 62

55 TABLE XII Relationship of Five Factors to Tax Effort In 67 Low Wealth Districts Factors Multiple R r2 2 R Change Simple R h 0.32414 0.10507 0.10507 0.32414 2 0.40287 0.16231 0.05724 0.29084 1 0.46864 0.21962 0.05732 0.12984 5 0.49280 0.24285 0.02323 -0.08486 . 6 0.49818 0.24819 0.00534 -0.08752 A factor correlation matrix was computed (Table XIII) for the factor scores in the 67 low wealth districts. Factor 1 had a negative correlation of -.4053 that was significant at the .001 level with factor 2. Factor 2 had a positive correlation with factor 3 of .2445 at a significant level of .02. . TABLE XIII Factor Correlation Matrix for 67 Low Wealth Districts FACTORS Factors 1 2 3 4 5 6 1 1.000 -0.4053 -0.1447 0.0386 0.1708 -0.1042 2 1.0000 0.2445 0.1699 0,1625 0.0683 3 1.0000 0.1818 -0.0440 -0.0812 4 1.0000 -0.1875 0.0094 5 1.0000 0.0738 6 1.0000

PAGE 63

CHAPTER IV FINDINGS AND ANALYSIS n This study sought to determine if certain socioeconomic characteristics of school districts were related to tax effort in Vermont school districts during the 1972-1973 school year. Tax effort for each school district was computed by dividing local revenue per pupil in average daily membership by the equalized grand list of property valuation per pupil in average daily membership. The literature and research were reviewed to determine what variables might be correlated to tax effort. Twenty-five independent variables were selected for the study. Data were collected from the Vermont State Department of Education, the 1970 United States Bureau of the Census, the Vermont State Tax Department and the National Educational Finance Project. The data were analyzed by means of stepwise multiple regression. This statistical analysis was first performed on all districts and then on districts whose equalized grand list valuation of property per pupil in average daily membership placed them in the top and bottom third of all districts In property wealth. Data were then subjected to factor analysis. The resultant six factors were identified and factor scores computed. The factors were then subjected to a stepwise multiple regression with tax effort as the dependent variable. This statistical analysis was performed on all 201 districts and on the 67 school districts whose equalized grand list of property valuation per pupil In average daily 56

PAGE 64

57 membership placed them in the top one-third of all districts in property wealth and on the 67 school districts whose equalized grand list valuation of property placed them in the lowest one-third of all districts in property wealth. This section will summarize and analyze the findings. Initially, the simple correlations of independent variables with tax effort will be discussed. Next, the results of the regression equation using the independent variables will be presented. The results of the factor analysis will likewise be presented and finally the regression equations using factor scores will be analyzed. The independent variables with significant simple correlations with tax effort are shown in Table XIV. The variable with the highest simple correlation with effort was one relating to the main source of property taxes residential property. Districts deriving much of their local taxes from residences (X22) tended to make high effort. This could be expected since the majority of school districts are rural and most property is residential or agricultural in nature in Vermont. School districts with high expenditures (variables X2 and X^) tended to make high effort as did districts with high average daily membership (X^) . In school districts characterized by a high percentage of economically deprived children (X2-^) there tended to be low effort (-.23). Low effort in these districts might be related to poverty levels, the availability of alternative funding sources and the low educational aspiration levels of individuals at poverty levels. Districts which have a high percentage of population in school, ages 5-19 (Xg) make a high effort as do districts characterized by a high

PAGE 65

58 TABLE XIV Independent Variables Significantly Correlated with School District Tax Effort (In Descending Order of Correlation) Variables percent of residential property value X2 total school expenditures X-j^ average daily membership X^ total school current expenditures ^21 percentage of economically deprived children Xg percentage of population in school age 5-19 Xj^3 percentage of family income $10,OOO-$24,999 X25 percent property value farms X^ percent population 3-34 in parochial school K-12 Xg percent population 25-over completing high school ^15 percent population family income wage and salary X/ total listed value real estate .4400* .2992* .2957* .2881* -.2315* .1990** .1586** .1517** .1517** .1472** .1382** .1268** * ** significant at .01 level significant at .05 level

PAGE 66

59 percentage of families with incomes between $10,000 to $24,999. It is interesting to note that a simple R of .1517, significant at the .05 level, was characterized for variable (percent property value farms) and variable (percent population 3-34 in parochial schools K-12) . This is probably due to the fact that most of the remaining parochial schools in Vermont are located in the wealthier urban areas, where income levels are somewhat higher and income (^^^^ is positively correlated with tax effort. Districts which have a high percentage of population 25 and over completing high school (Xg) and districts which have a high percent of population deriving family income from wage and salary (Xj^^) make substantially higher effort in supporting their schools. Finally, districts with a higher total listed value of real estate also made high effort. Residential property tax source, income and education all had higher correlation with effort than total value of real estate. Individual Variables in Regression Equation with Effort 201 districts In the stepwise regression equation showing the cumulative relationship between the independent variables and tax effort for all school district six variables accounted for 35 percent of the variation in tax effort as shown in Table XV. Percent of residential property value accounted for 19 percent of the variance in tax effort. Five percent of the variance was attributable to farm property values. For all school districts in Vermont, 25 percent of the variance in school district tax effort could be explained by these two variables alone. Therefore, in districts where the local school revenue was derived from taxes mostly on residences and farms, the school district tax effort tended to be high.

PAGE 67

60 TABLE XV Regression Analysis of the Relationship of Local Tax Effort to Significant Independent Variables for All 201 School Districts Variables . R R Change Simple R ^22 percent residential property value .19356 .19356 .43995 X25 percent property value farms .24859 .05503 .15167 ^8 percent population in school age 5-19 .28813 .03954 .19875 X2 total school expenditures .31374 .02561 .29917 X24 percent property value industrial .33356 .01981 .07122 *20 percent population family income other .35244 .01888 -.01265 Variables Xg, X2, ^2^, and X2Q accounted for an additional 10 percent in the variance in tax effort. Districts which have a high percentage of children ages 5-19 in school explained 4 percent of the variance in tax effort while 3 percent of the variance was explained by total school expenditures. Income was correlated negatively with effort which could indicate that it (income) was not a determiner of effort in the 201 school districts. It should be noted that variable X^^ (percent property value commercial) dropped out of the regression equation. The reason seems to be that it was insignificantly correlated with the other variable and consequently did not provide enough extra explanation of the variance in tax effort to be retained in the equation. Individual Variables in Regression Equation with Effort in 67 High Wealth Districts ~~ ' In the stepwise regression equation for the high property wealth

PAGE 68

61 districts, the cumulative relationship between tax effort and the independent variables, eight variables accounted for 14 percent of the variance in effort as shovm in Table XVI. The percent of residential property value accounted for 4 percent of the variance in tax effort and the percent of population with family income fron non-farm sources accounted for 3 percent. In relatively wealthy districts the value of residential property positively related to high effort. Likewise, the non-farm income factor would suggest that the individuals in wealthy districts may be white collar workers deriving their income from other than farm sources. Likewise, an additional conclusion to be drawn may be that the high effort districts in the wealthier areas are not farming communities. Average daily membership, the percent of economically deprived children and the percent of population 18-65 years of age, each explained less than 2 percent of the variance in tax effort for the support of schools. Each of these variables correlated negatively with tax effort which would indicate that the number of children in the district is not an influential factor in the tax effort they made. Likewise, these districts more than likely have a small number of economically disadvantaged children. With the exception of percent of residential property value, none of the variables in the equation had a simple R above .2161. Individua l Variables in Regression Equation With Effort in 6 7 Low Weal t h Districts In the stepwise regression of the low property w*nitl» districts showing the cumulative relationship between effort nnd the independent variables, eight variables accounted for 42 percent of the variance in tax effort as shown in Table XVII.

PAGE 69

62 TABLE XVI Regression Analysis of the Relationship of Local Tax Effort to Significant Independent Variables for School Districts with High Assessed Valuation of Property per Pupil in Average Daily Membership (N=67) Variables r2 2 R Change Simple R ^22 percent residential property value .04673 .04673 .21617 ^16 percent population family income non-farm .07867 .03194 .16361 ^1 average daily membership .09518 .01651 -.05390 ^21 percent economically deprived children .10283 .00765 -.13198 ^5 percent population 18-65 years of age .11336 .01053 -.06090 ^8 percent of population in school ages 5-19 .13030 .01694 .00475 percent population 25 and over completing high school .14019 .00989 .09880 ^24 percent property value industrial .14879 .00860 -.08453

PAGE 70

63 TABLE XVH Regression Analysis of the Relationship of Local Tax Effort to Significant Independent Variables for School Districts with Low Assessed Valuation of Property per Pupil in Average Daily Membership (N=67) Variables R R Change Simple R X X 25 15 17 20 16 percent property value farms .13453 .13453 .36678 average daily membership .23593 .10140 .32449 total listed value real estate .33294 .09701 .14594 percent of population family income .37165 .03871 .25277 wage and salary percent of population family income farm .38428 .01264 .31920 percent of population family income other .40037 .01609 .10095 percent of population family Income .41062 .01025 ,17363 non-farm percent of property value commercial .42019 .00957 .21962

PAGE 71

64 The percent of property value farms explained a little more than onequarter of the variance in tax effort for these districts. Average daily membership accounted for 10 percent of the variance and the total listed value of real estate 9 percent. The variable percent of population family income wage and salary explained 3 percent while percent of population family income farm, percent of population family income other and percent of population family income non-farm, each explained 1 percent of the variance in tax effort. Although relationship is not high enough to be conclusive, what is suggested from these results is that districts with a high percentage of farm property make a high effort in the support of education. Likewise, districts with expanding enrollments or high enrollments made a significant tax effort as do districts which have a high total listed value of real estate. The income variables, i.e., percent of population family income wage and salary, percent of population family income farm, percent population family income other and percent of population family income non-farm had little significance in explaining tax effort in the 67 less wealthy districts. While the variable (^2^) percent property value commercial explained less than one percent of the variance in effort, it did have a Simple R of .2196. The findings indicate that the most significant variable for all districts and the 67 wealthier districts was the percent of all property which is residential. For the 67 low wealth districts the most significant variable was the percent of property value which was farm property. District which had high residential and farm property were evidently usually willing to make greater tax effort in the support of schools.

PAGE 72

65 Since few income variables were included in the regression equation, it might be concluded that income should not be considered a determiner of tax effort. Factor Analysis of Variables As noted in Chapter III, it was suggested that many of the independent variables might be highly inter-correlated and that in the regression equation some variables might "mask" the significance of others. The independent variables were therefore subjected to factor analysis. The factor analysis does not itself identify the factors that are extracted from the data. There is, however, a correlation coefficient between each variable and each factor. Thus, relationships between factors and variables provide clues regarding the nature of the factors. In the following summary table (Table XVIII) all correlations of .40 or higher between the twenty-five independent variables and the six factors are reported. The .40 cutoff was chosen after consultation with statisticians who suggested the lowest limit generally considered as aiding in the description between a variable and a factor as being .30.^ The .40 level was chosen as being rather conservative. In factor 1, total school expenditures, correlated at .98, while average daily membership and total school current expenditures correlated at .99. Variable (total listed value real estate) correlated at .82 and percent property value commercial correlated at .48. So factor 1 related strongly to expenditures, students, and property values, and was given the label of "wealth and expenditures." In factor 2, three variables correlated at .82 or higher. These three variables were percent population 18-65 years of age, percent of '^Leonard Tashman and David Bee, private discussion at University of Vermont, February and March, 1975.

PAGE 73

66 TABLE XVIII Independent Variables Correlating with Factors at .AO or Higher Variable 1 2 Factors 3 4 5 6 Average daily membership .99 Total school expenditures .98 3 Total school current expenditures .99 4 Total listed value of real estate .82 Percent population 18-65 years of age .84 \ Percent population 65 and over .0/ Percent population in parochial schools K-12 Percent population in school, age 5-19 .49 Percent population 25 and over completing high school .51 ho Percent population 18-64 in labor force .82 11 Percent population family income $1000$5999 .45 .40 -.51 ^12 Percent population family income $6000$9999 .41 .85 ^13 Percent population family income $10,000-$24,999 •51 .41 ^14 Percent population family income $25,000-over 15 Percent population family income wage and salary .87 ^16 Percent population family income non-farm ^17 Percent population family income farm .86 ^18 Percent population family income social security and retirement .87 ^19 Percent population family Income welfare ^20 Percent population family income other .40 .41 .46 ^21 Percent economically deprived children ^22 Percent residential property value ^23 Percent property value commercial .48 ^24 Percent property value industrial hs Percent Property value farms .61

PAGE 74

67 population 18-64 in labpr force, and percent population with family income from wage and salaries. Percent of population 25 and over completing high school, and percent of population with family income $10,000-$24,999 correlated at .51. Three additional variables correlated between .40 and .49. These variables were percent population family income other .40, percent family income $6000-$9999, .41, and percent of population in school ages 5-19, .49. Factor 2 suggests some form of urbanization, since high levels of education and income are related and a high percentage of people in the labor force drawing their Income from wage and salaries tend to be located in most urban areas. This factor, then, was labeled "urbanization." In factor 3, percent population 65 and over and percent population family income social security and retirement correlated at .87 with the factor. Percent population family income $1000-$5999 correlated at .45 and percent population family income other at .41. • Factor 3, therefore, would seem to predominate in areas where there is a relatively older population with limited incomes. These could be rural districts where the young people are moving out of. This factor was designated "elderly limited income." Percent population family income farm correlated at .86 with factor 4 and percent property value farms correlated at .61. Percent population family income $1000-$5999 correlated at .40. The label provided this factor was "farming district." In factor 5 percent population family income non-farm and family income other correlated at .45 and .46, respectively. Percent population family income $1000-$5999 had a negative correlation of -.51 with the factor, while percent population family Income $10, 000$24, 999 had a

PAGE 75

68 positive correlation of .41. This factor would suggest a rural middle class "bedroom" or residential conmiunity. The factor designation, therefore, was "residential middle class district." In factor 6 the only variable with a correlation above .40 was percent of population family income $6000-$9999. This variable correlated with factor 6 at .85. This would indicate a "low middle class district" and the factor was so labeled. A factor correlation matrix was completed to see if any of the factors correlated significantly with each other. (See Table VIII) The study indicated that no one factor was significantly correlated with another. As a matter of fact, factor 2, designated urbanization, had the highest correlation of .0768 with factor 5, residential middle class district. This suggests that no one factor could be Identified as being related to tax effort. Regression Analysis of Factor Scores As indicated in Chapter III, factor scores for each district were computed and the factors placed in a stepwise multiple regression equation with tax effort as the dependent variable. This was done for all 201 school districts and the five factors which remained in the equation explained a little more than 8 percent of the variance in effort. Wealth and expenditures explained 6 percent of the variance, while urbanization explained approximately 2 percent. The other three factors, low middle class, residential middle class and farming districts explained less than one-half percent of the variance in tax effort. When all independent variables were placed in the regression equation, they explained 40 percent of the effort instead of the 8 percent explained by the factor scores. The reason for this could be that the factors represented the averaging of a number of related variables. Thus, a greater

PAGE 76

69 proportion of the variance in effort would be explained by individual variables since they were not comingled with groups of other variables which might reduce their individual correlation with effort. The high wealth districts were next examined. In the factor correlation matrix the simple correlation with the exception of two, all increased, but not significantly. The correlations between the factors wealth and expenditures and residential middle class decreased, as did the correlation between farming districts and residential middle class. The highest correlations were between urbanization and elderly limited income, a negative -.27; and wealth and expenditures and low middle class, a negative -.28. Even though these factors are not significantly correlated, it would tend to indicate that the urbanized high wealth areas do not have the element of elderly with limited income. Likewise, the high wealth districts with a substantial number of low income residents do not have high expenditur In the stepwise regression of the high wealth districts, using factor scores, 7 percent of the variance in effort was atrributable to 4 factors. The factors identified as urbanization and farming were eliminated from the equation. As a matter of fact, factor 5, residential middle class explained 5 percent of the variance, while the remaining three factors together explained 2 percent of the variance in tax effort. Since the variance in tax effort of high wealth districts as identified through the stepwise regression equation is so slight, the conclusion reached is that high wealth districts cannot be predicted through factors of socioeconomic variables. The low wealth districts were then examined. In the factor correlation matrix the simple correlations between expenditures and wealth and urbanization had a negative correlation of -.40, a substantial increase over the .09 correlation between the same factors in high wealth districts.

PAGE 77

70 This indicated that the low wealth urbanized areas do not make high expenditures or have the property wealth to do so. The urbanized low wealth districts would also tend to have a higher percentage of "elderly with limited income" residing in them as evidenced by the positive simple correlation of .24. In the stepwise regression for the low wealth districts, all five factors combined explained 25 percent of the variance in tax effort. The factor "elderly with limited income" was dropped from the equation. Farming districts explained 11 percent of the variance in tax effort, while urbanization explained 5 percent, and wealth and expenditures explained 6 percent, while residential middle class was responsible for 4 percent. This would tend to indicate that many of the low wealth districts are farming communities and they make the greatest effort in the support of schools in these low wealth districts. Urbanized and residential middle class districts made about the same amount of effort. Finally, it should be noted that a greater amount of variance in tax effort was explained by the independent variables and factors in the low wealth districts than in all districts and in high wealth districts. In the regression equation using the independent variables, the total amount of variance in effort explained for all 201 school districts was 4l percent. In the high wealth districts it was 7 percent; but in the low wealth districts, it was 42 percent. Similarly when the factors were used in the regression equation for all districts, they explained but 8 percent of the variance in effort. In the high wealth districts 7 percent of the variance in effort was explained and in the low wealth districts a total of 25 percent of the variance in effort was explained.

PAGE 78

71 The findings suggest that for all 201 school districts, as well as for the 67 high wealth and 67 low wealth districts, variables other than demographic ones may be more important in determining the amount of local tax effort a community makes in the support of education. Some of these non-demographic variables might be a type of local power structure, local tradition regarding education or the attitudes of community leaders toward schooling. Since Vermont is a small rural state organized into many small sparcely populated school districts, the accuracy of the fourth count census data may be questioned. Complete census surveys were conducted in approximately 8% of the districts used in this study. The remaining districts responded to a 15% sampling and thus the socioeconomic variables may have lacked precision of scale to make reliable predictions.

PAGE 79

CHAPTER V CONCLUSIONS AND RECOMMENDATIONS This study found no significant factors correlated to tax effort which would predict the type of districts that make a substantial tax effort for the support of schools in Vermont. However, a study of the independent socioeconomic variables identifies districts with high residential property values as the ones which make a greater tax effort in the support of schools in all districts and high wealth districts. Likewise, it appears, but is not conclusive, that in low wealth districts, districts with high farm value make a greater effort in the support of education than do other types of communities. This study also suggests, but once again is not statistically significant, that in districts with high concentrations of economically disadvantaged children, less tax effort is made for the support of education. It has been found in at least one other Vermont study^ that low income individuals generally have low educational aspiration levels and, therefore, do not consider education important enough to make a significant effort in its support. Most studies have generally found one or more socioeconomic variables reflecting demands for education to be significant. However, there has not been agreement on which variables are crucial. Even where identical models have been used for several states, the importance and even the sign of such variables are not always consistent. As was indicated earlier, the method ''Steven F. Hochschild, "Postsecondary Education Access Study," State of Vermont, Commission on Higher Education Facilities, Montpelier, Vt., 1972. 72

PAGE 80

73 of collecting socioeconomic data and the lack of precision of scale of these variables may have added to the source of measurement error in this particular study. Another factor which could have contributed to the low correlations between the socioeconomic variables and the dependent variable tax effort, aside from those already brought forth is the optional local effort which permits school districts to decide on the amount of money they elect to raise for the support of schools and methods used in property tax administration. It appears, even though the correlations were low, that the wealthy residential areas consistently tend to tax themselves proportionately more than poor rural districts. A state local finance system permitting optional local effort which operates to the benefit of these residential districts tends to provide the greatest amount of educational resources per child in those districts. This study, as many others, did not highly correlate property valuation and family income. These two variables measure different aspects of fiscal ability, that is to say that commercial and business properties are not included or reflected in personal Income; yet taxes which individuals pay, regardless of the base, are paid from current income. Although the study does not fully explain the reasons for the wide variations in tax effort for the support of schools In Vermont, wide variations do in fact exist. Because of this wide variation, the state should intervene and provide assistance which would reduce the disparities in expenditures for education. Further studies are recommended to determine if geographic and political differences in school districts affect the level of the effort they make in the support of education in Vermont.

PAGE 81

APPENDIX A RANK ORDER OF VERMONT SCHOOL DISTRICTS BY TAX EFFORT INDEX FROM HIGH TO LOW 1973 Rank School District Index Rank School District Index 1. Bristol 3.06 36. Westford 1.94 9 2.82 37. East Haven 1.90 3. Hinesburg 2.80 38. Vergennes I.D. 1.90 k. Middlesex 2.70 39. Moretown 1.89 5. Barnet 2.61 40. Pittsford 1.85 6. Hardwick 2.55 41. Benson 1.84 7 RraflfnrH T D 2.51 42. Colchester 1.84 8. Whiting 2.45 43. Randolph 1.84 9. Calais 2.43 44. Castleton 1.83 10. Monkton 2.35 45. Newport City 1.83 11. South Burlington 2.34 46. Orwell 1.82 1 2 2.31 47. Richmond 1.82 13. Shelburne 2.31 48. Huntington 1.81 14. Williston 2.31 49. Sunderland 1.81 15. Charlotte 2.30 50. Waterford 1.80 16. Chester 2.24 51. Hartford 1.79 1 7 1/ . Water bury 2.24 nyae rarK 1 TO i . /o 18. Berlin 2.23 53. Putney 1.78 19. Cornwall 2.21 54. Middlebury I.D. 1.77 20. Johnson 2.19 55. Readsboro 1.7721. Cabot 2.16 56. Derby 1.76 22. New Haven 2.15 57. Springfield 1.76 23. Duxbury 2.13 58. Wlnooski 1.75 24. Milton I.D. 2.10 59. Orange 1.74 25. Proctor 2.10 60. Arlington 1.73 26. Rockingham 2.07 61. Richford 1.73 27. St. Johnsbury 2.05 62. Clarendon 1.72 28. Franklin 2.03 63. Ferrisburg 1. 72 29. Jericho 2.03 64. St. Albans (Uty 1.72 30. Swan ton 2.00 65. Watprville 1.71 31. Danby 1.99 66. Newark 1.69 32. Weybridge 1.98 67. Worcester 1.69 33. Brookf ield 1.97 68. Lowel] 1.68 34. Brattleboro 1.95 69. Williamstown 1.68 35. Fair Haven 1.95 70. West Rutland 1.67 74

PAGE 82

Appendix A Continued 75 Rank School District Index Rank School District Index 71. Albany 1.66 111. Whitingham 1.46 12. Ludlow 1.66 112. Coventry 1.45 73. Rutland City 1.66 113. Washington 1.45 t H • 1 ft'i 114. Braintree 1.44 75. Troy 1.64 115. Craftsbury 1.43 76. Weathersf ield 1.64 116. Fletcher 1.43 11. West Fairlee 1.64 117. Sutton 1.43 78. Bridport 1.62 118. Newfane 1.41 7Q onorenaiu 1 1 Q 1.41 80. Alburg 1.61 120. Middletown Springs 1.40 81. Burlington 1.61 121. Shaft sbury 1.40 82. Hart land 1.61 122. Berkshire 1.39 83. Leicester 1.61 123. Bridgewater . 1.39 OH . weoi^iuxus Lcr X^H . IT WW Lid X 85. Charleston 1.59 125. Athens 1.36 86. Fairfield 1.58 126. Bolton 1.36 87. Kirby 1.58 127. Reading 1.36 88. Norton 1.58 128. Windham 1.36 89. 5?t Alhanfi Tnun A. • J t 129 90. Georgia 1.55 130. Sheldon 1.34 91. Glover 1.55 131. Wolcott 1.34 92. Highgate 1.55 132. Bethel 1.33 93. Barre City 1.54 133. Woodstock 1.32 94. 1 1A X . JH X . JX 95. Northfield 1.52 135. Enosburg Falls 1.31 96. Peacham 1.52 136. Fairlee 1.31 97. Danville 1.51 137. Starksboro 1.31 98. Concord 1.50 138. West Haven 1.31 99. ±.1. ClOL^Ui. g X . J\J 139 1 X . jV 100. Lincoln 1.50 140. Poultney 1,29 101. Rutland Town 1.50 141. Salisbury 1.29 102. Cambridge 1.49 142. Browning ton 1.28 103. Manchester 1.48 143. Rochester 1.28 104. Windsor 1.48 144. Baltimore 1.27 105. Addison 1.47 145. Rupert 1.27 106. Newport Town 1.47 146. Tunb ridge 1.26 107. Barre Town 1.46 147. Vershire 1.26 108. Wallingford 1.46 148. Isle LaMotte 1.25 109. Chelsea 1.46 149. Waitsfield 1.22 110. Ira 1.46 150. Eden 1.21

PAGE 83

Appendix A Continued 76 Rank School District Index Blank School District Index 151. Thetford 1.19 191. Vernon . /Z 152. Westfield 192. Warren 79 153. Burke 1. 18 193. Marlboro . /I 154. Tinmouth 1.18 194. Dover .66 155. Strafford 1.17 195. Wells .63 156. West Windsor 1. 17 196. Goshen . ,yi 157. Morristown 1. 16 197. Peru 158. Stockbridge 1. 16 198. Fayston .30 159. Chittenden 1.15 199. Landgrove .43 160. Wilmington 1.14 200. Plymouth .34 161. Pomfret 1.13 201. Winhall .26 162. Baker sfield 1.11 163. St owe 1. 10 164. Elmore 1.09 165. Walden 1.09 166. Woodford 1.09 167. Granville 1.06 168. South Hero 1.05 169. Belvidere 1.02 170. Jamaica 1.01 171. Pawlet 1.01 172. Londonderry .96 173. Montgomery .96 174. Pittsfield .96 175. Dorset .92 176. Lunenburg .92 177. Halifax .91 178. Barnard .87 179. Greensboro .86 180. Guildhall .86 181. Grafton .85 182. Jay .82 183. Hubbard ton .81 184. Mt. Holly .81 185. Ripton .80 186. Sherburne .80 187. Hancock .79 188. Morgan .75 189. Shrewsbury .75 190. North Hero .73

PAGE 84

X o CO 00 SO o CM O cn in 00 CM O oo2!;^SSSSS^tC5SS oiSJ5So?;iCS3 dddcsdddr^dddddddoOOOOOOCDOO ^ ^ S 3 S g § s I S 2 3 S S § I ^ 1 S ^ 5 S S S § g g S s § s § § g 2 s s s § 1 5 s § s § § s 5 s s ddddddr^lddd ddddcsoooooooooo III II ' 111 H z u M u H W O o z o M ^ td OS PS O U x\ in ^1 CM CM VO CM in CM CO CO 00 00 O CM O r~rH O vO O cn in in VO eg in O vO tH CM M o CO in 00 vO O CJ> r~vO O d I .H O 00 O 00 o so O vO O . vo in r«. 00 o rH CM • • O O I vO r-» CM rCO sr 00 rH CO 00 in CM Ov rH On O 00 CTv m CM CM fH CM 00 O O in rH in CO CO so r*> OS sr 00 rH 00 CM St CM CO CO 00 m «n CM O Ov m ON St vO 00 ON o o I O ON in CM o CM vO 00 o I o o CO o d I O rH CO CO rCO .H O rH ^ m in vO ON r>. in rON CM in o o o o o o vO CM in CM O I CO ON o in CM o o CO m CM o o o o m CM rH CM d I o 00 OO CJN CO • o I 00 O r«m in r«» r-~ CM ON 00 O O CM CO • • • o o o I 00 St ON VO O • o I o ON o CM o Cvl in in 00 VO rH CM CM CO CO VO O O St 00 vO r>» rH CO CO CM O CO in •JCO CM o I ^^OrHrocMrOONCMOON(^rHOrHvOONOO.vor>«oovooovoooNOin c^|2?-OMOTOr2ScMScni*ONi--coooinr>.M-^ SKor>^oSSrHrHr-.COOI^rHCMOCM.fOvO>£OOCOqin«CMON^OOrHvOjH^ >< vOOr->int>-stvOCOCMrH<»ONOr^vOincOstvOCMr>.CMrHrHCMCOin-* OvoinostcMsfcMvo~*r^coOr^rHinoNOcMr>inin^cMvo rH|OrHvOCOOvOCMr^vOCOCOOONOCOCO»a-OcOrHCMOOOvOvO JxJ OvOVOt^OCMvOrHStrHrHOOONvOCOr».v©ONI*«rHCOStrHrH OONONinrHrHOrHrHrHrHrHCMrHOCMOOOr-ICOStrHl-IO rnddddddoooooooooooooooooo II III OrHCMCOStmvOf^OOONOrHCMCO-a-u rH CM CO St m VO r-» 00 CJN rH rH rH rH H '-^ ^ ^ ^T* .i^ .T*

PAGE 85

H 00 rH cn in vO 1^ o\ rH CO CO CO X i-i rH o m o VO • vO • rH • CO • O O • o • o • o • o • o 1 • o o o o o tN n o CM cn -300 r^ CN iH CM r-rH o\ CO CM ^ rH ON X o iH tH CM iH o CM CM CM CM CO • a « • • • • • • • • o 1 O o 1 o 1 O o o 1 O o O O o 00 o CM 00 VO CM CO m rH o 00 o rH a\ Ov vO 00 vO 00 u-i ON 00 CO CO r~ tH iH O 00 00 u-i CM Ov CO X o rH tH rH CM CM o CO -* rH m o o o O o O O O o O O o o I I I I 78 ON vO in m rH o rvj m vO vO in O 00 CM rH o 00 m O Ml csl ON o r>ON in rH o> 00 1 00 00 vO m o co ON vO vO CM CM o rH CM CM CM o • • • • • • • • • • • o o o o o d o tH o o O 1 o o 1 vO CM -aCO o * CM vO in eg -3lO on 00 CM o rH CM CO ON vO vO CM t/S lO rH o CO O CM r~ CO Q ON o in CO in in CO in 1 CO V—' Q rH CM rH o CM • rH o * o O — ^ 1 • 1 v^ • Q • o 1 o 1 o o o 1 in in 00 00 O CO rH CM CJN 3vO CM CM CM in CO r>. ON y 1 00 in rH ON m CM rH rH CM r-H vO ON vO o CO r~~ vD rH CO CM Q rH CO rH CM CM O • • • • • • • d d o o d rH d O o O o 1 O O 1 00 o rH o 00 rH in ON CM m CM Cvl CO O o CM 00 00 in ON iH r» ON rH vO o m m rH in ON 00 CM rH CM o o o vD o 00 CM 00 rH vO r>. O O CM o o CM o CM CO O CO O m O o o o rH O o vO rH X VD in CO CO CM OO CM in CO CM vO vO 00 CM CM O ON m CM rH rH CO CO 1^ 00 ON o m O ON ON CM rH >J rH S)O CM ON 00 ON VtO CM CO in ON ON 00 o O vO 00 rH in -aCJN VO vO 00 CO <3N iH «* O -* O CO O o \0 o o 00 00 VO in H O O 0 o o 1 I m rH X CO vO o rH r«. O CO CM CM o o vO 00 CJN o rH CO VO 00 CM CO 00 CO CO rH vo in O r-aCO O m 00 m 00 O 00 3O rO CO CO o CM m vO vO ^ VO VO 00 O ON m CM VO sr in ON CO vO O 00 -* ON VO On r» rH CM CM CO 00 •* O vO 00 H CO in CO O o I o I o o o o o o o I o I X CO ro CJN CO CM CO CM CM o CO o CM o CJN in ON m in CO CM vO St 00 Cvl -3St in o o vO CM -cf aO CM in rH O O vO St CO O 00 o CO o in o vO O o o St rH VO O 00 O ON o m CM vO St CO CM CM o o CM ON vo CM m O 00 m o CO m 00 VO o •ao o o VO CM ON 00 vO 00 o o o I o o o I o I CO rH X in o CJN o CM vO m 00 ON rH rH -3CM vO On r~ Cvl O CM CM CM 00 r» r-. St -JCO 00 CM CO CM o m St 00 in 00 O CO vo in CM vO O vO CM -ao rH O ON O 00 o o o 00 CO m CO 00 On in VO VO CO O vO St 00 CM m r~. CM O CO vO in O ON o m m CO in CM o m rH t>m O 00 r>. vo rH O vO CO m . St m rH 00 ON CJN rH St St CM CO m rH ON in o CM 00 00 CM rH rH CM ON r>CO CM CM m o O O rH O ON o 00 o o rH o> O CO o ON NO O CM 00 rH m rH -arH vo in rH vO CO CO O « St rH St rH vO CM CM rH O 0 o 1 I O I o I o o o I o I o o o o o I o I rCM o CO vo rH ON 00 O CJN CO ON r>. O rH CO 00 rH rH CO CJN On m CM St vO CaCO in Cvl CM vo CO o 1^ o rH O vO O rH O r~ p«. ON Cvl CO -aCM >ao o m OO ON O O r~ vo St o% CO OO rH CO O vo m St CO sr ON sr 00 in -aCM ON CM CO CO CO ON CM CO CO m 00 CM O ON m o O St O 00 CM vO O O d CO d 0 o 1 I o o I I o I o o o I I o I rHCMC0StmvOr>.00ONrX X X CMCOStinvOr^OOOvOrHCMCO»3-m rHrHrHrHrHrHrHrHCMCMCMCMCMCM ^ ^ ^ ^ ^ ^ ^ ^ ^ ^

PAGE 86

79 lO CM on vO o in CM iH 00 o o cn O o • • • • o d o o o d m r-~ O o in O I o I o I in 1-1 vD o vO o o in o CM m tH O m r«. CO m O CO o o o rH o cn o o • « o d o d d d inorHinooiHONi^mo CMvOOOOOOCMvOOOOOOr^ •^vOiHOvor^moocooocovo CMrH0NV0r-|C0-*Or~O-*^ |x!i-tOOCMCMrHiHi-ICMCMO »••••••••• ooo'oooooooo rH CO CM vO CO . in vO ~300 o in vO in vO in rH CO -* vO . m CO vO o 00 o rH O o CO O CO CM CM CM o CM o rH CO O 1 o 1 o O 1 O O 1 O 1 O 1 O 1 o 1 o 1 o 1 o 1 c o CM o m 00 O o 00 c^ m VO o sr in o CM m CM 00 H CO ON CM CO rH sf CM CO r>. vD sf m -4m X 4CO CO O CM CM o iH o o O O o o o O -3St CO in 00 CM m O m o o sr CO o m rH 00 00 CM CM vO in vO o 00 CM rH 1^ ON ON sr CO in CO o sr rH o O 00 ON ON O o\ o rH O in VO rH vO 00 CO m ON CO o VO VO m rH O CM rH o rH o CM CM o CM rH o o o O O o o o o o rH o o o o o -a a rH CJV rH in 00 CM o CO CM CO m VO sr Ov CM CM O VO sr CO in o vO in in CO o sr ON vO r>. o CJv VO m m CM 00 O CO \o rH CM 00 CO Cvl CM sr 00 H rH CM CO in CO VO CO m m CO r-» o <3N VO CM CM rH 00 CO rH 00 rH 00 vO 00 CM CO CO CM m ON CM o CO CO o 00 CO m CO X! CO CM CM CM O CM o O rH o CO o o rH o o CO CO rH •H o CM O rH rH o I o I o I o I o I o I o o I o I o I o I o I o I c 0) a tw' *^ K> K/* Krfrf »'S >^ •'S >^ »^ 'S »^ in vo 00 CM o i-i CN cn
PAGE 87

BIBLIOGRAPHY Adams, Perry R. , "Socioeconomic Factors Associated with Patterns of School Fiscal Policy in Kentucky," Doctoral Dissertation, University of Florida, Gainesville, 1965. Alkin, Marvin, "Religious Correlates of School Expenditures," A paper prepared for the American Research Association, Chicago, February 11, 1965, ERIC, ED 011143. Campbell, Alan, "The Socioeconomic, Political and Fiscal Environment of Educational Policy Making in Large Cities," Michael Kirst, (Ed), The Politics of Education , McCutchan Publishing Company, Berkeley, California, 1970. Campbell, Alan and Shalala, Donna, "Resource Literature and Educational Revenue," Theory Into Practice , Ohio State University, Vol. XI, No. 2 April 1972. Carlson, DeVere, "Patron Attitudes Toward Selected Educational Issues in Communities with a Dual Educational System," Doctoral Dissertation, University of Pittsburgh, 1969, Dissertation Abstracts, Vol. 30. Clune, William H. , "Taxing and Spending for Public Schools; The Origin Descriptions and Effects of Non-School Taxes and Industrial and Commercial Property," Education and Government Division, Illinois Bureau of the Budget, Chicago, 1971. Cooley, W. and Lohnes, P., Multivariate Procedures for the Behavioral Sciences , John Wiley and Sons, New York, 1962. Cubberley, E., Public Education in the United States , Houghton Mifflin Co. Riverside Press, Cambridge, Massachusetts, 1934. Cubberley, E., School Funds and Their Apportionment , Teachers College, Columbia University, New York, 1905. Dixon, W. T., (Ed.) BMD, Biomedical Computer Programs , University of California Press, Berkeley, 1968. Dye, Thomas, "Politics, Economics, and Educational Outcomes in the States, . University of Georgia, Athens, 1967. Ellis, J., Jr. "A Study of the Relationship Between Local Expenditures on Education and Local Income," Doctoral Dissertation, University of Virginia, 1967, Dissertation Abstracts, Vol. 27. Earner, Frank, "Economic Sociological and Demographic Characteristics of Oregon School Districts and Their Relationship to District Financial Practices," University of Oregon, Bureau of Educational Research and Services, Eugene, April 1966. 80

PAGE 88

81 Fisher, Jack, "A Comparison Between Central Cities and Suburbs and Local Ability to Support Public Education," Doctoral Dissertation, University of Florida, Gainesville, 1972. Gallup, George, "Fourth Annual Gallup Poll of Public Attitudes Toward Education," Phi Delta Kappan , Vol. LIX, No. 1, September, 1972. Garrison, C. B., "An analysis of the Interrelationships of Economic Index of Taxpaying Ability for Schools of Arkansas Counties," Doctoral Dissertation, University of Arkansas, 1965, Dissertation Abstracts, Vol. 26. Gentry, Gilbert, "The Relationship of Certain Cultural Factors to Initiative in the Local Support of Education in Florida," Doctoral Dissertation, University of Florida, Gainesville, 1959. Harvey, E. L., "Property Tax Determinants of Educational Expenditures," Doctoral Dissertation, Stanford University, 1969, Dissertation Abstracts, Vol. 30. Hendrix, V. and Alkin, M., "Population Age Distribution and Public Educational Expenditures." A paper. Educational Research Association, New York, February 1967. Hochschild, Steven F., "Postsecondary Education Access Study," State of Vermont, Commission on Higher Education Facilities, Montpelier, Vt., 1972 Hooper, Harold, "Socioeconomic Factors Associated with Patterns of School Fiscal Policy in Florida," Doctoral Dissertation, University of Florida, Gainesville, 1965. "How to Evaluate Your District's Financial Effort," School Management , January 1965. "Income v. Effort," School Management , January, 1969. James, Thomas; Kelley, J.; and Garms, W., "Determinants of Educational Expenditures in Large Cities of the United States," School of Education, Stanford University, Palo Alto, California, 1966. James, H.; Thomas, A.; and Dych, H. , "Wealth Expenditures and Decision Making for Education," USOE, Research Project No. 1241, Stanford University, Stanford, California, June 1963. Johns, R.; Alexander, K. ; and Jordan, K. F., Financing Education Fiscal and Legal Alternatives , Charles E. Merrill, Columbus, Ohio, 1972. Johns, R. ; Alexander, K. ; and Stollar, D.; (Eds.) Status and Impact of Educational Finance Programs , Vol. 4, National Educational Finance Project, Gainesville, Florida, 1971. Johns, R. and Kimbrough, R., "The Relationship of Socioeconomic Factors, Educational Leadership Patterns and Elements of Community Power Structures to Local School Fiscal Policy," USOE, Project No. 2842, May 1968, ERIC ED 021336.

PAGE 89

82 Kay, Harold B., "A Study of the Relationships Between Selected Socioeconomic Variables and Local Tax Effort to Support Public Schools in Kentucky," Doctoral Dissertation, University of Florida, Gainesville, 1973. King, Charles R. , "Socioeconomic Factors Associated with Patterns of School Fiscal Policy in Georgia," Doctoral Dissertation, University of Florida, Gainesville, 1965. Lazier, Willard E., "School Finance: A Determinant Model for Eligibility," Doctoral Dissertation, University of Utah, Salt Lake City, 1972. MacDougall, M. A., "An Analysis of Income Flow to Educational Expenditures," Doctoral Dissertation, University of Virginia, 196A, Dissertation Abstracts, Vol. 24. Martin, Charles E., "The Relationship of Social and Economic Characteristics to Local Initiative in the Financial Support of Public Schools in Mississippi," Doctoral Dissertation, University of Southern Mississippi, 1964, Dissertation Abstracts, Vol. 23. Metzcus, Richard H. , "Community Human Resources and Local Financial Support for Public Schools," Doctoral Dissertation, University of Illinois, 1969, Dissertation Abstracts, Vol. 30. Meyers, Alfred V., "The Financial Crisis in Urban Schools-Patterns of Support and Non-Support Among Organized Groups in Urban Communities," Doctoral Dissertation, Wayne State University, 1965, Dissertation Abstracts, Vol. 25. Miner, Jerry, Social and Economic Factors in Spending for Public Educations , Syracuse University Press, Syracuse, 1963. Mort, P. R. , and Reusser, W. C, Public School Finance , McGraw-Hill Book Co., Inc., New York, 1941. Myers, H. 0., "A Study of Certain Phases of Local Tax Effort in Relation to Taxpaying Ability in Florida," Masters Thesis, University of Florida, Gainesville, 1950. National Educational Finance Project, "Future Directions for School Financing," Gainesville, Florida, 1971. Nie, Norman; Hull, C. Hadley; and Bent, Dale H., Statistical Package for the Social Sciences , McGraw-Hill, New York, 1970. Patterson, Wade, and Schoonhoven, John, "A Comparative Study of Inconsistent Voter Behavior in School Budget Elections," November 1966, ERIC, ED 011135. Petersen, Thor, "School Approval-Disapproval and Educational Enlightenment of Parents Based on Occupation, Educational Level, Age, Race, Geographic Area and Length of Residency," Doctoral Dissertation, Michigan State University, 1971, Dissertation Abstracts, Vol. 32.

PAGE 90

83 Peterson, L. , Rossmiller, R. , North, S., and Wakefield, H., "Economic Impact of State Support Models on Education," University of Wisconsin, Madison, 1963. Photiadas, J., and Zeller, F., "Attitudes Toward State and Local Taxes in West Virginia: The Preliminary Results of a Survey," University Center for Appalachian Studies and Development, Morgantown, West Virginia, 1968. Quick, W. J. "Socioeconomic Factors Associated with Patterns of School Fiscal Policy in Illinois," Doctoral Dissertation, University of Florida, Gainesville, 1965. Rodriguez v. San Antonio Independent School District, 337 F. Supp. 280 (1971) Reversed 41 "Law Week" 4407 March 21, 1973. Serrano v. Priest, 96 Cal. Rptr ., 601, 487 P. 2d 1241 (1971). Strayer, G . , and Haig, R., "The Financing of Education in the State of New York," Report of the Educational Finance Inquiry Conmiission, Vol. I, Macmillan Co., New York, 1923. Turck, M., "A Study of the Relationships Among the Factors of Financial Need, Effort and Ability in 581 High School Districts in Michigan," Doctoral Dissertation, Michigan State University, 1966, Dissertation Abstracts, Vol. 21. United States Bureau of the Census, "Public Use Samples of Basic Records from the 1970 Census," Washington, D. C, United States Department of Commerce, April 1972. Vermont Department of Education, "Vermont State Aid," Montpelier, Vermont 1972. Von Hatley, R. F., "Family Income Voting Behavior and Financial Referendums: Educational Finance and Politics in Albuquerque, 1968-69," Doctoral Dissertation, University of New Mexico, 1971, Dissertation Abstracts, Vol. 31. Witt, Irving, and Pearce, Frank, "A Study of Voter Reaction to a Combination Bond-Tax Election on March 26, 1968," ERIC, ED 019945.

PAGE 91

BIOGRAPHICAL SKETCH Edward Joseph Fabian was born October 7, 1937 in Middle Granville, New York. He graduated with a B.S. in Education from Castleton State College in 1961, from St. Michael's College in 1966 with a M.Ed, in Guidance and Administration, from the University of Vermont in 1973 with a C.A.G.S. in Educational Planning and Administration, and a Ph.D. in Educational Administration from the University of Florida in 1976. Mr. Fabian worked as an elementary and junior high school teacher in Orwell, Vermont and as an elementary principal in Orwell, Vermont. He moved to the State Department of Education in 1966 as Chief of Education Field Services and currently serves as Deputy Commissioner of Education. Mr. Fabian has been included in Outstanding Young Men of America (1967), Who's Who in State Government (1975) and is a member of Phi Delta Kappa, Vermont Education Association, American Association of School Administrators, as well as many other organizations. He is married to the former Martha Anne Ryan and has five children: four boys, Charles, Kevin, Christopher, Edward, Jr. and one girl, Jullanne. 84

PAGE 92

I certify that I have read this study and that, in my opinion, it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. ^ / V . (. C / .\ Kern Alexander, Ohairman Professor of Educational Administration I certify that I have read this study and that, in my opinion, it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. ' William Alexander, Professor of Curricultim and Instruction I certify that I have read this study and that, in my opinion, it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. I' ^ Arthur J. Lewi,s, Professor of Curriculum and Instruction I certify that I have read this study and that, in my opinion, it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. 1 / / /• . • <' f — . X. — ' I ^ Michael Nunnery, Professor of Educatlxinal Administration y

PAGE 93

This dissertation was submitted to the Dean of the College of Education and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August 1976 Dean, Graduate School



PAGE 1

AN ANALYSIS OF THE PERCEPTIONS OF THE LEADERSHIP BEHAVIOR OF MALE AND FEMALE UNIVERSITY OF FLORIDA ADMINISTRATORS by Barbara Jean Keener A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1976

PAGE 2

ACKNOWLEDGEMENTS The writer wishes to express her sincere appreciation to the many persons who assisted in making this study a reality. Members of the Supervisory Committee, Dr. James L. Wattenbarger and Dr. Arthur Sandeen, Co-Chairmen , Dr. Dorothy Neville and Dr. Harold Riker, should be recognized for their constant scholarly criticism and support. Dr. John Nickens is thanked for his assistance with the statistical computation. A very special appreciation goes to Dr. Wattenbarger for his continual inspiration and guidance throughout the doctoral program. Thanks is also extended to the administrators and their associates who so willingly participated in the research. Encouragement and advice came from many others, too numerous to list here. The writer is indebted to each individual who played a role in her completion of the doctoral degree. ii 1

PAGE 3

TABLE OF CONTENTS Page ACKNOWLEDGEMENTS ii ABSTRACT v CHAPTER I INTRODUCTION 1 Statement of the Problem 2 Delimitations 6 Justification for the Study 7 Definition of Terms 8 Prodcedures 8 Summary 17 II SELECTED REVIEW OF RELATED LITERATURE 19 Leadership Behavior 19 Females in Elementary and Secondary Education Administration 25 Females in Higher Education Administration ... 29 Summary 37 III A DESCRIPTION OF THE UNIVERSITY AND THE SAMPLE POPULATION 39 The University of Florida 39 Administrators' Professional Career Interviews 41 Summary. . 43 IV FINDINGS 44 Summary of the L . B . D . Q . -XI I Findings 45 V DISCUSSION, CONCLUSIONS AND RECOMMENDATIONS 75 Discussion 75 Conclusions 85 Recommendations 87 iii

PAGE 4

TABLE OF CONTENTS (continued) Page APPENDICES A AN INTRODUCTION: THE LEADER BEHAVIOR DESCRIPTION QUESTIONNAIRE — FORM XII 88 B ADMINISTRATOR'S PROFESSIONAL CAREER INTERVIEW GUIDE 98 C LETTER OF AUTHORIZATION 100 REFERENCES 101 BIOGRAPHICAL SKETCH 112 iv

PAGE 5

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy AN ANALYSIS OF THE PERCEPTIONS OF THE LEADERSHIP BEHAVIOR OF MALE AND FEMALE UNIVERSITY OF FLORIDA ADMINISTRATORS by Barbara Jean Keener March, 1976 Chairperson: Dr. James L. Wattenbarger Co-Chairperson: Dr. C. Arthur Sandeen Major Department: Educational Administration The purpose of this study was to determine if significant differences existed in reported perceptions of leadership behavior between University of Florida male and female administrators. The sample population participating was: male and female administrators, their immediate superordinates , and a sample of their subordinates. The Leadership Behavior Descrip tion Questionnaire-Form XII and a Professional Career Interview Guide were used as research tools. Twelve constructs were developed in accordance with the dimensions of the L.B.D.Q. --XII . The constructs were tested at the .05 level of significance. Each construct v\:as tested in four areas: subordinate responses, superordinate responses, V

PAGE 6

administrator responses, and comparisons between groups. Interview information was examined in three areas: career orientation, career development and career aspirations. Comparisons of male and female responses were made. The study found, through the examination of each constrt that there was little difference in the leadership behavior o male and female administrators at the University of Florida. Based on this research, there appears little justification to conclude that female administrators behave differently as leaders than males. The professional career interviews did reveal some differences between male and female administrators in the areas of career orientation, career development and career aspirations . vi

PAGE 7

CHAPTER I INTRODUCTION A niomber of people have pointed out that there are few women in higher education leadership positions. For example, Oltman (1970) wrote, "The actual participation of women in administrative policy-making in higher education is conspicuously lacking." The Carnegie Commission (1973) reported, "If women are thinly represented on faculties, especially in traditionally male fields, they are so rarely represented in top academic administrative positions as to be practically nonexistent in the upper echelons." There are many reasons for such a dearth; but an investigation of all these reasons was not the pursuit of this endeavor. Instead, the major concern in this study was to examine the leadership performance of men and women university administrators. At the heart of the matter is the question of sex qualification. In other words, are men more qualified to fill these leadership positions than women? Are there restrictions which are placed on women that denigrate their leadership potential? Research on women's administrative behavior (as compared to men's administrative behavior) is limited and this study assumes that more should be carried out. 1

PAGE 8

2 Such an investigation will not reveal however, all the explanations as to why there are so few women in policy-making positions. This study came to grips, on the other hand, with the issue of women's performance as university administrators. The concern herein was to isolate the sex of the administrator as a factor in leadership qualities. Since societal stereotypes and expectations rarely "assign" women to decision-making or leadership roles, there appears to be a need to know more about the women who are in these positions. There is a ne^d to discover how they got there and how well they perform. There is a need to determine whether or not there is any perceived difference based solely on sex. Statement of the Problem This study was designed to compare the leadership behavior of male and female university administrators in order to determine to what extent, if any, they behave differently as leaders. The study revolved around the following question: In what ways do male and female administrators perform their leadership functions differently? Precisely, the purpose was to examine the relationship between the leader behavior of men and of women as perceived by the following role categories: 1. The leader behavior descriptions of male and of female university administrators as perceived by their respective, immediate superordinates . 2. The leader behavior descriptions of male and of female university administrators as perceived by their respective subordinates.

PAGE 9

3. The leader behavior descriptions of male and o£ female university administrators as perceived by themselves. From these data, the investigation sought answers to the following questions: 1. Is there a difference in the way superordinates perceive male and female university administrator's leadership behavior? 2. Is there a difference in the way subordinates perceive male and female university administrator's leadership behavior? 3. Is there a difference in the way male and female university administrators perceive their own behavior? The following questions were answered according to the rating received by the male and female administrators on twelve leadership behavior constructs. These constructs are based on the twelve leadership behavior dimensions of the Leadership Behavior Description Questionnaire-XII . Construct 1 : Representation Is there a difference between the perceived degree to which male and female university administrators speak and act as the representative of the group as reported by the various groups, i.e., superordinates, administrators, and subordinates? Construct 2 : Demand Reconciliation Is there a difference between the perceived degree to which male and female university administrators reconcile conflicting demands and reduce disorder to the system as reported

PAGE 10

4 by the various groups, i.e., superordinates , administrators, and subordinates? Construct 5 : Tolerance for Uncertainty Is there a difference between the perceived degree to which male and female university administrators are able to tolerate uncertainty and postponement without anxiety and upset as reported by the various groups, i.e., superordinates, administrators, and subordinates? Construct 4 : Persuasiveness Is there a difference between the perceived degree to which male and female university administrators use persuasion and argument effectively and exhibit strong convictions as reported by the various groups, i.e., superordinates, administrators, and subordinates? Construct 5 : Initiation of Structure Is there a difference between the perceived degree to which male and female university administrators clearly define their own role and let followers know what is expected of them as reported by the groups, i.e., superordinates, administrators, and subordinates? Construct 6 : Tolerance of Freedom Is there a difference between the perceived degree to which male and female university administrators allow followers scope for initiative, decision and action as reported by the groups, i.e., superordinates, administrators, and subordinates? Construct 7 : Role Assumption Is there a difference between the perceived degree to

PAGE 11

which male and female university administrators actively exercise the leadership role rather than surrender leadership to others as reported by the group, i.e., superordinates , administrators, and subordinates? Construct 8 : Consideration Is there a difference between the perceived degree to which male and female university administrators regard the comfort, well-being, status and contribution of followers as reported by the group, i.e., superordinates, administrators, and subordinates? Construct 9 : Production Emphasis Is there a difference between the perceived degree to which male and female university administrators apply pressure for productive output as reported by the groups, i.e., superordinates, administrators, and subordinates? Construct 10 : Predictive Accuracy Is there a difference between the perceived degree to which male and female university administrators exhibit foresight and ability to predict outcomes accurately as reported by the groups, i.e., superordinates, administrators, and subordinates? Construct 11 : Integration Is there a difference between the perceived degree to which male and female university administrators maintain a closely-knit organization and resolve intermember conflict as reported by the groups, i.e., superordinates, administrators, and subordinates?

PAGE 12

6 Construct 12 : Superior Orientation Is there a difference between the perceived degree to which male and female university administrators maintain cordial relations with superiors, have influence over them, and are striving for higher status as reported by the groups, i.e., superordinates , administrators, and subordinates? Each construct was examined using Fisher's exact test and chi square at the .05 level of significance. In addition to the data collected for examining differences between male and female administrators for each of the stated constructs, a professional career interview was conducted with each administrator. The interviews were conducted in order to provide greater depth regarding individual perceptions and careers. Delimitations The following restrictions were observed in conducting the study: 1. The study of leadership qualities was limited to individuals who were currently employed at the University of Florida in 1975. 2. The data collection was limited to responses to the leadership questionnaire and the responses to the personal interviews. 3. The scope of the study was confined to measurable leadership qualities as defined in the L.B.D.Q. --XII and personal observations concerning professional career patterns. 4. The sex of the superordinates and subordinates was not considered as a factor in this study.

PAGE 13

5. The researcher conducting the personal interviews with the male and female administrators was female. 6. The researcher conducting the personal interviews also distributed and analyzed the survey instrument. Justification for the Study There appears to be a paucity of information providing a basis for comparison of female and male university administrators. Little has been done to seek out and define the leadership performance in terms of a female-male contrast. The recent emphasis on involving more women in administration occasioned by legislature, both state and national, as well as societal influences requires that more information about perceptions of administrative performance of men and women be collected and analyzed. These responses, coupled with knowledge of current female administrators' backgrounds and career patterns, should reveal fundamental information about female administrators in higher education. This type of survey should be able to provide guidelines to those persons counseling future administrators. The career patterns and performances which are discerned could be of benefit to those involved in the hiring and/or promotion of females in administration. For those women seeking higher management positions, a knowledge of other females' performances could be beneficial. In general, this study should assist in the clarification

PAGE 14

8 and understanding of women's administrative leader behavior as compared to men's in a university setting. Definition of Terms University Administrator : An employee of the University of Florida who has faculty rank (courtesy or appointed) and has reported fifty percent or more time devoted to administrative duty. The employee should have rank of departmental chairperson or above. This includes institute directors, deans, associate deans and assistant deans. Leader : An individual who, on the basis of his office or official status in an organization, is placed in the position of being able to influence the activities of that organization as it attempts to achieve its goals. In this study, the leader is identified as a university administrator. Leadership : The overt actions in which a leader engages in influencing organizational activities. Superordinate : A person with rank above the identified administrator. The person to whom the identified administrator reports organizationally. Subordinate : A person with rank below the identified administrator; is directly responsible to the identified administrator. Procedures The procedures section is divided into three parts. The first part includes the study's design and the selection of the sample. The second area is an explanation of the development of the instruments and the data collection process. The

PAGE 15

last part deals with the treatment of the data after collection. Design and Sample Female administrators . Female administrators were located by computer survey of the University of Florida faculty. The computer program located all female faculty members (courtesy or appointed) , with rank of department chairperson or above and time records of fifty percent or more devoted to administrative duty. Nineteen female administrators were located by this method. One of the administrators served on the dissertation committee for this work and therefore was not used in the survey. Consequently, the actual number of female administrators participating in this study was 18. In accordance with the "administrator" definition, these administrators were in mid-management positions or above. They included institute directors, assistant institute directors, co-ordinators , deans, assistant deans, associate deans, and department chairpersons. Male administrators . Male administrators were located using the same basic computer survey as that for female adminis trators. After the male administrators were identified, they were matched to the 18 female administrators on the basis of compatible degree, title, and faculty rank. All male administrators with job descriptions similar to the 18 female administrators were numbered and selected by a simple random method. Kendall's (1960) Table of Random Numbers was used.

PAGE 16

10 Due to the nature o£ the "administrator" definition, these administrators were in mid-management positions or above They included institute directors, co -ordinators , deans, assistant deans, associate deans and department chairpersons. Superordinates . The administrators were requested to identify the person who was their direct superior or superordinate. One superordinate was identified for each administrator. The study identified the highest ranking female administrators at the University of Florida in order to observe them as administrators . All participating superordinates were male. Subordinates . The administrators were requested to identify a person directly responsible to them. This person was designated the subordinate. One subordinate was identifie for each administrator. The sex of the subordinate was not considered as a factor in this study. However, it can be note that a majority of the female administrators' participating subordinates were female and a majority of the male administrators' participating subordinates were male. The selection of only one subordinate for each administrator in the study could bias the results since the selection was done by the administrator being rated. However, the same bias would apply equally to all the subordinate respondents. It should also be noted that about one-third of the administrators studied actually had only one subordinate. A majority of the administrators with only one subordinate were female.

PAGE 17

11 Instrumentation and Data Collection Instrument . The instrument used in this study was the Leader Behavior Description Questionnaire-Form XII developed by Ralph Stogdill (1963). The questionnaire was designed to obtain descriptions o£ leaders, through 12 dimensions o£ leader behavior. The leader behavior is perceived objectively in terms o£ the 12 dimensions' frequency of occurrence. A copy of the instrument is included in Appendix A. In addition to the basic instrument, a professional career interview was conducted (by this researcher) with each administrator. These interviews were employed to obtain information on the administrators' career development, orientation and aspirations. A copy of the interview guide is included in Appendix B. The interview guide was constructed on the basis of questions asked by Arter (1972) in her survey of female university administrators at state universities and land-grant colleges. The questions were adjusted to the personal interview setting. Data Treatment L.B.D.Q. --XII consists of 100 items describing leader behavior. Each item is answered with a forced choice format, with one of five possible responses: always, often, occasionally, seldom and never. Each item receives a score from five to one. Each subscale score consists of the sum of the scores from the items of the subscale.

PAGE 18

12 Stogdill (1970) concluded that subscales of the L.B.D.Q. --XII were differently related to different dimensions of leader personality, member satisfaction, and group performance. His theoretical work was based on the factors of identifiable behavior patterns. The following 12 dimensions of leader behavior were defined in the L.B.D.Q. --XII : 1. Representation. The perceived degree to which an » individual speaks and acts as the representative of the group. 2. Demand Reconciliation. The perceived degree to which an individual reconciles conflicting demands and reduces disorder to the system. 3. Tolerance for Uncertainty. The perceived degree to which an individual is able to tolerate uncertainty and postponement without anxiety and upset. 4. Persuasiveness. The perceived degree to which an individual uses persuasion and argument effectively and exhibits strong convictions. 5. Initiation of Structure. The perceived degree to which an individual clearly defines his own role and lets followers know what is expected of them. 6. Tolerance of Freedom. The perceived degree to which an individual allows followers scope for initiative, decision, and action. 7. Role Assumption. The perceived degree to which an individual actively exercises the leadership role rather than surrender leadership to others.

PAGE 19

f 13 8. Consideration. The perceived degree to which an individual regards the comfort, well-being, status, and contribution of followers. 9. Production Emphasis. The perceived degree to which an individual applies pressure for productive output. 10. Predictive Accuracy. The perceived degree to which an individual exhibits foresight and ability to predict outcomes accurately. 11. Integration. The perceived degree to which an individual maintains a closely-knit organization and resolves intermember conflict. 12. Superior Orientation. The perceived degree to which an individual maintains cordial relations with superiors; has influence over them, and is striving for higher status. A letter authorizing the use of the L.B.D.Q. --XII in the study is included in Appendix C. Validity In Stogdill's (1970) Review of Research on the L.B.D.Q. --XII , he explained the test of validity given. In order to test the validity of the subscales of the L.B.D.Q. --XII , Stogdill (1969) with the assistance of a playwright, wrote a scenario for each of the subscales. The items in a subscale were used as a basis for writing the scenario for that pattern of behavior. Experienced actors played the role of supervisor and workers. Each role was played by two different actors. Motion pictures were made of the role performance. Observers used L.B.D.Q. --XI I to describe the behavior of the supervisor. No significant difference was found between two

PAGE 20

14 different actors playing the same role. However, the actors playing a given role were described significantly higher in that role than in other roles. Since each role was designed to portray the behaviors represented by the items in its respective subscale, and since the same items were used by the observers to describe the playing of the role, it can be concluded that the scales measure what they are purported to measure (p . 5) . Reliability A modified Kuder-Richardson formula was used to determine the reliability of the L.B.D.Q. --XII , Each item was correlated with the remainder of the items in the subscale. The resulting reliability coefficients ranged from .54 to .87 for nine different groups of leaders, indicative of sufficient reliability for use in this study (Stogdill, 1963). Administration The Leader Behavior Description Questionnaire-Form XII can be used by a leader to describe his own behavior, or by the leader's associates to describe a given leader. The L.B.D.Q. --XII (Appendix A) was administered to each identified administrator, the administrator's superordinate , and a subordinate of the administrator. The administrator was interviewed using the professional career interview guide (Appendix B) . Method of Securing Data Initial contact with the administrators was made by telephone (by this researcher) when interview appointments were scheduled. Following the interviews, administrators were

PAGE 21

15 requested to complete the L . B . D . Q . -XI I . Copies o£ the surveywere left with the administrators. Administrators were also requested to identify one of their superordinates and one of their subordinates. Identified superordinates and subordinates were contacted in person, in their offices, and requested to participate in the study. A copy of the L.B.D.Q. --XII was left with each superordinate and subordinate. ' Instructions for completing the survey were placed on the first page of the instrument (Appendix A). The instrument required 30 minutes to complete. Confidentiality was promoted with each subject asked to return his/her survey directly to the researcher. A series of follow-up contacts were made with those administrators, superordinates, and subordinates who were slow in returning their surveys. Method of Statistical Analysis The data received from the L.B.D.Q. --XII were analyzed, using several techniques and processes. Responses were treated in six categories: 1. Female administrator's self -evaluation. 2. Female administrator's superordinate ' s evaluation. 3. Female administrator's subordinate's evaluation. 4. Male administrator's self -evaluation. 5. Male administrator's superordinate ' s evaluation. 6. Male administrator's subordinate's evaluation.

PAGE 22

16 The University of Florida Computer Center services were employed to analyze the data for each group: superordinates , subordinates and administrators. The computer test was run after the results for each participant were scored. Hollerith cards were key punched for each participant. Each card was punched with the score for each test item, sex, and group of the respondent. The cards were coded and classified in order to assess the following: Superordinates-superordinates ' perceptions of male administrators, and superordinates' perceptions of female administrators; Administrators -males ' perceptions of their own leader behavior, and females' perceptions of their own behavior; Subordinates-subordinates ' perceptions of male administrators, and subordinates' perceptions of female administrators. This procedure was done in order to determine if the three groups perceived the leader behavior of male and female administrators differently. If differences existed, methods were used to determine in which dimensions of the L.B.D.Q. --XII such differences occurred. Testable Questions In order to answer the questions of how the male and female administrators were perceived, a Fisher's exact test and chi square were run at the .05 level of significance using data from the L.B.D.Q. --XII for each construct. Each question was asked in regard to the respondents' groups -subordinates ,

PAGE 23

17 superordinates and administrators. Two computer programs were run, using the Statistical Package for the Social Services -Version 6.00. The first computer analysis was used to identify the mean scores of all the administrators on each of the twelve constructs. Administrators were assigned to category 1 or category 2 as follows: Category 1 contained all scores equal to the mean or higher. Category 2 contained all scores below the mean. Male and female administrators were compared by group and tests of statistical significance for differences were performed by use of chi square and Fisher's exact test in the second computer analysis. The comparison was based on a 2 x 2 test using male administrators female administrators and category 1 category 2. A personal interview was held with each male and female administrator in the sample. The professional career interview guide (Appendix B) was used as a basic format for each interview. Descriptive information from each interview was compiled and reviewed. Patterns and trends were ret'ognized in relation to the administrators' sex. Summary This study's sample population was composed of 18 female administrators and an equal sample of male administrators at the University of Florida. The total population included the identified administrators, their superordinates, and their specified subordinates.

PAGE 24

18 The Leader Behavior Description Questionnaire-Form XII , developed by Ralph Stogdill was the instrument used to assess perception of leader behavior. For this research study, leader behavior was measured in terms of perceptions by the administrators, superordinates , and subordinates. The validity and reliability of this instrument was established by Ralph Stogdill through the Ohio State Leadership Study group. A professional career interview guide was developed to direct questions to the male and female administrators. The administrators participated in open ended interviews concerning their careers. Information gained was used to compare the administrators on the basis of sex. The data for each of the 12 constructs of the L.B .D.Q . XII were separated by sex of the administrator and groupadministrators, superordinates and subordinates. They were then subjected to chi square and Fisher's exact test to determine if there were any differences between sex and group categories at the .05 level of significance.

PAGE 25

CHAPTER II SELECTED REVIEW OF RELATED LITERATURE The purpose of this chapter is to review research literature relevant to the study of male and female universityadministrators. This study focuses on leadership behavior of university administrators. However, a review of females involved in all levels of education administration is pertinent to the major theme of the study. The review will include a synopsis of literature dealing with females in all levels of educational administration. The chapter contains three sections entitled: Leadership Behavior, Females in Elementary and Secondary Education Administration, and Females in Higher Education Administration. Leadership Behavior Recent leadership theories are reviewed in light of the leadership theme in this study. There appear to be six categories of leadership theories. This specific research effort is aligned with the "interaction-expectation" theories. Literature covering the other five types of leadership will only be briefly discussed in this chapter. Environmental Theories Murphy (1941) contended that leadership qualities are not a factor of the individual, but a function of the occasion. 19

PAGE 26

20 The leader evaluates a situation and is the instrumental factor through which a solution is achieved. The cultural setting is the controlling factor according to Schneider (1937) . He found that great military leaders in England emerged in proportion to the number of conflicts. Great Man Theories Dowd (1936) concluded that individuals possess different degrees of intelligence, energy, and moral force, therefore, a superior few will inevitably emerge as leaders. Bernard (1926) advanced similar conclusions in his study of social psychology. He maintained that leadership can be explained in terms of traits of personality and character. Thus, trait theories could be used to explain leadership. Personal-Situational Theories A combination of the two theories previously mentioned provides another method for describing and studying leadership. Case (1933) suggested that leadership is produced by three factors: (1) the personality traits of the leader, (2) the nature of the group and of its members, and (3) the event (change or problem) confronting the group. Gibb (1954) maintained that when group formation and interaction takes place leadership is a natural interactional phenomenon. Bennis (1966) dealt with five considerations when developing leadership theory: (1) impersonal bureaucracy and rationality of measures, (2) informal organization and

PAGE 27

21 interpersonal relations, (3) benevolent autocracy that gets results because it structures the relationship between superiors and subordinates, (4) job enlargement and employeecentered supervision that permits individual self-actualization, and (5) participation management and joint consultation that allow integration of individual and organizational goals. Cattell (1951) suggested that leadership represents a dynamic interaction between the goals of the leader and the goals and needs of the followers. It functions to help the group decide upon a goal and to help the group find the means to a goal. Humanistic Theories The development of effective and cohesive organizations was a major concern of Argyris and McGregor. Argyris (1964) saw a basic conflict between the organization and the individual. He stated that it is the individual's character to be self -directive and to seek fulfillment through exercising initiative and responsibility. It is a tendency of organizations to structure member roles and to control performance in the interest of achieving specified objectives. An organization which can balance the followers' creative contributions as a natural outgrowth of their needs for growth and self-expression will be the most effective. McGregor (1966) perceived leadership on the basis of two organizational types-Theory X and Theory Y. Theory X attempts to direct and motivate people on the assumption that people are

PAGE 28

22 passive and resistant to organizational needs. Theory Y presumes that people already possess motivation and desire for responsibility and therefore attempts to arrange organization conditions in order to make possible fulfillment of their needs while directing their efforts to achieve organizational objectives. Likert (1967) observed that leadership is based on interaction. The leader builds group cohesiveness and motivation for productivity by providing freedom for responsible decision making and exercise of initiative. Exchange Theories Jacobs (1971) conceptualized leadership in terms of a social exchange theory. It is based on the assumption that the group provides status and esteem satisfaction to the leader in exchange for his contributions to goal attainment. Leadership implies an equitable exchange relationship between leader and followers. Acknowledgement of role obligations allows each party to satisfy the expectations of others on an equitable basis . Interaction-Expectation Theories Homans ' (1950) theory of the leadership role is based on three variables: action, interaction and sentiments. Leadership is defined in terms of the organization of interaction. The greater the frequency of interaction and participation between members, the greater the mutual liking and clarity of group norms. The higher the rank of a person within the group,

PAGE 29

23 the wider the interaction, the more likely his activities conform to group norms, and the greater the number of group members for whom he originates interactions. When group tasks are dependently related to one another and to a solution of a common problem, leadership arises according to Hemphill (1955) . This type of theory was the basis for the development of Stogdill's (1959) expectancy-reinforcement theory of role attainment. According to Stogdill, as group members interact, their roles are defined by mutually confirmed expectations relative to the performances and interactions. As each individual interacts, he is judged by the contributions he makes to the group. The initiation and maintenance of structure defines the level of interaction and expectation. In the predistinguished leadership positions the leader is expected to play a role that differs from the roles of other group members. This expectancy reinforcement model appears to be applicable to the study of university administrators. Stogdill's research has yielded several leadership surveys. One form, Leader Behavior Description Questionnaire-Form XII , will be used in this study. The leader behavior approach has been used previously in a variety of manners for the study of school administration. For example, Halpin (1956) studied 50 Ohio school superintendents to determine the relationship between the superintendent's own perception of how he or she behaved on the Leader

PAGE 30

24 Behavior Description Questionnaire as compared with the school board and staff perceptions. Brown (1966) focused on the leader behavior of 170 principals in Alberta, Canada. Using the L.B.D.Q. --XII , the findings indicated that: (1) teachers' estimates of the school's performance was not sensitive to the perception of leadership in the school, (2) teacher satisfaction was sensitive to the perception of the school leadership, and (3) conficence in the principal was related to the school leadership. Jacobs (1965) compared "high" and "low" innovating principals using the L.B.D.Q. --XII . He found that the high innovating principals received higher ratings on six dimensions of leader behavior: Initiating Structure, Predictive Accuracy, Representation, Integration, Persuasion, and Consideration. Additional literature addressing leadership theories which are relevant to this study will be briefly discussed before the remaining sections of this chapter. Kimbrough (1968) wrote that leadership must involve more than the personal characteristics of the leader. He suggested that the leadership role is either enhanced or suffers depending on how it is valued by members of the system. This perception is seen as an important facet in the value differentiation of leader behavior. Halpin (1966) contended that what the leader does and how he does it as perceived by others who work with the leader is the nucleus of leader behavior. Evaluation of leader behavior can be made in terms of the individual, the group, or both.

PAGE 31

25 According to Owens (1970) , the study of leadership should examine and measure the performance of the leader rather than human traits or other phenomena. The focus should be on observed behavior. In other words, researchers should give attention to what has happened or appears to be happening rather than on finding the cause of observed behavior. Females in Elementary and Secondary Education Administration As previously mentioned, there is a paucity of research dealing with females as education administrators. Numerous pages could be filled with publications on education administration and higher education which fail to mention the particu lar role and scope of female participation. At the same time, there is current literature pertaining to women in higher education which says little, if anything, about the administra tive realm. A recent Change publication, Women on Campus (1974) , offers sixteen articles, none of which discuss the female as an administrator nor a potential administrator. The Carnegie Commission: Opportunities for Women in Higher Education (1973) appears to be a thorough investigation Yet a discussion of female administrators is limited to one page dealing with the decline in the number of women holding administrative positions in coeducational institutions. The Commission suggests that women are so rarely represented in top administrative positions as to be practically nonexistent.

PAGE 32

26 Data provided reinforce this analysis. Females were found to be a low percentage of academic administrators in 454 institutions surveyed. This type of information partially explains why such a dearth of research exists -there are few female administrators in higher education. Howeverstudies of females in elementary and secondary education are also few in number. Studies from that area seem particularly pertinent here. Taylor's (1971) survey found women constitute 67% of the total elementary and secondary teaching force, while 971 of secondary principals and more than 99% of the superintendents are men. The percentage of women elementary principals (211) is actually lower today than it was in past decades. Taylor also showed that all other things being equal, superintendents (male) were not likely to hire women as administrators. She concluded that the only factor which appeared to have any significance on the hiring process was that of sex. The other variables -age , type of position, length of experience, size of school district-did not have any valid correlation with the hiring process. Hemphill, Griffiths, and Frederiksen (1962) found that male principals were preferred by boards of education, although they did not demonstrate superior performance. The study concluded that women tend to score higher than men in ability to work with teachers and outsiders, were more concerned with objectives, possessed greater knowledge of teaching techniques.

PAGE 33

27 and were able to gain positive reactions from teachers and superiors . A University of Florida-Kellogg leadership study team (Grobman and Hines, 1966) attempted to identify and clarify good and poor principal behavior. The team concluded that women were more democratic than men and outscored them in using effective administrative practices. Morsink (1966) studied leader behavior of men and women secondary principals. She discovered that men had more tolerance for freedom. Women scored better in being persuasive in argument, emphasizing production, speaking and acting as a representative of the group, maintaining cordial relations with superiors in exerting influence and striving for higher status. Gross and Trask (1964) conducted part of a National Principalship Study on the difference between men and women elementary principals. They found that women principals gave greater importance to the differences between individual students, placed more emphasis on the detection and assisting of delinquency-prone pupils in their schools and generally were more concerned for the students in their schools. No differences between the sexes were found to exist in the importance they attributed to the academic performance of pupils or in the emphasis placed on the discipline of pupils. Women principals were more likely to require teachers to conform to their standards. No difference was found between the sexes in the amount

PAGE 34

28 of support given to a teacher in a conflict with a pupil or in the extent to which parents were involved in school activities. The sex of the principal was not related to the morale of the teachers in the school. Hoyle (1967) in a Texas study found that women principals were more often aware of the problems facing their teachers than were male principals. No difference was found between the sexes in terms of involving teachers in decision making or encouraging teacher initiative. A study of New York State administrators (Mann, 1971) categorized principals as delegates, trustees, or politicos. A delegate was considered a principal who was guided by citizen preferences even at the expense of his/her own judgment. A trustee was a principal whose decisions were usually based on his own values even if they conflicted with the community sentiment. A politico principal was one who borrowe from both trustee and delegate as aictated by the situation, ' but who had some internally consistent reason for doing so. Women were most likely to be delegates. Proportionately fewer women were trustees and no women were politicos. Longstreth (1973) analyzed the perceptions of leadership behavior of male and female principals in Florida. In a test of twelve leadership dimensions, differences between the sexes were found as follows: — 1. Female principals perceived themselves as regarding the comfort, well-being, status and contribution of their

PAGE 35

29 staffs to a significantly higher degree than did the male principals . 2. Male staff perceived male and female principals as applying a higher degree of pressure for production output than did female staff. 3. Male and female staff perceived female principals as speaking and acting as the representative of the staff to a significantly higher degree than male principals. 4. Males perceived female principals as either "subordinate centered" or "boss centered" and male principals as "boss centered." Females perceived female principals as "subordinate centered" and male principals as "boss centered." (Longstreth, 1973, page 96) In measures of significant difference, however, Longstreth concluded that a principal's sex is not a significant factor in overall leader behavior. A position paper prepared by the Recruitment Leadership and Training Institute (1974) summarized the status of women in public educational administrative positions. It cited the exclusion of women from administrative positions and urged that continued research and discussion be done in an effort to identify areas for innovation and change. F emales in Higher Education Administration The majority of research in this area has been done in the form of doctoral dissertations. These studies, along with additional research directly related to female administrators

PAGE 36

30 in higher education, are reported on in this section of the literature review. In 1961, Kaufman conducted a study to determine the status of women in higher education in selected institutions in the United States. The purpose of the study was to identify and analyze employment practices in appointing women to administrative positions. Kaufman hypothesized that sex would be the determining factor in making administrative appointments when all other variables were equal. Among the findings of the investigation, Kaufman discovered that: (1) there was an incongruence between theory and practice in regard to sex when making administrative appointments, (2) in the institutions surveyed, the proportion of women administrators was small, (3) women were appointed to positions considered typically female positions, (4) previous experience and education were of prime importance in preparing women administrators, (5) there was a lack of qualified women available to appoint to administrative positions, and (6) when equally qualified, the men generally advanced more readily. Gardner (1966) conducted a survey to determine career patterns of women administrators in Illinois. The study was an attempt to provide useful information to women interested in administrative careers in higher education. Gardner's subjects were classified in four administrative areas: head librarians, deans of women, registrars, and other deans (a miscellaneous category) .

PAGE 37

31 For all respondents in general, Gardner found that: (1) prior to holding their present administrative position, 48% o£ the women had been assistants to the chief administrators in their respective areas; (2) one half of the respondents indicated that they had begun their administrative careers within the range of ages 26 through 35; and (3) over one half had not been assisted in obtaining their positions. Gardner found that women administrators tended to be promoted to such positions from office staff and teaching positions. Since the majority of the respondents indicated that few courses were of particular help to them in their careers, Gardner deduced that the respondents had not been trained for administrative positions. She recommended further studies to be made to determine helpful courses of study for women administrators. Simpson C1970) conducted a study of six colleges in Pennsylvania to determine whether those persons responsible for employing faculty and administrators would express a preference for male candidates. Additionally, he sought to determine whether those responsible for making academic appointments who repeatedly excluded female candidates would relegate women to lesser positions. Simpson's research instrument utilized brief descriptive resumes of fourteen individuals paired as candidates for theoretical administrative positions. The results of the survey indicated that those responsible for employing faculty and

PAGE 38

32 administrators did show discrimination against female candidates when considering equally qualified male and female candidates. Harris' research (1970) evolved from testimony before the Special House Subcommittee on Education in 1970. She found women faculty tended to be excluded from positions in prestigious universities. Women were generally in the lower professional ranks, finding it difficult to rise to the rank of full professor. The study also showed that women administrators were scarce in number and generally held appointments in areas traditionally considered female. At the same time a number of women with qualifications equal to many directors were appointed as assistant directors doing the same work and for less financial renumeration. Although inequities do exist, Harris concluded it is most difficult to determine between de facto and de jure sexual discrimination. An attempt to define the attitudes of women in administrative positions as reflected by their involvement on campus and to create an awareness of discrimination where it might exist was the purpose of an American Association of University Women survey to which 454 member institutions responded (Oltman, 1970) . The major findings of the study indicated that (1) although their promotional policies were the same for men and women, 34 educational institutions had no women department heads and the

PAGE 39

33 mean number of women department heads in all institutions was less than three, (2) although 92% indicated their institutions included women in upper level administrative positions, women were seldom employed in positions which involved critical decision making, nor were women actively recruited to upper level positions, (3) women were generally found in positions which . involved minor policy making decisions at the middle management level or were in positions typically considered female, and (4) greater opportunities for women were found in the administration of women's colleges. In an article about women in higher education administration, Carroll (1972) expressed concern for the lack of women in upper level administrative positions in coeducational colleges and universities. No longer can women claim the position of dean of women. In many institutions this position has been eliminated with the creation of the dean of students' position to which a man has generally been appointed. Carroll hypothesized that there were three reasons for the scarcity of women administrators in higher education: (1) women tended not to seek administrative positions, (2) when administrators vacated positions they do not recommend women for their positions, and (3) women are not sought for administrative positions by those individuals responsible for selecting administrators . Noll (1973) investigated the opinions of policy making officials towards the hiring of women administrators, in public, two-year educational institutions. The study found no

PAGE 40

34 relationship between female policy makers and the number of female top-level administrators. Both male and female candidates were expected to possess the same personal characteristics of primary importance-ability and professional experience. Emotional stability was considered the third most important characteristic for females, whereas organizational ability rated number three for the males. The majority of policy-making officials in the two-year institutions surveyed would be willing to recommend a female for a top-level administrative position in their own district/ institution. Fecher (1972) studied the career patterns of the 650 women administrators in positions not typically held by women in public coeducational institutions listed in the Education Directory 1970-71 : Higher Education , other than deans of women, deans of schools of nursing, deans of schools of home economics and librarians. Among the major findings were the following: (1) a large percentage of the females reported that they served on policy making committees but that they had little influence on policy decisions; (2) women administrators in higher education in positions not traditionally held by women suffered the same restrictions on sex and employment that are apparent throughout society; (3) the field of student personnel services appeared to have offered greater employment opportunities for women administrators in higher education than other areas in the administration of higher education; (4) women administrators felt that

PAGE 41

35 being married was neither a disadvantage nor an advantage as an administrator; (5) most females in administration accepted new positions within the same institution rather than seeking new positions elsewhere. The Pfiffner (197 2) study attempted to determine some characteristics that women in the highest three levels of administration in the California public community colleges had in common. This research effort concluded that most women administrators did not feel they were discriminated against because of being a woman administrator. The five personal characteristics which women felt were most important for a toplevel administrative career were: (1) the ability to work with others, (2) a strong personal value system, (3) fairness and objectivity, (4) sensitivity toward people and (5) a sense of humor and humility. It is reported in the study that few women became academic administrators for two main reasons. First, the sex-role women learned to play did not include this occupation as an option. Second, the discriminatory attitude exhibited by both men and women toward women becoming administrators precluded their doing so. In order to develop a larger number of women leaders, Pfiffner urged that some attitude changes by society toward the development of each human being's potential be made. She also suggested that pressure from various sources, especially legal and economic pressures related to the provisions under government contracts, would enable more women to become involved in administration.

PAGE 42

36 Arter (1972) surveyed the role of women in the administration of state universities and land-grant colleges. In the investigation it was found that over one half of the state universities and land-grant colleges did not have women in toplevel administrative positions. Over half of the institutions queried did not appoint women to administrative posts in the last five-years prior to the study. Over one third of the institutions did not consider women for administrative posts during the last five year period. Ninety-three percent of the institutions surveyed stated that they would consider qualified women for top-level administrative posts. A profile of the women administrators surveyed showed them to have prime responsibility for personnel and academic programs; confidence and authority to make decisions; responsible to administrators other than the governing board, chancellor, president, or vice-president; personally responsible for carrying out policy, delegating authority, overseeing implementation, and transmitting decisions. The women administrators studied planned to remain in administration. Service, dedication, and challenge ranked highest as their reasons for working. LaPuma (1972) examined the literature of higher education in order to determine the attitudes toward the employment of women in higher education. The study was concerned with selected books that dealt with personnel policies and practices relative to higher education, published or reprinted between 1960 and 1970.

PAGE 43

37 Research revealed that discriminatory attitudes towards women faculty members and administrators declined during the decade of the sixties. However, respondents stated that women were less qualified and less committed to an academic career. At the same time, the literature suggested that colleges and universities do not provide women with the same opportunities they grant men. The research seemed to illustrate that sex is a determinate in the academic marketplace. Summary Very little research is available concerning females in education administration. Comparisons of male and female administrators are particularly scarce. Studies, of any nature, of female administrators in higher education are also few. Fundamental to this lack of research is a dearth of female administrators. According to recent surveys, there is a low percentage of female administrators at every level of education. Higher education seems to display the lowest number of female administrators. Research to date has indicated certain differences between male and female elementary-secondary school administrators. Yet, overwhelming differences in leadership effectiveness were difficult to demonstrate. While some variations in the leadership function could be correlated with sex, overall performance of males and females was similar.

PAGE 44

38 Females in higher education administration are found clustered in the mid-management levels or below. Recent researchers have found that it remains difficult for women to be employed as administrators. Those who are administrators are expected to possess the same leadership and personal qualities as their male counterparts. Yet few are in decision making positions. With affirmative action programs in progress, discrimination should be declining. It appears that an understanding of females in higher education administration is vital. The literature reviewed here reveals that further study is necessary to this aspect of higher education administration. Recent legislation and current studies that have been completed emphasize the urgency to place women in leadership positions. This acknowledgement clearly applies to higher education as well as to other realms. Such declarations make it imperative that research be expanded.

PAGE 45

CHAPTER III A DESCRIPTION OF THE UNIVERSITY AND THE SAMPLE POPULATION This chapter provides background relative to the University of Florida and a report of the responses to the professional career interviews. The University of Florida This study was carried out at the University of Florida, one of nine universities in the Florida State University System. The University is under direct supervision of the Board of Regents, a group of nine citizens appointed by the governor for terms of one to nine years. The Board of Regents nominates the president of the University of Florida, as the university's chief executive officer. He is appointed by the State Board of Education. The president has veto power over all actions of committees, college faculties, councils and the University Senate. The University is located in the northern center of Florida at Gainesville. Historically, it is a combined State University and land-grant college. In 1905, the Florida legislature established the University of Florida for men and placed it under the direction of the Board of Control. In 1947 the University was made coeducational. 39

PAGE 46

40 Currently, the University is composed of 16 colleges and two schools. It is unique as all these programs are located on a single campus. The campus also includes more than 40 centers, bureaus and institutes. Academic units are: the Colleges of Agriculture, Architecture and Fine Arts, Arts and Sciences, Business Administration, Dentistry, Education, Engineering, Health Related Professions, Journalism and Communications, Law, Medicine, Nursing, Pharmacy, Physical Education, Health and Recreation, and Veterinary Medicine, along with the University College, School of Forest Resources and Conservation and the Graduate School. As of 1975, the University employed 2,693 faculty members, 5,217 in career service and 289 in administrative and professional positions. Nine percent of the academic staff were classified in the administration, student services, libraries category. An analysis of the University faculty showed 2,242 male faculty members and 280 female faculty members. More than 90 percent of the students attending the University came from within the state. Enrollment figures at the University for 1974-75 were 28,332 total students (17,749 men and 10,503 women), including 4,353 graduate students. The University has conferred more than 98,000 degrees since its founding. In 1973-74 it awarded: 4,445 Bachelors, 393 D.D.'s, 71 M.D.'s, 1,177 Masters, 227 Ph.D.'s, 49 Ed.D.'s, 5 Engineering Specialists, and 71 Education Specialists for a total of 6,435 degrees.

PAGE 47

41 Graduate study has had a phenomenal growth since its beginning at the University. In 1930, 33 degrees were awarded in twelve fields. In 1940, 66 degrees were awarded in 16 fields. In 1971-72 the total number of graduate degrees awarded was 1,636 in more than 90 fields. The proportion of doctoral degrees has also increased. In 1950, 18 Ph.D.'s and 5 Ed.D.'s were awarded. In 1973-74 the total was 224 Ph.D.'s and 49 Ed.D's. The University conducts research in nearly all fields of knowledge. Through its competition for sponsored research and training funds the University acquired over 34 million dollars in grants and contracts in fiscal 1973. The University of Florida is the largest state supported institution of higher learning in Florida. In relation to this study, it should be pointed out that although it is the state's oldest institution, it did not admit women students until 1947. Thus much of its history is based on its background as a public men's university. This information on the University provides a background for the conduct of this study. Administrators' Professional Career Interviews A portion of the information received from the professional career interviews will be included in later chapters. The results reported in this section are predominantly discussed later in relation to their impact on the study's conclusions. The administrators interviewed ranged in age from 32 years to 64 years. A majority of the males and females were in the age range 40-50 years. All of the males interviewed were married.

PAGE 48

42 The females were evenly divided between married, single, and divorced. All o£ the married females had children, as did all the males. A majority of the males were 35-40 years when they first became an administrator. The females tended to be older when; they acquired their first administrative position, a majority were 40-45 years. There was a contrast between the sexes in their total years of administrative experience. A majority of the females had had one to three years in administration. A majority of the males had had four to six years in administration. The females interviewed tended to be relatively new to their present position. A majority had held their current post for less than three years. A majority of the males had a slightly higher average. A majority of the male administrators had held their present position for two to four years. The administrators reported heavy work loads. A majority of both the males and females averaged 50-60 hours of work per week. One male and one female explained that their total work effort was over 70 hours per week. Five females and five males stated they worked an average of 65 hours per week. A large majority of the administrators hold doctorate degrees. Almost half of both the males and females hold a degree in education. A majority of the remaining administrators hold Ph.D.'s in various disciplines. More females than males interviewed hold Ed.D.'s. More males than females interviewed hold Ph.D.'s in education.

PAGE 49

43 Both male and female administrators tended to belong to a number of professional associations. Each administrator reported membership in six to ten organizations. Male administrators tended to publish more frequently than female administrators. A majority of the males reported three to four publications per year.A majority of the females cited one to two publications per year. Additional answers and discussion from the administrators' interviews will be discussed in Chapters IV and V. Summary The University of Florida was the location for this research. All administrators and associates surveyed are employed at this University which is the largest state supported institution of higher learning in Florida. It was originally established as a university for men and became coeducational in 1947. Its 1975 enrollment exceeded 28,000 students . A review of the professional career interviews included comparisons of male and female administrators. Responses were similar in categories of age, work loads, degrees, and organizational membership. Male administrators reported more years in administration and more years in their present position than the female administrators. Males also cited more publications per year than the females. Further conclusions and discussion of the interviews are included in Chapter IV and V.

PAGE 50

CHAPTER IV FINDINGS The analysis of the data in accordance with the procedures set forth in Chapter I is reported in this chapter. Data for analysis were drawn from the administration of the Leader Behavior Description Questionnaire-Form XII . The chapter also relates the findings of the personal interviews held with the male and female administrators. The L . B . D .Q . -Form XII was administered to all participants in the study; administrators, superordinates , and subordinates. Fisher's exact test and chi square were used to answer the twelve questions concerning the differences between male and female administrators' leadership behavior. The significance level was set at .05. Each question was answered on the basis of the ratings of male and female administrators by three groups-subordinates , superordinates and the administrators. Two computer programs were run, using the Statistical Package for the Social Services -Version 6.00. The first computer analysis was used to identify the mean scores of all the administrators on each of the twelve constructs. Administrators were assigned to category 1 or category 2 as follows: Category 1 contained all scores equal to the mean or higher. Category 44

PAGE 51

45 2 contained all scores below the mean. Male and female administrators were compared by group and tests of statistical significance for differences were performed by use of chi square and Fisher's exact test in the second computer analysis. The comparison was based on a 2 x 2 test using male administrators-female administrators and category 1 category 2. Summary of the L.B.D.Q.--XII Findings Of 108 questionnaires distributed, fifty-eight were returned, yielding a 551 total rate of return. There were 35 respondents (64%) to the female administrator evaluation and 23 respondents (42%) to the male administrator evaluation. There were 20 subordinate respondents, 18 superordinate respondents, and 20 administrator respondents. The total respondents by category were as follows: Respondents to Male Administrator Evaluation: Subordinates 7 Superordinates 8 Administrators 8 Respondents to Female Administrator Evaluation: Subordinates 13 Superordinates 10 Administrators 12 Statistical analysis of the L.B.D.Q. -XII results were applied to each of the twelve constructs and interpreted using chi square and Fisher's exact test. Tests were run for comparisons on the basis of the administrators' sex and for comparisons by level of the respondent. The results are reviewed by each of the twelve constructs.

PAGE 52

46 Construct 1: Representation There was no significant difference at the .05 level in the ratings as reported by the subordinates, superordinates and administrators between the perceived degree to which male and female university administrators speak and act as the representative of the group. Table 1.1 reports the results of the subordinate responses for the first construct. It shows that the difference between the subordinate responses was not significant at the .05 level. Table 1.2 reports the results of the superordinate responses for the first construct. It indicates that the difference between the superordinate responses was not significant at the .05 level. Table 1.3 reports the administrator responses for the first construct. It shows that the difference between the administrator responses was not significant at the .05 level. The difference in the categories of responses made by the three group levels was not significant at the .05 level. Table 1.4 reports this result. According to these responses, subordinates, superordinates, and administrators similarly perceived the degree to which male and female university administrators speak and act as representative of the group.

PAGE 53

47 Table 1.1 Number of Subordinate Responses by Category and Sex on Construct 1: Representation (N=20) SEX TOTAL CATEGORY 1 -Within Mean 2-Below Mean or Above Male Adm. 7 5 2 Female Adm. 13 6 7 Total 20 11 9 Fisher's Exact Test = 0.27245 Table 1.2 Number of Superordinate Responses by Category and Sex on Construct 1: Representation (N=18) SEX TOTAL CATEGORY 1 -Within Mean 2-Below Mean or Above Male Adm. 8 5 3 Female Adm. 10 4 6 Total 18 9 9 Fisher's Exact Test = 0.31859 Table 1.3 Number of Administrator Responses by Category and Sex on Construct 1: Representation (N=20) SEX TOTAL CATEGORY 1-Within Mean 2-Below Mean or Above Male Adm. 8 5 3 Female Adm. 12 8 4 Total 20 13 7 Fisher's Exact Test = 0.74923

PAGE 54

48 Table 1.4 Number of Responses by Category and Group on Construct 1: Representation (N=58) GROUP , TOTAL CATEGORY 1-Within Mean 2-Below Mean or Above Subordinate 20 11 9 Superordinate 18 9 9 Administrator 20 13 7 Total 58 53 25 Chi Square = 0.91394 Significance = 0.6352 Construct 2: Demand Reconciliation There was no significant difference at the .05 level in the ratings as reported by the subordinates, superordinates , and administrators between the perceived degree to which male and female university administrators reconcile conflicting demands and reduce disorder to the system. Table 2.1 reports the results of the subordinate responses for the second construct. It shows that the difference between the subordinate responses was not significant at the .05 level. Table 2.2 reports the results of the superordinate responses for the second construct. It indicates that the difference between the superordinate responses was not significant at the .05 level. Table 2.3 reports the administrator responses for the second construct. It shows that the difference between the administrator responses was not significant at the .05 level. The difference in the categories of responses made by the three group levels was not significant at the .05 level. Table

PAGE 55

49 2.4 reports this result. According to these responses, subordinates, superordinates and administrators similarly perceived the degree to which male and female university adminis trators reconcile conflicting demands and reduce disorder to the system. Table 2.1 Number of Subordinate Responses by Category and Sex on Construct 2: Demand Reconciliation (N=20) SEX TOTAL CATEGORY 1 -Within Mean or Above 2 -Be low Mean Male Adm. Female Adm. Total 7 13 20 5 7 12 2 6 8 Fisher's Exact Test = 0.39164 Table 2.2 Number of Superordinate Responses by Category and Sex on Construct 2: Demand Reconciliation (N=18) SEX TOTAL CATEGORY Male Adm. Female Adm, Total 10 18 1-Within Mean or Above 5 5 10 2-Below Mean 3 5 8 Fisher's Exact Test = 0.47984

PAGE 56

50 Table 2.3 Number of Administrator Responses by Category and Sex on Construct 2: Demand Reconciliation (N=20) SEX TOTAL CATEGORY 1 -Within Mean or Above 2 -Below Mean Male Adm. 8 7 1 Female Adm. 12 8 4 Total 20 15 5 Fisher's Exact Test = 0 . 30650 Table 2.4 Number o£ Responses by Category and Group on Construct 2: Demand Reconciliation (N=58) GROUP TOTAL CATEGORY 1 -Within Mean or Above 2-Below Mean Subordinate 20 12 8 Superordinate 18 10 8 Administrator 20 15 5 Total 58 37 21 Chi Square = Significance = 1. 74091 0.4188 Construct 3: Tolerance for Uncertainty There was no significant difference at the .05 level in the ratings as reported by the subordinates, superordinates and administrators between the perceived degree to which male and female university administrators are able to tolerate uncertainty and postponement without anxiety and upset. Table 3.1 reports the results of the subordinate responses for the third construct.

PAGE 57

51 It shows that the difference between the subordinate responses was not significant at the .05 level. Table 3.2 reports the results of the superordinate responses for the third construct. It indicates that the difference between the superordinate responses was not significant at the .05 level. Table 3.3 reports the results for the administrator responses for the third construct. It shows that the difference between the administrator responses was not significant at the .05 level. The difference in the categories of responses made by the three group levels was not significant at the .05 level. Table 3.4 reports this result. According to these responses, subordinates, superordinates , and administrators similarly perceived the degree to which male and female university administrators are able to tolerate uncertainty and postponement without anxiety and upset. Table 3.1 Number of Subordinate Responses by Category and Sex on Construct 3: Tolerance for Uncertainty (N=20) SEX TOTAL CATEGORY ~~ ~~ 1-Within Mean 2-Below Mean or Above Male Adm. 7 4 3 Female Adm. 13 4 9 Total 20 8 12 Fisher's Exact Test = 0.25077

PAGE 58

52 Table 3.2 Number o£ Superordinate REsponses by Category and Sex on Construct 3: Tolerance for Uncertainty (N=18) SEX TOTAL CATEGORY 1-Within Mean 2or Above Below Mean Male Adm. 8 3 5 Female Adm. 10 5 5 Total 18 8 10 Fisher's Exact Test = 0.47984 Table 3.3 Number of Administrator Responses by Category and Sex on Construct 3: Tolerance for Uncertainty (N=20) SEX TOTAL CATEGORY 1 -Within Mean or Above 2Below Mean Male Adm. 8 6 2 Female Adm. 12 6 6 Total 20 12 8 Fisher's Exact Test = 0.25961 Table 3.4 Number of Responses by Construct 3: Tolerance Category and Group on for Uncertainty (N=58) GROUP TOTAL CATEGORY 1-Within Mean 2-Below or Above Mean Subordinate 20 8 12 Superordinate 18 8 10 Administrator 20 12 8 Total 58 28 30 Chi Square Significance = = 1.75534 = 0.4157

PAGE 59

53 Construct 4: Persuasiveness There was no significant difference at the .05 level in the ratings as reported by the subordinates, superordinates , and administrators between the perceived degree to which male and female university administrators use persuasion and argument effectively and exhibit strong convictions. Table 4.1 reports the results of the subordinate responses for the fourth construct. It shows that the difference between the subordinate responses was not significant at the .05 level. Table 4.2 reports the results of the superordinate responses for the fourth construct. It indicates that the difference between the superordinate responses was not significant at the .05 level. Table 4.3 reports the results of the administrator responses for the fourth construct. It shows that the difference between the administrator responses was not significant at the .05 level. The difference in the categories of responses for the three group levels was not significant at the .05 level. Table 4.4 reports this result. According to these responses, subordinates, superordinates, and administrators similarly perceived the degree to which male and female university administrators use persuasion and argument effectively and exhibit strong convictions.

PAGE 60

54 Table 4.1 Number of Subordinate Responses by Category and Sex on Construct 4: Persuasiveness (N=20) SEX . TOTAL CATEGORY 1 -Within Mean T^^eTow'Tiean or Above Male Adm. 7 2 5 Female Adm. 13 5 8 Total 20 7 13 Fisher's Exact Test = 0.52574 Table 4.2 Number of Superordinate Responses by Category and Sex on Construct 4: Persuasiveness (N=18) SEX TOTAL CATEGORY 1-Within Mean 2-Below Mean or Above Male Adm. 8 4 4 Female Adm. 10 6 4 Total 18 10 8 Fisher's Exact Test = 0.52016 Table 4.3 Number of Administrator Responses by Category and Sex on Construct 4: Persuasiveness (N=20) SEX TOTAL CATEGORY 1-Within Mean 2-Below Mean or Above Male Adm. 8 6 2 Female Adm. 12 7 5 Total 20 13 7 Fisher's Exact Test = 0.39164

PAGE 61

55 Table 4.4 Number of Responses by Category and Group on Construct 4: Persuasiveness (N=58) GROUP TOTAL ^ CATEGORY 1-Within Mean 2 -Below Mean or Above Subordinate 20 7 13 Superordinate 18 10 8 Administrator 20 13 7 Total 58 30 28 Chi Square = 3.75772 Significance = 0.1528 Construct 5: Initiation of Structure There was no significant difference at the .05 level in the ratings as reported by the superordinates and administrators between the perceived degree to which male and female university administrators clearly define their own role and let followers know what is expected of them as reported by the superordinates and the administrators. However, there was a significant difference of .04427 reported by the subordinates' responses. The subordinate responses to female administrators rated the female administrators significantly higher than the subordinate respondents to the male administrators rated the male administrators. Table 5.1 reports the results of the subordinate responses for the fifth construct. It shows that there was a significant difference at the .05 level between the subordinate responses. Table 5.2 reports the results of the superordinate responses

PAGE 62

56 for the fifth construct. It indicates that the difference between the superordinate responses was not significant at the .05 level. Table 5.3 reports the administrator responses for the fifth construct. It shows that the difference between the administrator responses was not significant at the .05 level. The difference in the categories of responses made by the three group levels was not significant at the .05 level. Table 5.4 reports this result. According to these responses, superordinates and administrators similarly perceived the degree to which male and female university administrators clearly define their own role and let followers know what is expected of them. There was a significant difference perceived by the subordinates' responses . Table 5.1 Number of Subordinate Responses by Category and Sex on Construct 5: Initiation of Structure (N=20) SEX TOTAL CATEGORY 1-Within Mean or Above 2Below Mean Male Adm. 7 0 7 Female Adm. 13 6 7 Total 20 6 14 Fisher's Exact Test 0.04427* .05

PAGE 63

57 Table 5.2 Number of Superordinate Responses by Category and Sex on Construct 5: Initiation of Structure (N=18) SEX TOTAL CATEGORY 1 -Within Mean or Above 2 Below Mean Male Adm, 8 4 4 Female Adm. 10 6 4 Total 18 10 8 Fisher's Exact Test = 0.52016 Table 5.3 Number of Administrator Responses by Category and Sex on Construct 5: Initiation of Structure (N=20) SEX TOTAL CATEGORY 1 -Within Mean or Above 2 -Below Mean Male Adm. 8 3 5 Female Adm. 12 5 7 Total 20 8 12 Fisher's Exact Text = 0. 61189 Table 5.4 Number of Responses by Catej Construct 5: Initiation of jory and Group on Structure (N=58) GROUP TOTAL CATEGORY 1-Within Mean or Above 2 -Below Mean Subordinate 20 6 14 Superordinate 18 10 8 Administrator 20 8 12 Total 58 24 34 Chi Square = 2.57462 Significance --^ 0.2760

PAGE 64

58 Construct 6: Tolerance of Freedom There was no significant difference at the .05 level in the ratings as reported by the subordinates, superordinates , and administrators between the perceived degree to which male and female university administrators allow followers scope for initiative, decision and action. Table 6.1 reports the results of the subordinate responses for the sixth construct. It shows that the difference between the subordinate res-onses was not significant at the .05 level. Table 6.2 reports the results of the superordinate responses for the sixth construct. It indicates that the difference between the superordinate responses was not significant at the .05 level. Table 6.3 reports the results for the administrator responses for the sixth construct. It shows that the difference between the administrator responses was not significant at the .05 level. There was a significant difference of .0258 reported in the categories of responses made by the three group levels. The administrators ranked themselves lower in ratings than did the subordinate and superordinate respondents. Table 6.4 reports this result. According to these responses, subordinates and superordinates similarly perceived the degree to which male and female university administrators allow followers scope for initiative, decision and action. As a group, the administrators perceived themselves lower on this construct.

PAGE 65

59 Table 6.1 Number of Subordinate Responses by Category and Sex on Construct 6: Tolerance of Freedom (N=20) SEX TOTAL CATEGORY 1 -Within Mean 2 -Below Mean or Above Male Adm. 74 3 Female Adm. 13 8 5 Total 20 12 ' 8 Fisher's Exact Test = 0.74923 Table 6.2 Number of Superordinate Responses by Category and Sex on Construct 6: Tolerance of Freedom (N=18) SEX TOTAL CATEGORY 1 -Within Mean 2 -Below Mean or Above Male Adm. 8 6 2 Female Adm. 10 7 3 Total 18 13 5 Fisher's Exact Test = 0.61765 Table 6.3 Number of Construct Administrator 6: Tolerance Responses by Category and Sex on of Freedom (N=20) SEX TOTAL CATEGORY 1-Within Mean 2-Below Mean or Above Male Adm. 8 3 5 Female Adm. 12 3 9 Total 20 6 14 Fisher's Exact Test = 0.86275

PAGE 66

60 Table 6.4 Number of Responses by Category and Group on Construct 6: Tolerance of Freedom (N=58) HROTIP TOTA T vj r\ 1 \j kj i\ 1 1 -Within Mean or Above 2-Below Mean Subordinate 20 12 8 SuDerordinate 18 13 5 Administrator 20 6 14 Total 58 31 17 Chi Square = Significance = 7.31448 0. 0258* •* *£ = .05. Construct 7: Role Assumption There was no significant difference at the .05 level in the ratings as reported by the subordinates , superordinates , and administrators between the perceived degree to which male and female university administrators actively exercise the leadership role rather than surrender leadership to others. Table 7.1 reports the results of the subordinate responses for the seventh construct. It shows that the difference between the subordinate responses was not significant at the .05 level. Table 7.2 reports the results of the superordinate responses for the seventh construct. It indicates that the difference between the superordinate responses was not significant at the .05 level. Table 7.3 reports the administrator responses for the seventh construct. It shows that the difference between the administrator responses was not significant at the .05 level.

PAGE 67

61 The difference in the categories of responses made by the three group levels was not significant at the ,05 level. Table 7.4 reports this result. According to these responses, subordinates, superordinates and administrators similarly perceived the degree to which male and female university administrators actively exercise the leadership role rather than surrender leadership to others. Table 7.1 Number of Subordinate Responses by Category and Sex on Construct 7: Role Assumption (N=20) SEX TOTAL CATEGORY 1-Within Mean 2 -Below Mean or Ab ove Male Adm. 7 2 Female Adm. 13 5 Total 20 7 Fisher's Exact Test = 0.52574 Table 7.2 Number of Superordinate Responses by Category Construct 7: Role Assumption (N=18) and Sex on SEX TOTAL CATEGORY 1 -Within Mean or Above 2Below Mean Male Adm. 8 4 4 Female Adm. 10 4 6 Total 18 8 10 5 8 13 Fisher's Exact Test = 0.81578

PAGE 68

62 Table 7.3 Number of Administrator Responses by Category and Sex on Construct 7: Role Assumption (N=20) *J Lj J\. TOTAT CATEGORY 1 -Within Mean or Above 2Below Mean Male Adm. 8 6 2 Female Adm. 12 6 6 Total .20 12 8 Fisher's Exact Test 0.63003 Table 7.4 Number o£ Responses by Category and Group on Construct 7: Role Assumption (N=58) GROUP TOTAL CATEGORY 1-Within Mean or Above 2Below Mean Subordinate 20 7 13 Superordinate 18 8 10 Administrator 20 12 8 Total 58 27 31 Chi Square = Significance = 2.55853 0. 2782 Construct 8: Consideration There was no significant difference at the .05 level in the ratings as reported by the subordinates, superordinates and administrators between the perceived degree to which male and female university administrators regard the comfort, well-being, status and contribution of followers. Table 8.1 reports the results of the subordinate responses for the eighth construct. It shows that the difference between the subordinate responses

PAGE 69

63 was not significant at the .05 level. Table 8.2 reports the results of the superordinate responses for the eighth construct. It indicates that the difference between the superordinate responses was not significant at the .05 level. Table 8.3 reports the results of the administrator responses for the eighth construct. It shows that the difference between the administrator responses was not significant at the .05 level. The difference in the categories of responses made by the three group levels was not significant at the .05 level. Table 8.4 reports this result. According to these responses, subordinates, superordinates and administrators similarly perceived the degree to which male and female university administrators regard the comfort, well-being, status and contribution of followers. Table 8.1 Number of Subordinate Responses by Category and Sex on Construct 8: Consideration (N= 20) SEX TOTAL CATEGORY 1-Within Mean or Above 2Below Mean Male Adm. 7 3 4 Female Adm. 13 6 7 Total 20 9 11 Fisher's Exact Test = 0.63003

PAGE 70

64 Table 8.2 Number of Superordinate Responses by Category and Sex on P f"i"n Q T r1 p T ?1 1* 1 n n r\I= 18) SEX TOTAL CATEGORY 1 -Within Mean or Above 2Below Mean Male Adm. 8 3 5 Female Adm. 10 4 6 Total 18 7 11 Fisher's Exact Test = 0.64781 Table 8.3 Number of Administrator Respons-s by Category and Sex on Construct 8: Consideration (N=20) SEX TOTAL CATEGORY 1-Within Mean or Above 2Below Mean Male Adm. 8 2 6 Female Adm. 12 5 7 Total 20 7 13 Fisher's Exact Test = 0.39164 Table 8.4 Number of Responses by Category and Group on Construct 8: Consideration (N=58) GROUP TOTAL CATEGORY 1-Within Mean 2-Below Mean or Above Subordinate 20 9 11 Superordinate 18 7 11 Administrator 20 7 13 Total 58 23 35 Chi Square Significance = 0.42429 0 .8088

PAGE 71

65 Construct 9: Production Emphasis There was no significant difference at the .05 level in the ratings as reported by the subordinates, superordinates and administrators between the perceived degree to which male and female university administrators apply pressure for productive output. Table 9.1 reports the results of the subordinate responses for the ninth construct. It shows that the difference between the subordinate responses was not significant at the .05 level. Table 9.2 reports the results of the superordinate responses for the ninth construct. It indicates that the difference between the superordinate responses was not significant at the .05 level. Table 9.3 reports the results of the administrator responses for the ninth construct. It shows that the difference between the administrator responses was not significant at the .05 level. The difference in the categories of responses made by the three group levels was not significant at the .05 level. Table 9.4 reports this result. According to these responses, subordinates, superordinates and administrators similarly perceived the degree to which male and female university administrators apply pressure for productive output.

PAGE 72

66 Table 9.1 Number of Subordinate Responses by Category and Sex on Construct 9: Production Emphasis (N=20) SEX TOTA T CATEGORY 1 -Within Mean or Above 2Below Mean Male Adm. 7 5 2 . Female Adm. 13 8 5 Total 20 13 7 Fisher's Exact Test = 0.52574 Table 9.2 Number of Superordinate Responses by Category and Sex on Construct 9: Production Emphasis (N=18) SEX TOTAL CATEGORY 1-Within Mean or Above 2Below Mean Male Adm. 8 4 4 Female Adm. 10 5 5 Total 18 9 9 Fisher's Exact Test = 0.68141 Table 9.3 Number of Administrator Responses by Category and Sex on Construct 9: Production Emphasis CN=20) SEX TOTAL CATEGORY Male Adm. Female Adm. Total 1-Within Mean or Above 2 -Below Mean 8 12 20 3 6 9 5 6 11 Fisher's Exact Test = 0.46499

PAGE 73

67 Table 9.4 Number of Responses by Category and Group on Construct 9: Production Emphasis (N=58) GR OUP TOTAL CATEGORY 1 -Within Mean 2 -Below Mean or Above Subordinate 20 13 7 Superordinate 18 9 9 Administrator 20 9 11 Total 58 31 27 Chi Square = 1.73238 Significance == 0.4206 Construct 10: Predictive Accuracy There was no significant difference at the .05 level in the ratings as reported by the subordinates, superordinates and administrators between the perceived degree to which male and female university administrators exhibit foresight and ability to predict outcomes accurately. Table 10.1 reports the results of the subordinate responses for the tenth construct. It shows that the difference between the subordinate responses was not significant at the .05 level. Table 10.2 reports the results of the superordinate responses. It indicates that the difference between the superordinate responses was not significant at the .05 level. Table 10.3 reports the results of the administrator responses for the tenth construct. It shows that the difference between the administrator responses was not significant at the .05 level. The difference in the categories of responses made by the three group levels was not significant at the .05 level. Table

PAGE 74

68 10.4 reports this result. According to these responses, subordinates, superordinates , and administrators similarly perceived the degree to which male and female university adminiS' trators exhibit foresight and ability to predict outcomes accurately. Table 10.1 Number of Subordinate Responses by Category and Sex in Construct 10: Predictive Accuracy (N=20) SEX TOTAL CATEGORY 1 -Within Mean 2 -Below Mean or Above Male Adm. 7 1 6 Female Adm. 13 2 11 Total 20 3 17 Fisher's Exact Test = 0.72982 Table 10.2 Number of Superordinate Responses by Category and Sex on Construct 10: Predictive Accuracy (N=18) SEX TOTAL CATEGORY 1 -Within Mean or Above 2Below Mean Male Adm. 8 4 4 Female Adm. 10 5 5 Total 18 9 9 Fisher's Exact Test = 0.68141

PAGE 75

69 Table 10.3 Number o£ Administrator Responses by Category and Sex on Construct 10: Predictive Accuracy (N^20) SFX TOTAL CATEGORY 1 -Within Mean or Above 2-Below Mean Male Adm. 8 3 5 Female Adm. 12 4 8 Total 20 7 13 Fisher's Exact Test = 0 . 74923 Table 10.4 Number of Responses by Category and Group on Construct 10: Predictive Accuracy (N=58) GROUP TOTAL CATEGORY 1-Within Mean or Above 2-Below Mean Subordinate 20 3 17 Superordinate 18 9 9 Administrator 20 7 13 Total 58 19 39 Chi Square = 5.33819 Significance = 0.0693 Construct 11: Integration There was no significant difference at the .05 level in the ratings as reported by the subordinates, superordinates , and administrators between the perceived degree to which male and female university administrators maintain a closely-knit organization and resolve intermember conflict. Table 11.1 reports the results of the subordinate responses for the eleventh construct.

PAGE 76

70 It shows that the difference between the subordinate responses was not significant at the .05 level. Table 11.2 reports the results of the superordinate responses for the eleventh construct. It indicates that the difference between the superordinate responses was not significant at the .05 level. Table 11.3 reports the results of the administrator responses for the eleventh construct. It shows that the difference between the administrator responses was not significant at the .05 level. The difference in the categories of responses made by the three group levels was not significant at the .05 level. Table 11.4 reports this result. According to these responses, subordinates, superordinates , and administrators similarly perceived the degree to which male and female university administrators maintain a closely-knit organization and resolve inter-member conflict . Table 11.1 Number of Subordinate Responses by Category and Sex on Construct 11: Integration (N=^20) SEX TOTAL CATEGORY 1 -Within Mean 2-Below Mean or Above • 5 2 6 7 11 9 Fisher's Exact Text = .027245 Male Adm. 7 Female Adm. 13 Total 20

PAGE 77

71 Table 11.2 Number of Superordinate Responses by Category and Sex on Construct 11: Integration (N=18) SEX TOTAL CATEGORY 1 -Within Mean or Above 2Below Mean Male Adm. 8 5 3 Female Adm. 10 7 3 Total 18 12 6 Fisher's Exact Test = 0.79864 Table 11.3 Number of Administrator Responses by Category and Sex on Construct 11: Integration (N=20) SEX TOTAL CATEGORY 1-Within Mean or Above 2Below Mean Male Adm. 8 4 4 Female Adm. 12 8 4 Total 20 12 8 Fisher's Exact Test = 0.38881 Table 11.4 Number of Responses by Category and Group on Construct 11: Integration (N=58) SEX TOTAL CATEGORY 1 -Within Mean or Above 2Below Mean Subordinate 20 11 9 Superordinate 18 12 6 Administrator 20 12 8 Total 58 35 . 23 Chi Square = 0.54037 Significance 0.7632

PAGE 78

72 Construct 12: Superior Orientation There was no significant difference at the .05 level in the ratings as reported by the subordinates, superordinates and administrators between the perceived degree to which male and female university administrators maintain cordial relations with superiors, have influence over them, and are striving for higher status. Table 12.1 reports the results of the subordinate responses for the twelfth construct. It shows that the difference between the subordinate responses was not significant at the .05 level. Table 12,2 reports the results of the superordinate responses for the twelfth construct. It indicates that the difference between the superordinate responses was not significant at the .05 level. Table 12.3 reports the results of the administrator responses for the twelfth construct. It shows that the difference between the administrator responses was not significant at the .05 level . The difference in the categories of responses made by the three group levels was not significant at the .05 level. Table 12.4 reports this result. According to these responses, subordinates, superordinates, and administrators similarly perceived the degree to which male and female university administrators maintain cordial relations with superiors, have influence over them, and are striving for higher status.

PAGE 79

1 73 Table 12.1 Number of Subordinate Responses by Category and Sex on Construct 12: Superior Orientation (N=20) SEX ^ TOTAL CATEGORY 1 -Within Mean 2 -Below Mean or Above 3 4 6 7 9 11 Fisher's Exact Test = 0.63003 Male Adm. 7 Female Adm. 13 Total 20 Table 12.2 Number of Superordinate Responses by Category and Sex on Construct 12: Superior Orientation (N=18) SEX TOTAL CATEGORY 1 -Within Mean 2 -Below Mean or Above Male Adm. 8 6 2 Female Adm. 10 5 5 Total 18 11 7 Fisher's Exact Test = 0.27828 Table 12.3 Number of Administrator Construct 12: Superior Responses by Category and Orientation (N=20) Sex on SEX TOTAL CATEGORY 1 -Within Mean or Above 2Below Mean Male Adm, 8 5 3 Female Adm. 12 5 7 Total 20 10 10 Fisher's Exact Test = 0.52496

PAGE 80

74 Table 12.4 Number of Responses by Category and Group on Construct 12: Superior Orientation (N=58) GROUP TOTAL CATEGORY 1 -Within Mean 2 -Below Mean or Above Subordinate 20 9 11 Superordinate 18 11 7 Administrator 20 10 10 Total 58 _30 28 Chi Square = 1.02114 Significance = 0.6002 Summary of the Administrators' Professional Career Interviews Information and insights gained from the interview portion of research are discussed in this part of the chapter. A copy of the interview guide is included as Appendix B. Three career subtopics emerged from the interviews: career development, career orientation, and career aspirations. Each of these is reviewed separately, with particular emphasis on a comparison of male and female responses. The major purpose of the interview process was to have the administrators assess their administrative involvement, which includes leadership functions. Answers acquired through the interview process are usually not of a quantitative nature. Therefore, discussion of the interviews revolves around overviews and general observations. The emphasis is on contrasting males and females as administrators.

PAGE 81

75 Career Development Female and male patterns o£ career development emerged quite differently. Educational background tended to yield the same result. The degrees held by males and females were similar. However, the acquisition of these degrees proved to be different in nature. It appeared that the key contrast between males and females was marriage and family status. Males continued their education without interruption, through their highest degree. Females took longer to complete their highest degree, with time off for child rearing and family responsibilities. Exceptions to the female pattern were the unmarried female administrators. Divorced female administrators also had a discernable pattern. They returned for advanced degrees and sought administrative positions after being divorced. Marriage status seemed to have no bearing on the career development of male administrators. A majority of the males were married, with families. Males and females were in similar age brackets. Most clustered around the mid-forty to mid-fifty range. Experience in administration was varied. Noteworthy was the fact that the males tended to climb the administrative ladder in a distinguishable manner. They had prior administrative experience, at least mid-management in status. A majority of the females were found to be in their first administrative role. Most had been teaching faculty, appointed "out of the ranks" to the administrative position.

PAGE 82

76 Career Orientation Males and females tended to give different reasons for assuming their current administrative position. This difference could be attributed in part to prior career background. Both male and female administrators saw their positions as advancements, providing opportunities for new experiences. The males defined the administrative role as an opportunity to become a change agent, to have an impact on decision-making. Females perceived the administrative position as one with lots of interaction. Females were eager to use their position to work with people; to be responsive to staff and students' needs. Both males and females pinpointed problem-solving as a positive aspect of the administrative assignments. Job duties as described by both males and females seemed to demand responsibilities for all of the roles mentioned above. The duties which were listed were numerous and varied. The important observation to make is that job responsibilities did not seem to differ between male and female administrators. A "typical day" was impossible for any of the administrators to describe. Females felt they were overburdened with university committee assignments as the "token female." However, males were found also to have heavy committee involvements. The administrators estimated their "average" work week to be 50-60 hours. The females often pointed out that 10-20 of these hours were sometimes spent with paper work taken home. Females with families stressed their tendency to work at home

PAGE 83

77 during evenings. Males stated that they worked straight through a day until a late dinner, or returned to the office during evening or weekend hours. Single females seemed to utilize a variety of work settings. Based on these responses, it could be suggested that male and female administrators work similar total hours. When questioned about job completion males and females responded almost identically. Both categories felt that requirements for paper shuffling were burdensome. Most also mentioned "red tape" as a thwarting factor. Both sexes cited budget cuts and the dollar crunch as limitations on their planning and programming. Female administrators did not note disrespect or disregard due to their sex. Males did not mention being hampered by affirmative action demands. Self-satisfaction and job-satisfaction were high for both sexes. The male and female administrators represented themselves as individuals who had matched their personal fulfillment needs with an appropriate employment environment. Both sexes seemed equally involved in their careers and equally pleased with their current job expectations. A majority of the administrators felt that their positions provided prestige, recognition, and visibility. Expectations for these factors may vary according to sex, but differences were not apparent in the interviews. Prestige did not appear to be deemed as vital as was a means to get things done. As long as superordinates and subordinates were aware of the adminis-

PAGE 84

78 trators' needs and methods, the administrator voiced positive feelings concerning competency and reputation. Males tended to emphasize superordinate interaction, females discussed both superordinate and subordinate, equally. Career Aspirations A majority of the female administrators interviewed were in their first administrative position. Few of them had been specifically seeking administrative careers. While they declared a satisfaction and challenge with their current role, they did not appear ambitious to assume a higher level administrative position. Additional power of top level administration did not appear to be alluring to most of the females. The males interviewed were more likely to picture themselves as upwardmobile, in their careers. Their long-range career objectives included a commitment to professional advancement. In fact, many of the male administrators claimed to have assumed their present position as a stepping stone to higher level administration. They described their current job as a testing ground for their administrative growth and development. Most of the administrators stated that they regretted their lack of classroom involvement. A majority of the females stressed that teaching was their primary reason for entering a higher education career. About one-half of the females said that they would eventually like to return to fulltirae teaching. The males noted teaching, but were less interested in teaching fulltime.

PAGE 85

79 Both tenured and nontenured administrators felt some obligation to publish and maintain professional organization memberships. Males seemed to put a greater emphasis on their publication efforts. However, females aspiring to administrative advancements felt publishing was a definite requirement for promotion. Current salary levels appeared acceptable to both sexes. A majority of the administrators felt university pay scales were somewhat below effort extended. Yet, neither males nor females felt that the current salaries were discriminatory on the basis of sex. The female administrators interviewed appeared eager to see additional females in university administration. The males interviewed did not voice opposition to the idea. Some of the males gave it avid support. Summary Data from the Leader Behavior Description Questionnaire XII were analyzed. Each of the twelve constructs were examined with the significant difference level set at .05. Subordinate, superordinate , and administrator responses were reviewed. Cross comparisons were also tabulated for the three groups. Findings from this procedure indicated that sex was not a significant factor in the leader behavior of University of Florida administrators for eleven of the twelve constructs. There was a significant difference reported by the subordinate responses to Contrust 5: Initiation of Structure.

PAGE 86

80 Information from the personal interviews with University of Florida administrators was examined in three areas: career development, career orientation, and career aspirations. Responses to the interview questions tended to show some contrasts in male and female patterns in these three areas .

PAGE 87

CHAPTER V DISCUSSION, CONCLUSIONS AND RECOMMENDATIONS Discussion Response Rate to the L.B.D.Q. --XII Twelve questions were addressed in this study, focusing on the twelve constructs of the L.B.D.Q. --XII . There were more respondents to female administrators than there were respondents to male administrators. Several factors should be considered in accounting for this difference. The female respondents were contacted earlier in the academic quarter. Therefore, these respondents had a greater amount of time to return the questionnaire. There was little difference in total number of returns of respondents, by group level, to male administrators. The largest number of respondents by group level were the subordinate respondents to the female administrators. About onethird of the female administrators reported that they had only one subordinate. They may have been in very close contact with the one individual. It could also be suggested that the subordinates might have had more time to devote to the study, while superordinates gave it a lower priority. The respondents averaged 30 minutes to complete the questionnaire. The difference in response rate was not large when compared within 81

PAGE 88

82 male and female respondent divisions. The greater distinction was between male and female respondents, as discussed earlier . The L.B.D.Q.--XII Findings Answers to the 100 items of the L.B.D.Q. --XII provided no significant difference at the .05 level between the way male and female administrators' leader behavior was perceived on ten constructs. Reasons for this similarity should be discussed. First, the survey may not have been able to discriminate between male and female university administrators. The leader skills and behavior may be extremely similar between all university administrators. It should be suggested that a certain level of leader behavior must be observable before an individual becomes a university administrator. It could be further contended that once an individual performs as an administrator within the university setting, he or she acquires characteristics which are alike in a majority of. administrative actions. This is not to suggest that all university administrators are identical, but it is to interpret the L.B.D.Q. --XII findings as an indication that there are few distinguishable differences in the leader behavior of male and female administrators at the University of Florida. Second, this study used sex of the administrator as the basis for comparing the administrators' leader behavior. Characteristics other than sex provide a wider range of responses on the L.B.D.Q. --XI I . Previous experience, educational background, and career aspirations might all be used as categories

PAGE 89

83 for comparison. However, it should be noted here that differences in these categories could often be correlated with the sex of the administrator. Third, similarly it might be suggested that administrators are perceived as administrators, not as male administrators nor female administrators. Thus subordinates see their "superiors" as university administrators, not expecting nor interpretating their behavior on the basis of the administrators' sex. Superordinates perceive the administrators as their subordinates and respond to their leader behavior without regard to a "male" subordinate or a "female" subordinate category. Also the administrators observe themselves as university administrators, without the identity of "male" administrator nor "female" administrator. An overview explanation of the fact that there were few significant differences reported by this study could be the inability of this particular instrument to identify distinguishable contrasts between the leader behavior of male and female university administrators. On construct five, the subordinate respondents to female administrators were significantly different than the subordinate respondents to the male administrators. The subordinate respondents perceived female administrators as having a higher degree to which they clearly define their own role and let followers know what is expected of them. (A significant difference was not perceived by the superordinates nor the administrators for the same question, however.)

PAGE 90

84 This measurable difference can be related to the personal interviews with the administrators. The female administrators emphasized their interest in their subordinates more often than did the male administrators. Although this emphasis was not apparent in the administrators' responses to this construct of the L.B.D.Q. --XII , apparently it was noticed by the subordinates . Perhaps since the female administrators were more likely to be relatively new to their position as well as to the administrative role, the subordinates noted that they were more concise about that role in communication with the subordinate. It is possible that the females were more recently in roles similar to their subordinate than were the male administrators, thus they functioned more definitively by the subordinates' standards . On construct six, a significant difference was found in the perceived degree to which the three groups rated the administrators' tendency to allow followers scope for initiative, decision and action. There was no significant difference in the ratings of male administrators compared to female administrators, by each group. The subordinate and superordinate respondents rated the administrators significantly higher than the administrators rated themselves. This observable difference could be explained by the difficulty of self -evaluation in respect to action towards others. Although all the constructs dealt with interaction, this one

PAGE 91

85 pinpointed the administrators perceiving themselves as more "confining" to their followers than did the followers. The superordinates' perception of the interaction agreed with the subordinates'. Perhaps, the administrators saw themselves as less tolerant of freedom because they wished to stress their administrative role for action as more vital than the subordinates'. On the other hand, the subordinates may have interpreted the concept as a positive one. They may have rated the administrators high for granting them a level of independent thinking and doing. The superordinates may have interpreted the behavior in the same manner. Administrators' Professional Career Interviews The administrators' professional career interviews focused on the descriptive data that was collected. The administrators answered questions concerning their career development, orientation, and aspirations. The male and female administrators inter viewed appeared to be similar in many respects. Observations of the interview discussions are general because of the nature of the interviews. Exceptions to each generality can be assumed Patterns attributable to the administrators' sex did emerge. These patterns appeared to revolve around the contrasting career development of males and females. As of 1975, females appeared to have taken a different route to attain administrative positions. While the role and scope of female and made administrators seemed to be similar, sometimes the personal purposes and pursuits appeared to vary according to sex.

PAGE 92

86 It is important to note that both sexes claimed to perceive the administrative responsibilities to include leadership functions. Neither sex felt that sex was correlated to their leadership performance. While this information provided certain insights, it is difficult to relate and compare it with the L.B.D.Q. --XII . The conclusion is similar to the administrators' evaluations of themselves in the L.B.D.Q.-XI I . However, contrasts between the results of the L.B.D.Q. --XI I and the professional career interviews should be made before further conclusions can be drawn. The L.B.D.Q. --XII survey found that there are no significant differences between the leader behavior of male and female university administrators, on the basis of ten leadership behavior dimensions, as reported by superordinates , subordinates, and the administrators. Conclusions from the interviews were based solely on the administrators' observations of themselves. At the same time, portions of the personal interviews pertained to background and performance of the administrators not directly correlated to leader behavior. Responses to construct 5 seemed to partially support the observation that L.B.D.Q. --XI I responses and individual interviews were somewhat different. While superordinates and administrators found no significant difference between the way male and female administrators relate to followers, the subordinate respondents fall into the significant difference category. As the interviews had shown, subordinates reported that female

PAGE 93

87 administrators rate higher in terms o£ clearly defining their own roles and letting followers know what is expected of them. However, constructs 6 and 8 also dealt with questions of administrators' relation to followers and no significant difference between male and female administrators was found in these two constructs. On the other hand, although male administrators vocalized greater attention to superordinates and more anxiety over achieving a higher level administrative position, the L.B.D.Q. --XII found no significant difference between the sexes in this area. Construct 12 addressed itself to this particular dimension of leader behavior. The specific dimensions measured by the L.B.D.Q. -XI I appear similar to the topics discussed in the interviews with the administrators. Two of the dimensions previously discussed yielded results which contrast to a limited extent with the administrators' verbally reported perceptions. Nonetheless, it is difficult to generalize concerning the personal interview responses. Variations should be noted and thus, comparison of both research techniques should be considered. Conclusions of the studies reported in the literature review chapter should be contrasted with the findings of this research. Fecher (1972) observed that females have little influence on policy making decisions at colleges and universities. Yet, when interviewed. University of Florida female administrators claimed a "fair share" of participation in decision making.

PAGE 94

88 Fecher also found that marriage was neither an advantage nor a disadvantage to the female administrator. However, University of Florida female administrators emphasized that marriage had altered their career development. They reported that they have postponed their education, or interrupted it, for family purposes. Once the females became administrators, marriage status did not appear to affect the job performance. One of Pfiffner's (1972) major conclusions was that female administrators display a high interest in working with others. Similar findings appeared in the interview portion of this study. Female administrators, more often than male administrators, pointed out their interest and devotion to interaction with others as part of the administrative role. Arter's (1972) national survey of female administrators indicated that a majority of the women plan to remain in administration. The University of Florida female administrators did not seem to fit this pattern. A majority of them voiced an interest to "return" to teaching or half-time administrative responsibilities. In addition, Morsink (1969) reported that women in administration are striving for higher status. Yet, a majority of University of Florida female administrators pointed out that they preferred to remain in their present position or become fulltime teachers, A minority expressed an interest in acquiring higher level administrative positions.

PAGE 95

89 Contrary to Harris' (1970) research which concluded that females were in traditionally female roles in universities, University of Florida female administrators were in positions of responsibility. However, University of Florida female administrators are a lovi? percentage of total administrators at the University. They are also a low percentage of total female faculty and staff. These figures may be the type Harris used to support her study. Carroll (1972) suggested that females do not seek administrative posts. Interview information from the Florida female administrators did not thoroughly explain this phenomenon, but it did appear to agree with the observation. A majority of the female administrators noted that they had not actively sought their present position. General observations from the personal interviews are summarized as follows: 1. Males and females follow different routes to becoming administrators. The male administrators tended to complete their education without interruption and immediately pursue administrative positions. They appeared to move up a "management occupational ladder." Female administrators' patterns were much less discernible. The female's educational background and trends were diverse. Few females had completed their entire education without interruption. A majority of the females seemed to have become administrators without pursuing that role in preparatory education. Management positions seemed much more likely

PAGE 96

90 to be "unplanned promotions" for the female administrators. Thus, while the males seemed to be following a line design of administrative advancements, the females became administrators without a master plan for starting and continuing up the administrative ladder. 2. Both male and female administrators pointed out that their positions were time demanding , but extremely rewarding. Males tended to emphasize the opportunity to act as a change agent. Females more often mentioned the inter-action with personnel and students and the ability to respond to the individual's needs. Both sexes felt their position had an impact on the functioning of the university. It could be suggested that this difference in role orientation is due to the experience prior to the present position. Males were more likely to have had administrative training and experience. Females were more likely to have had teaching training and experience. 3. Long range career goals differed between the males and females. Males were more likely to aspire to higher management levels. Females were more likely to be satisfied with their current position or expressed a desire to return to fulltime teaching. A need to aspire to top-level administrative positions appears to be lower in the females surveyed than in the males. This difference might be attributed to the fact that a minority of the females had originally selected administration as a career goal, while a majority of the males had chosen administration as their life's vocation.

PAGE 97

91 Conclusions Male and female university administrators, their superordinates , and their subordinates were surveyed on 12 dimensions of leader behavior, measured by the L . B . D .Q . -XI I . Evidence from the survey instrument would tend to reinforce the conclusion that there is no significant difference in the perceived leadership behavior of University of Florida male and female administrators to the extent that the 12 constructs of the L.B.D.Q. --XII report leadership behavior. Answers to the questions posed at the outset of the study can be provided on the basis of the data: 1. There was a significant difference in the way subordinates perceived male and female university administrators in regard to one construct; the degree to which an individual clearly defines his or her own role and lets followers know what is expected of them. There was no significant difference as measured in terms of the other eleven constructs in the way subordinates perceived male and female university administrators' leadership behavior. 2. There was no significant difference in the way superordinates perceived male and female university administrators' leadership behavior. 3. There was no significant difference in the way male and female university administrators perceived their own behavior.

PAGE 98

92 It was pointed out in the individual personal interviews with the male and female administrators that there were differences in career development between the sexes, however, the central attitudes and approaches to the administrative role were similar. Males indicated in the interviews that they were more eager to obtain top level administrative positions than did females . Most of the contrasts between the male and female interview responses would appear to be the result of traditional female-male educational and career patterns. It should be noted that the females presently in university administrative positions are a small minority, but their background is mainly rooted in the pre-woman's liberation era. Consequently, it could be assumed that female and male administrators could each be aligned with certain traditional career patterns based on sex. Perhaps more important to the readers of this study is the way in which females perform as university administrators. The results of the instrument utilized in this research report sex is not a significant factor in overall leader behavior of University of Florida administrators. The findings of this study would lend support to the following implications: 1. Sex of the applicants should not be a factor in evaluating them as leaders for the university administrative setting.

PAGE 99

93 2. Leadership behavior and performance (as measured by the L.B.D.Q. --XII ) of a university administrator was not found to be significantly related to the administrator's sex. Recommendations for Future Study 1. Studies of male and female administrative leadership behavior should be conducted at other universities and colleges. Community college and university based studies should be compared. 2. Additional aspects of male and female administrators' backgrounds and behaviors should be researched. These might include decision making experience, career planning, and other categories of leader behavior, 3. Current educational programs preparing individuals for administrative roles should be reviewed. 4. Comprehensive studies regarding employment and promotion of individuals in higher education should be studied.

PAGE 100

APPENDIX A AN INTRODUCTION: THE LEADER BEHAVIOR DESCRIPTION QUESTIONNAIREFORiM XII This Leader Behavior Description Questionnaire-XI I (L.B.D.Q. --XII) is being used as part of a study whose purpose is to compare the leadership behavior of female university administrators and male university administrators. This instrument has been designed to provide information relative to this study. It is not a test and in no way a measure of the administrator's ability as an administrator. Your cooperation in filling out this survey is appreciated. The L.B.D.Q. --XII will be used to study the perception of the administrator's leader behavior as seen by the administrator's immediate superior (superordinate) , randomly selected members of the administrator's staff (subordinates), and the administrator's own perception of leader behavior. The data will be reported mainly in the form of statistical summaries. In all cases, the answers will be held in strict confidence. Therefore, you are urged to respond in a sincere and open manner. Thank you for your cooperation and participation. Barbara Keener, Doctoral Student Institute of Higher Education University of Florida 94

PAGE 101

95 LEADER BEHAVIOR DESCRIPTION QUESTIONNAIRE-Form XII Originated by staff members of The Ohio State Leadership Studies and revised by the Bureau of Business Research Purpose of the Questionnaire On the following pages is a list of items that may be used to describe the behavior of the administrator. Each item describes a specific kind of behavior, but does not ask you to judge whether the behavior is desirable or undesirable. Although some items may appear similar, they express differences that are important in the description of leadership. Each item should be considered as a separate description. This is not a test of ability or consistency in making answers. Its only purpose is to make it possible for you to describe, as accurately as you can, the behavior of the selected administrator. NOTE: The term, "group," as employed in the following items, refers to a department, division, or other unit or organization that is supervised by the person being described. In some cases the "group" may represent only one or two subordinates . The term "members," refers to all the people in the unit of organization that is supervised by the person being described. Please attempt to answer all the questions. Bureau of Business Research College of Commerce and Administration The Ohio State University Columbus, Ohio Copyright 1962

PAGE 102

96. DIRECTIONS: a. READ each item carefully. b. THINK about how frequently the leader engages in the behavior described by the item. c. DECIDE whether he/she (A) always, (B) often, (C) occasionally, (D) seldom or (E) never acts as described by the item. d. DRAW A CIRCLE around one of the five letters (A B C D E) following the item to show the answer you have selected. A » Always B » Often C » Occasionally D = Seldom E = Never e. MARK your answers as shoivn in the examples below. • Example: The administrator often acts as described . . A B C D E Example: The administrator never acts as described . . A B C D E Example: The administrator occasionally acts as described ABCDE 1. The administrator acts as the spokesman of the group. ABCDE 2. The administrator waits patiently for the results of a decision ABCDE 3. The administrator makes pep talks to stimulate the group ABCDE 4» The administrator lets group members know what is expected of them ABCDE S.. The adnjin i
PAGE 103

97.. A » Always B Often C » Occasionally D = Seldom E = Never 7. The administrator is friendly and approachable ... A B C D E 8. The administrator encourages overtime work A B C D E 9. The administrator makes accurate decision A B C D E 10. The administrator gets along well with the people above him/her ABCDE 11. The administrator publicizes the activities of the group ABCDE 12. The administrator becomes anxious when he/she cannot find out what is coming next ABCDE 13. The administrator's arguments are convincing .... ABCDE 14. The administrator encourages the use of uniform procedures ABCDE 15. The administrator permits the members to use their own judgment in solving problems ABCDE 16. The administrator fails to take necessary action . . ABCDE 17. The administrator does little things to make it pleasant to be a member of the group ABCDE 18. The administrator stresses being ahead of competing groups ABCDE 19. The administrator keeps the group working together as a team ABCDE 20. The administrator keeps the group in good standing with higher authority ABCDE 21. The administrator speaks as the representative of the group ABCDE 22. The administrator accepts defeat in stride ABCDE

PAGE 104

93 A * Always B = Often C = Occasionally D = Seldom E = Never 23. The administrator argues persuasively his/her point of view ABCDE 24. The administrator tries out his/her ideas in the group ABCDE 25. The administrator encourages initiative in the group members ABCDE 26. The administrator lets other persons take away his/her leadership in the group ABCDE 27. The administrator puts suggestions made by the group into operation ABCDE 28. The administrator needles members for greater effort ABCDE 29. The administrator seems able to predict what is coming next ABCDE 30. The administrator is working hard for a promosion . .ABCDE 31. The administrator speaks for the group when visitors are present ABCDE 32. The administrator accepts delays without becoming upset ABCDE 33. The administrator is a very persuasive talker . . . .ABCDE 34. The administrator makes his/her attitudes clear to the group ABCDE 35. The administrator lets the members do their work the way they think best ABCDE 36. The administrator lets some members take advantage of him/her. .' ABCDE

PAGE 105

99 A = Always B « Often C " Occasionally D = Seldom E =• Never 37. The administrator treats all group members as his/her equals ABCDE 38. The administrator keeps the work moving at a rapid pace ABCDE 39. The administrator settles conflicts when they occur in the group ABCDE 40. The administrator's superiors act favorably on most of his/her suggestions ABCDE 41. The administrator represents the group at outside meetings ••• ABCDE 42. The administrator becomes anxious when waiting for new developments ABCDE 43. The administrator is very skillful in an argument . .ABCDE 44. The administrator decides what shall be done and how it shall be done. ABCDE 45. The administrator assigns a task, then lets the members handle it ABCDE 46. The administrator is the leader of the group in name only ABCDE 47. The administrator gives advance notice of changes . .ABCDE 48. The administrator pushes for increased production . .ABCDE 49. Things usually turn out as he/she predicts ABCDE 50. The administrator enjoys the privileges of his/her position ..ABCDE 51. The administrator handles complex problems efficiently ABCDE

PAGE 106

100 A = Always B = Often C = Occasionally D = Seldom E = Never 52. The administrator is able to tolerate postponement and uncertainty ABCDE 53. The administrator is not a very convincing talker ABCDE 54. The administrator assigns group members to particular tasks ABCDE 55. The administrator turns the members loose on a job, and lets them go to it ABCDE 56. The administrator backs down when he/she ought to stand firm ABCDE 57. The administrator keeps to himself /herself ABCDE 58. The administrator asks the members to work harder. . ABCDE 59. The administrator is accurate in preducting the trend of events ABCDE 60. The administrator gets his/her superiors to act for the welfare of the group members ABCDE 61. The administrator gets swamped by details ABCDE 62. The administrator can wait just so long, then blows up A B-C D E 63. The administrator speaks from a strong inner conviction ABCDE 64.. The administrator makes sure that his/her part in the group is understood by the group members. . . ABCDE 65. The administrator is reluctant to allow the members any freedom of action ABCDE 66. The administrator lets some members have authority that he/she should keep ABCDE 67. The administrator looks out for the personal welfare of the group members ABCDE

PAGE 107

101 A Always B » Often C = Occasionally D Seldom E » Never 68. The administrator permits the members to take it easy in their work A B C D E 69. The administrator sees to it that the work of the group is coordinated ABCDE 70. The administrator's work carries weight with his/her superiors ABCDE 71. The administrator gets things all tangled up ABCDE 72. The administrator remains calm when uncertain about coming events ABCDE 73. The administrator is an inspiring talker ABCDE 74. The administrator schedules the work to be done . . .ABCDE 75. The administrator allows the group a high degree of initiative ABCDE 76. The administrator takes full charge when emergencies arise ABCDE 77. The administrator is willing to make changes ABCDE 78. The administrator drives hard when there is a job to be done ABCDE 79. The administrator helps group members settle their differences ABCDE 80. The administrator gets what he/she asks for from his/her superiors ABCDE 81. The administrator can reduce a madhouse to system and order ABCDE 82. The administrator is able to delay action until the proper time occurs ABCDE 83. The administrator persuades others that his/her ideas are to their advantage ABCDE

PAGE 108

102A = Always B » Often C « Occasionally D = Seldom E = Never 84. The administrator maintains definite standards of performance ABODE 85. The administrator trusts the members to exercise good judgment ABODE 86. The administrator overcomes attempts made to challenge his/her leadership ABODE 87. The administrator refuses to explain his/ her actions ABODE 88. The administrator urges the group to beat its previous record ABODE 89. The administrator anticipates problems and plans for them ABODE 90. The administrator is working his/her way to the top ABODE 91. The administrator gets confused when too many demands are made of him/her ABODE 92., The administrator worries about the outcome of any new procedure ABODE 93The administrator can inspire enthusiasm for a project ABODE 94. The administrator asks that group members follow standard rules and regulations ABODE 95. The administrator permits the group to set its own pace ABODE 96. The administrator is easily recognized as the leader of the group. . ABODE 97. The administrator acts without consulting the group ABODE

PAGE 109

103. A * Always B = Often C = Occasionally D = Seldom E « Never 98. The administrator keeps the group working up to capacity ABCDE 99. The administrator maintains a closely knit group ABCDE 100. The administrator maintains cordial relations with superiors ABCDE

PAGE 110

APPENDIX B ADMINISTRATOR'S PROFESSIONAL CAREER INTERVIEW GUIDE 1. General Information. A. What is your age? B. What is your marital status? C. Do you have children? 2. What was your age when you took your first administrative position? 3. How many years of experience do you have in administration? 4. How long have you held your present position? 5. What prior administrative experiences have you had? 6. What were your reasons for assuming your present position? 7. What is your educational background? 8. What degrees do you hold? 9. What are your present job responsibilities? 10. What do you feel are your road blocks to job completion? 11. What degree of self satisfaction do you get through your job? 12. What degree of career satisfaction do you get through your job? • 104

PAGE 111

105 13. What type of prestige, recognition and/or visibility do you get through your job? 14. What is your average weekly work load, in hours? 15. What are your long range career objectives? 16. What were your major reasons for entering your field? 17. Do you feel your salary is equitable? 18. To what professional organizations do you belong? 19. Do you feel a need to publish? How often do you publish? What have you published? 20. Do you have any additional coiimients concerning your career as an administrator?

PAGE 112

APPENDIX C LETTER OF AUTHORIZATION THE OHIO STATE UNIVERSITY October 16, 1974 Mrs. Barbara Keener 410 Victory Drive Apartment 115 Tallahassee, Florida 32301 Dear Mrs. Keener: You have our permission to use the Leader Behavior Description Questionnaire in your research. Research using the L.B.D.Q. is reviewed in the Hand book of Leadership by Ralph M. Stogdill, published by the Free Press. The address of the publisher is enclosed. Sincerely , /s/ Ralph Stogdill Ralph M. Stogdill Professor of Management Sciences RMS: je Enclosure 106

PAGE 113

REFERENCES American Association for Higher Education. Campus sex bias found widespread by AAUIV. College and Universi ty Bulletin, December 15 , 1970. ' American Association of American Colleges. Sex discrimination provisions concerning students and employees as contained" in the Higher Education Act of 1972 . Washington, D.C.: Project on the Status and Education of Women, 1972. American Association of School Administrators. Equal leadership opportunities for women . Washington, D.C.: Educational Research Service, 1972. American Association of School Administrators. Women superin tendents of schools . Washington, D.C.: Educational Research Service, 1972. Argyris, C. Executive leadership: An appraisal of a manager in action . New York: Archon, 1967. Argyris, C. Integrating the individual and the organization . New York: Wiley, 1964. Arter, M. H. The role of women in administration in state universities and land grant colleges . (Doctoral dissertation, Arizona State University) Ann Arbor, Michigan: University Microfilms , 1972, No. 72-13-006. Austin, H. S. The woman doctorate in America . New York: Russell Sage Foundation, 1969. Barter, A. S. The status of women in school administration. Educational Horizons , 1959 , 37_, 7275. Bartky, J. A. Administration as educational leadership . Palo Alto, California: Stanford University Press, 1956. Bartol, K. Closed loop--women as leaders. The Masters in Business Administration . December 2, 1972, 6-10. Bass, B. M. , Drusell, J., § Alexander, R. Male manager's attitude toward working women. American Behavioral Scientist , 1971, IS (2), 221-236^ 107

PAGE 114

108 Bell, D. Meritocracy and equality. The Public Interest , Fall, 1972 , 29^, 29-68. Bern, S. L. , ^ Bern, D. J. Training the woman to know her place : The power of a nonconscious ideology . Palo Alto, California : Stanford University Press, 1 9 71. Bern, S. L. , ^ Bern, D. J. Training the woman to know her place : The social antecedent of woman in the world of work . Harrisburg, Pennsylvania: State Department of Public Instruction, 1971. Bennet, W, Institutional barriers to the utilization of women in top management . (Doctoral dissertation, University of Florida) Ann Arbor, Michigan: University Microfilms, 1964, No. 64-25-225. Bennis, W. G. Changing organization . New York: McGraw-Hill, 1966. Bennis, W. G. Leadership theory and administrative behavior: The problem of authority. Administrative Science Quarterly , 1959, 4, 259-260. Bernard, J. S. Academic women . University Park, Pennsylvania: Pennsylvania State University Press, 1964. Bernard, L. L. An introduction to social psychology . New York: Holt, 1926. Brown, A. F. A perpetual taxonomy of the effective rated teacher. Journal of Experiential Education , 1966 , 35^, 1-10. Brown, A. F. Reactions to leadership. Educational Administra tors Quarterly , 1967, 3, 62-73. Bulwick, H. , 5 Elicks, S. Affirmative action for women: Myth and reality . Berkeley"! Institute of Business and Economic Research, University of California, 1972. Bureau of National Affairs. ASPA-BNA survey: Employment of women . Washington, D. CTl Bureau of National Affairs, 1970. Bureau of National Affairs. Personnel policies forum-survey No. 96 V'/omen and minorities in management and in personnel management . V/ashington, D. C: Bureau of National Affairs, 1971. Burke, W. W. Leadership behavior as a function of the leader, the follower and the situation. Journal of Pers onal ity, 1965, 33, 60-81. "

PAGE 115

109 Burns, D. Women in educational administration: A study of leadership in California public schools . (Doctoral dissertation, University of Oregon) Ann Arbor, Michigan: University Microfilms, 1964, No. 64-12-150. Euros, 0. (Ed.) The seventh mental measurement yearbook . Highlands Park, New Jersey: Gryon Press, 1972, Vol. 1. Campbell, R. F., § Faber, C. F. Administrative behavior: Theory and research. Review of Educational Re search, 1961, 31_, 352-367. Campbell, R. , § Gregg, R. T. Administrative behavior in education . New York: Harper, 1957. Carnegie Commission on Higher Education (Ed.) Opportunities for women in highe r education. New York: McGraw-Hill, T973: Carroll, M. A. Women in administration in higher education. Contemporary Education , 1972, 43, 214-218. Case, C. M. Leadership and conjuncture. Sociology and So cial Research, 1933, 17, 510-513. Cattail, R. B. New concepts for measuring leadership in terms of group syntality. Human Relations , 1951, 4, 161-184. Change Magazine (Ed.) Women on campus . New York: Change Magazine Press, 1975. Charters, W. W. , Jr. Teacher perceptions of administrator behavior . (Cooperative Research Project No. 929) Washington, D. C: Office of Education, 1964. Chase, F., § Cuba, E. Administrative roles and behavior. Review of Educational Research, 1955 , 2S_, 281-298 . Citizen's Advisory Council on the Status of Women. Women in 1970 . Washington, D. C: U. S. Government Printing Office, March, 1971. Citizen's Advisory Council on the Status of Women. Need for studies of sex discrimination in public schoo ls. Washington, D. C: Department of Labor, 1972 . Citizen's Advisory Council on the Status of Women. Women inl97iWashington, D. C: U. S. Government Printing Office, January, 1972. Cohen, A. C. Women and higher education: Recommendation^: for change. Phi Delta Kappan , 1973, 50(3), 164-167.

PAGE 116

110 Cook, A. H. Sex discrimination at the university. A.A.U.P . Bulletin , 1972 , 58_(3) , 279-82. Cook, E. V. Leadership behavior of elementary school prin cipals and the organizational climate of the schools which they administer . (Doctoral dissertation, Rutgers State University) Ann Arbor, Michigan: University Microfilms, 1965, No. 66-6769. Cribbin, J. J. Effective managerial leadership . New York: American Management Association, 1972. Croft, J. C. Dogmatism and perceptions of leader behavior. Educational Administrator Quarterly , 1965 , 60-71. Crowley, J. Facts and factions about working women explored. Institute for Social Research (University of Michigan) , 1972, 4-5. Day, D. R. Description of male and female behavior by male and female subordinates . Urbana: University of Illinois, Department of Industrial Administration, 1968. DeLamater, J., § Fidell, L. On the status of women. American Behavioral Scientists , 1971, 15_(2) , 163-171. Dowd, J. Control in human societies . New York: AppletonCentury, 1936. Ebel, R. (Ed.) Encyclopedia of educational research . (44th ed.) London: MacMillan Company, 1969. Equal Employment Opportunity Commission. Guidelines on dis crimination because of sex . Washington, D. C: U. S. Government Printing Office, 1969 (Chapter 14, Part 160A, as amended) . Equal Employment Opportunity Commission. Toward job equality for women . Washington, D. C: U. S. Government Printing Office, 1969. Erickson, D. A. The school administrator. Review of Educational Research , 1967 , 37_, 417-432. Ernst, R. J. An investigation of the relationship between selected characteristics of principals and organizational climates of elementary schools . (Doctoral dissertation, Florida State University) Ann Arbor, Michigan: University Microfilms, 1965, No. 65-15-460. Eurich, N. , Tompkins, P., 5 Eddy, E. The education of women. Saturday Review , 1963, 18_, 61-70.

PAGE 117

Ill Fecher, A. R. Career patterns of women in college and university administration . (Doctoral dissertation, Indiana University) Ann Arbor, Michigan: University Microfilms, 1972, No. 72-21-800. Ferber, M. , 5 Loeb, J. Sex as predictive of salary and status on a university faculty. Journal of Educational Measure ment , 1971 , 8 (4) , 235-44 . Fiedler, F. E. A theory of leadership effectiveness . New York: McGraw-Hill, 1967. Fiedler, F. E. Style or circumstance: The leadership enigma. In W. R. Lassey (Ed.), Leadership and Social Change . Iowa City: University Associates, 1971, 275-284. Firestone, S. The dialectic of sex, the case for feminist resolution . New'York: William Morrow, 1970. Gardner, H. R. Women administrators in higher education in Illinois : A study of current career patterns . (Doctoral dissertation , Indiana University) Ann Arbor, Michigan: University Microfilms, 1966, No. 66-12-655. Gentry, H. W. Patterns of behavioral characteristics exhibited by school administrators . (Doctoral dissertation. The University of Tennessee) Ann Arbor, Michigan: University Microfilms, 1957, No. 22-667. Getzels, J. W. A psychological framework for the study of educational administration. Harvard Educational Review , 1952 , 235-246. Gibb, C. A. Leadership. In G. Lindzey Handbook of Social Psychology . Cambridge, Mass.: Addison-Wesley , 1954. Gott, C. M. A study of perceptions and expectations of leader ship behavior of principals of Texas large senior high schools ^ (Doctoral dissertation. University of Texas) Ann Arbor, Michigan: University Microfilms,1966, No. 66-14-340. Griffith, D. E. The nature and meaning of theory in Behavioral Science and Educational Administration, The Sixty-third Yearbook of the National Society for the Study of" Education . (Part II) Chicago: The University of Chicago Press, 1964, 95-118. Grobman, H. G., § Hines, V. A. What makes a good principal? The Bulletin of the National Association of Secondary School Principals , 1966, 40(223), 5-6.

PAGE 118

112 Gross, N., 5 Harriott, R. Staff leadership in public schools : A sociological inquiry" New York; John Wiley and Sons, 1965. Gross, N.', § Trask, A. E. Men and women as elementary school principals (Cooperative Research Project No. 8 53) . Washington, D. C: U. S. Office of Education, 1964. Halpin, A. W. The leadership behavior of school superintendents . Columbus, Ohio: Ohio State University, 1956. Halpin, A. W. Theory and research in administration . Toronto, Canada: Collier MacMillan, 1966. Harris, A. S. The second sex in academia. A.A.U.P. Bullet in, 1970, 56, 283-295. Hedges, J. N. Women at work-women workers and manpower demands in the 1970's. Monthly Labor Review , 1970, 19-29. Hemphill, J. K. Leadership behavior associated with the administrative reputations of college departments. Journal of Educational Psychology , 1955 , 4_6, 385-401. Hemphill, J. K., Griffiths, D. E., ^ Frederiksen, N. Adminis trative performance and personality . New York: Bureau of Publications of Columbia University, 1962. Holden, L. W. Administrative roles in secondary education as identified by secondary principals and teachers . (Doctoral dissertation, Stanford University) Ann Arbor, Michigan : University Microfilms, 1959, No. 59-1424. Homans , G. C. The human group . New York: Harcourt, Bruce, 1950. Horner, M. A bright woman is caught in a double bind: In achievement-oriented situations she worries not only about failure but also success. Psychology Today , 1971, 46, 36-38. Hoyle, J. A. Who shall be principal--a man or a woman? The National Elementary Principal , 1969, £8(3), 23-24. Hoyle, J. R. Problemattack behavior and its relationship to the sex, prior teaching experience, and college prepara tion of selected elementary school principals . (Doctoral dissertation, Texas A § M Lfniversity) Ann Arbor, Michigan: University Microfilms, 1967, No. 67-9789. Jacobs, J. W. Leader behavior of the secondary school principal. National Association of Secondary School Principals, The Bulletin , 1965, 49, 13-17.

PAGE 119

113 Jacobs, T. 0. Leadership and exchange in formal organizations . Alexandria, Va. : Human Resources Research Organization, 1971. Johnson, D. L. Ms. administrators, where are they: The Schoo l Administrator (American Association of Schools Administrators Newsletter), 1972, 19. Kaufman, H. The status of women in administration in selected institutions of higher education in the United States . (Doctoral dissertation. New York University) Ann Arbor, Michigan: University Microfilms, 1961, No. 62-1443. Kaufman, S. Few women get positions of power in academe, survey discloses. The Chronicle of Higher Education , 1970, _5(10) , 14. Kendall, M. G. Table of random numbers . London: Cambridge University Press., 1960. Kimbrough, R. B. Administering elementary schools: Concepts and practices . New York: The MacMillan Company, 1968. Knezevich, S. J. Administration of public education . New York: Harper, 1969. Koontz, E. The best kept secret of the past 5,000 years: Women are ready for leadership in education . Bloomington, Indiana: Phi Delta Kappa Educational Foundation , 197 2. Krohn, B. The puzzling case of the missing Ms. Nation's Schools and Colleges , 1974 , 58^, 32-38. LaPuma, B. D. A study of attitudes toward the employment of women in higher education as revealed in the literature of higher education . (Doctoral dissertation. New York University) Ann Arbor, Michigan: University Microfilms, 1972, No. 72-20, 641. Levitin, T., Quinn, R. P., § Stiles, G. L. Sex discrimination against the American working women. American Behavi oral Scientist , 1971, 1^(2), 237-254. Levitt, M. J. Political attitudes of American women: A study of the effects of work and education on their political role . (Doctoral dissertation, University of Maryland) Ann Arbor, Michigan: University Microfilms, 1965, No. 66-933. Likert, R. The human organization . New York: McGraw-Hill, 1967.

PAGE 120

114 Longstreth, C. A. An analysis of the perceptions of the leader ship behavior o"f male and female secondary school prin cipals in Florida . (Doctoral dissertation, University of Miami) Coral Gables, Florida, 1973. Mann, D. Administrative-community-school relationships in New York State. Report for the New York State Commission on the Quality, Cost and Financing ot Elementary and Secondary Education, 1971 . McGregor, D. An analysis of leadership. In W. R. Lassey (Ed.), Leadership and Social Change . Iowa City: University Associates, 1971, 17-25. McGregor, D. Leadership and motivation . Cambridge, Mass.: MIT Press, 1966. Megargee, E. Influence of sex roles on the manifestation of leadership. Journal of Applied Psychology , 1969, 55(5) , 337-382. Mitchell, J. M. , § Starr, R. A regional approach for analyzing the recruitment of academic women. American Behavioral Scientist , 1971, 15(2), 183-203. Moore, L. L. The relationship of academic groups membership to the motive to avoid success in women . (Doctoral dissertation. University of Virginia) Ann Arbor, Michigan: University Microfilms, 1971. No. 72-7220. Mors ink, H. M. A comparative study of the leader behavior of men and women secondary school principals . (Doctoral dissertation. University of Michigan) Ann Arbor, Michigan: University Microfilms, 1966, No. 27-2793A. Morsink, H. M. Leader behavior of men and women principals. The Bulletin (National Association of Secondary School Principals), 1970, 5£(347) , 80-87. Murphy, A. J. A study of the leadership process. American Sociology Review , 1941, 6, 674-687. National Council of Administrative Women in Education of the National Education Association. Women, a significant national resource . Washington, D. C. : National Education Association, 1971. National Education Association Research Division. Professional women in public schools, 1970-71 . Washington, D. C: National Education Association, 1 9 71. National Organization for Women (New York Chapter) . Report on sex bias in the public schools (Rev. ed.) New York : National Organization for Woiiien, 19 72.

PAGE 121

115 Noll, N. L. Opinions of policy-making officials in two-year public educational institutions toward the employment of women administrators . (Doctoral dissertation, Arizona State University) Ann Arbor, Michigan: University Microfilms, 1973, No. 73-20-447. Oltman, R. Campus 1970: Where do women stand? American Association of University Women Journal , 1970, 2^, 14-15. Owens, R. G. Organizational behavior in schools . Englewood Cliffs, New Jersey: Prentice-Hall, 1970. Pfiffner, V. T. Factors associated with women in major adminis trative positions in California community colleges . (Doctoral dissertation. University of Southern California) Ann Arbor, Michigan: University Microfilms, 1972, No. 72-19-716. Prather, J. Why can'twomen be more like men? American Be havioral Scientist , 1971, (2) , 172-182 . Recruitment Leadership and Training Institute. Women in administrative positions in public education . Washington, D. C. : U. S. Office of Education, 1974. Rossi, A. S. Discrimination and demography restrict opportunities for academic women. College and Univ ersity Business, 1970, 48(76), 72-78. Schneider, J. The cultural situation as a condition for the achievement of fame. American Sociology Review, 1937, 2, 480-491. Simpson, L. A. A myth is better than a miss: Men get the edge in academic employment. College and University Business , 1970, £8(76), 68-72. Stodt, M. M. Autonomy and complexity in women teachers in leadership positions . (Doctoral dissertation, Columbia University) Ann Arbor, Michigan: University Microfilms, 1972, No. 72-19-528. Stogdill, R. M. Individual behavior and group achievemen t. New York: Oxford University Press, 1959, Stogdill, R. M. Manual for the Leader Behavior Descriptio n Questionnaire-Form XII, an experimental revision . Columbus, Ohio: Bureau of Business Research, College of Commerce and Administration, Ohio State University, 1963. Stogdill, R. M. A review of research on Leader Behavior Description QuestionnaireForm XlT. Columbus, Ohio : Ohio State University, College of Administrative Science. 1970.

PAGE 122

116 Stogdill, R. M. , 5 Coons, A. E. Leader behavior: Its descrip tion and measurement . Columbus, Ohio: Ohio State University, Bureau of Business Research, College of Commerce and Administration, 1957. Stogdill,' R. M. , Scott, E. L., § Jaynes, W, E. Leadership and role expectations . Columbus, Ohio: Ohio State University Bureau of Business Research, College of Commerce and Administration, 1965. Stogdill, R. M. , 5 Shartle, C. L. Methods in the study of administrative leadership . Columbus, Ohio: Ohio State University , Bureau of Business Research, College of Commerce and Administration, 1955. Stogdill, R. M. , di Shartle, C. L. Patterns of administrative performance . Columbus, Ohio: Ohio State University, Bureau of Business Research, College of Commerce and Administration, 1956. Suelzle, M. A questionnaire-sexism in American schools. Learning: The Magazine for Creative Teaching , 1972, 1(1) , 81-84. Tannenbaura, R. , Weschler, I. R. , 5 Massarik, F. Leadership and organizations: A behavioral science approach . New York: McGraw-Hill, 1961. Taylor, S. S. The attitudes of superintendents and board of education members in Connecticut toward the employment and effectiveness of women as public school administrators . (Doctoral dissertation. University of Connecticut) Ann Arbor, Michigan: University Microfilms , 1971, No. 71-18-452. Tipple, M. E. Attitudes toward the hireability of women for professional administrative positions in public education . (Doctoral dissertation. University of Michigan) Ann Arbor, Michigan: Univers ity Microfilms , 1973, No. 73-12-028. Toporoff, R. Generating role types concerning the occupational participation of women in the twentieth century . (Doctoral dissertation, Washington State University) Ann Arbor, Michigan: University Microfilms, 1972, No. 72-18-494. United States Department of Health, Education, and Welfare. Report of the women's action program , 1972. United States Department of Labor. The myth and reality: Male workers more equal than female workers ? Washington, D. C: Women's Bureau Department of Labor, 1972, 1-3.

PAGE 123

117 United States Equal Employment Opportunity Commission, Toward job equality for women . Washington, D, C; Equal Employment Commission, 1972. University of Florida. The office of academic affairs fact book . Gainesville: University of Florida, February, 1975. University of Florida. The university record of the University of Florida . Gainesville: University ot Florida, March, 1974, No . T'. University of Florida. The university graduate school record . Gainesville: University of Florida, 1973-74. VanMeier, E. J. Leadership behavior of male and female elementary principals . (Doctoral dissertation, Northern Illinois University) Ann Arbor, Michigan: University Microfilms, 1971, No. 71-29, 823. Warwick, E. B. Attitudes towards women in administrative posi tions as related to curricular implementation and change . (Doctoral dissertation. University of Wisconsin) Ann Arbor, Michigan: University Microfilms, 1967, No. 67-9024. Weir, V. J. T. Leadership among administrative women in public education in Nebraska . (Doctoral dissertation. University of Nebraska) Ann Arbor, Michigan: University Microfilms, 1961, No. 62-142. Willower, D. J. Leadership styles and leaders perception of subordinates. Journal of Educational Sociology , 1960, 34 , 58-64. Zimmerman, J. N. The status of women in educational administra tion positions within the central office of public schools . (Doctoral dissertation. Temple University), Ann Arbor, Michigan: University Microfilms, 1971, No. 71-26,538.

PAGE 124

BIOGRAPHICAL SKETCH Barbara Jean Keener received her initial schooling in Springfield, Missouri, graduating from Parkview High School in 1963. She earned her Bachelor of Arts degree in Political Science from The Colorado College in 1967. She received her Master of Arts degree in Speech Communication from the University of Wyoming in 1969, and a Master of Science degree in Student Personnel Counseling from Miami University (Ohio) in 1971. In 1971-72 she served as Assistant to the Dean of Student Affairs at Rollins College. She began her Educational Administration doctoral studies at the University of Florida in the summer of 1972. During her doctoral studies, Ms. Keener held a Kellogg Fellowship for 1972-74 and served as an intern with the Florida State Legislature in 1974-75. As an undergraduate she was active in student government, intercollegiate debate, journalism, and Delta Gamma social sorority. She was a teaching fellow and assistant debate coach while earning her degree at the University of Wyoming, At Miami University she served as Freshman Adviser with the Office of Dean of Women. Ms. Keener holds memberships in various professional and honorary organizations, including Phi Delta Kappa, Kappa Delta Pi and Pi Lambda Theta education honoraries. 118

PAGE 125

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. es L. 'Watrenbarger rofessor of Educati dfial A Chairman Administration I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Arthur Sandeen, Co-Chairman Associate Professor of Educational Administration I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Dorothyj Nevill Assistant Professor of Psychology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Harold C. Riker Professor of Counselor Education

PAGE 126

This dissertation was submitted to the Graduate Faculty of the College of Education and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. March, 1976. Dean, College of Education Dean, Graduate School