Citation
The application of consumer choice theory to study of interdistrict enrollment options

Material Information

Title:
The application of consumer choice theory to study of interdistrict enrollment options
Creator:
Burke, Diane Marie
Place of Publication:
Denver, Colo.
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
x, 140 leaves : ; 29 cm

Thesis/Dissertation Information

Degree:
Doctorate ( Doctor of Philosophy)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
School of Education and Human Development, CU Denver
Degree Disciplines:
Education
Committee Chair:
Murphy, Michael
Committee Members:
Sanders, Nancy
Linn, Robert

Subjects

Subjects / Keywords:
School choice -- Colorado -- El Paso County ( lcsh )
School choice ( fast )
Colorado -- El Paso County ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references.
General Note:
Submitted in partial fulfillment of the requirements for the degree, Doctor of Philosophy, School of Education.
General Note:
School of Education and Human Development
Statement of Responsibility:
by Diane Marie Burke.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
25352786 ( OCLC )
ocm25352786
Classification:
LD1190.E3 1991d .B87 ( lcc )

Full Text
THE APPLICATION OF CONSUMER CHOICE THEORY
TO STUDY OF INTERDISTRICT ENROLLMENT OPTIONS
by
Diane Marie Burke
B.A., Elmhurst College, 1971
M.Ed., College of William and Mary, 1973
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
School of Education


1991 by Diane Marie Burke
All rights reserved.


This thesis for the Doctor of Philosophy
degree by
Diane Marie Burke
has been approved for the
School of
Education
by
/
Robert Linn
Da


Burke, Diane Marie (Ph.D., Education)
The Application of Consumer Choice Theory to Inter-District
Enrollment Options
Thesis directed by Professor Michael Murphy
The purpose of this study was to test the application of
consumer choice theory in the public education setting by
studying the data generated when students in El Paso County,
Colorado, during the 1989-90 school year chose to exercise an
interdistrict enrollment choice option. A basic premise of
consumer choice theory is that consumers make choices to
maximize utility. This study required an assumption about what
educational utility parents and students maximized when they
chose schools. For the purposes of this research, it was assumed
that when students and parents exercised the interdistrict
enrollment option they maximized school quality as measured by
student achievement and socioeconomic status.
In order to determine the validity of the argument, three
hypotheses were developed and tested. One: When students
cross school district boundaries to attend school, the test scores
of the receiving schools are higher than the test scores of the
sending school. Two: When students cross school district
boundaries to attend school, the SES of the receiving school is


higher than the SES of the sending school. Three: When students
cross school district boundaries to attend school, the receiving
schools test scores are higher than the sending schools scores
when the schools populations are controlled for SES. All schools
in El Paso County, Colorado, were assigned two values, one based
on student achievement as measured by standardized test scores
and one based on SES as measured by the percentage of students
on free and reduced lunch. During the 1989-90 school year, 446
students crossed district boundaries to attend school. The
movement of these students across school district boundaries
was traced.
Dependent sample t-tests were run on the data for
hypotheses one and two. The results of the data analysis
indicated that when parents and students acted as consumers,
they did maximize utility as measured by student achievement
and socioeconomic status. Hypotheses one and two were
accepted. Hypothesis three was not tested after the running of a
Pearson-product moment correlation determined the test scores
and SES were inversely related negating the value of controlling
for SES.
The form and content of this abstract are approved. I
recommend its publication.
Signed
v


CONTENTS
Tables......................................... viii
CHAPTER
1. BACKGROUND AND INTRODUCTION...................1
Purpose of the Study..........................11
Methodology...................................13
2. LITERATURE REVIEW...............................15
Current Choice Options Available..............16
External Options........................16
Internal Options........................19
The History of School Choice in the
United States.................................26
Libertarian Arguments Concerning School
Choice........................................32
Research Base Surrounding Choice..............36
Economic Efficiency Arguments Concerning
School Choice.................................42
Excellence Arguments Concerning School
Choice........................................49
Consumer Choice Theory........................57
Student Achievement as Consumer Utility . .60
Outlier Studies.........................61
Case Studies............................62


Program Evaluations.....................64
Socioeconomic Status as Consumer Utility ... 67
Summary.......................................68
3. METHODOLOGY.....................................70
Procedures....................................71
Instrumentation...............................73
Sub-Populations Analyzed...................74
Statistical Tests.............................75
Limitations of the Study ..................76
4. DATA ANALYSIS...................................78
Hypothesis One.............................78
Hypothesis Two.............................93
Hypothesis Three..................... ... 104
Summary of the Findings......................107
5. CONCLUSIONS, WHAT THE FINDINGS MAY MEAN
FOR SCHOOL CHOICE, AND RECOMMENDATIONS
FOR FURTHER RESEARCH.........................113
Conclusions..................................113
What Findings May Mean for School Choice . 123
Recommendations for Further Research . . . 127
BIBLIOGRAPHY..........................................131
vii


TABLES
Table
4.1. Comparison of Sending and Receiving Schools When Student
Cases Were Used.......................................80
4.2. Comparison of Sending and Receiving Schools When Same
Achievement Tests Were Used in Both Schools...........83
4.3. Comparison of Sending and Receving Schools When Different
Achievement Tests Were Used...........................84
4.4. Comparion of Sending and Receiving Schools When No
Tuition Was Paid......................................85
4.5. Comparison of Sending and Receiving Schools When
Partial Tuition Was Paid..............................86
4.6. Comparison of Sending and Receiving School When
Elementary Student Cases Were Used....................88
4.7. Comparison of Sending and Receiving Schools When Middle
School Student Cases Were Used........................89
4.8. Comparison of Sending and Receiving Schools When High
School Student Cases Were Used........................ 89


4.9. Comparison of Sending and Receiving Schools When Only
Building Level Student Achievement Test Data
Were Used.............................................90
4.10. Comparison of Sending and Receiving Schools When District
Level Student Achievement Test Data Were Used......91
4.11. Comparsion of Sending and Receiving Schools When All
Student Cases Were Used............................. 95
4.12. Comparison of Sending and Receiving Schools When Same
Achievement Test Was Used In Both Schools...........97
4.13. Comparison of Sending and Receiving Schools When
Different Achievement Tests Were Used................98
4.14. Comparison of Sending and Receiving Schools When No
Tuition Was Paid....................................99
4.15. Comparison of Sending and Receiving Schools When Partial
Tuition Was Paid....................................100
4.16. Comparison of Sending and Receiving Schools When
Elementary Student Cases Were Used...................100
4.17. Comparison of Sending and Receiving School When Middle
School Student Cases Were Used..................... 101
4.18. Comparison of Sending and Receiving Schools When High
School Student Cases Were Used.................... 102


4.1.9 Comparison of Sending and Receiving Schools When Only
Building Level SES Data Were Used.................103
4.20. Comparison of Sending and Receiving Schools When District
Level SES Data Were Used.........................104
4.21. Correlation Coefficients of Sending School Test Scores,
Receiving School Test Scores, Sending School SES and
Receiving School SES..............................................106
x


CHAPTER ONE
BACKGROUND AND INTRODUCTION
Perhaps the single most promising idea is
choice. It responds to a simple but quite serious
problem. In most places around the country students
are arbitrarily assigned by their school systems to a
single public school. If that school is a bad one, its
students are trapped. Their parents have no chance
to shift them to another public schoolmaybe just a
few city blocks awaythat has better teaching or
better discipline or just plain higher quality over all.
It's a system of self-perpetuating mediocrity.
Poor schools have no incentive to improve; their
students are captive clients; and parents have no
opportunity to take their business elsewhere. (Bush,
1989)
This statement by President Bush during his first month in
office indicates the current political climate supporting schools of
choice. The advocates of increased choice in education faced a
very favorable environment in the 1980s. They had a President
and administration that favored increased choice and advocated
voucher and tuition tax credits. The Mueller v. Allen Supreme
Court decision in 1983 established a precedent in their favor and
the recommendations of the National Governors' Association
report, Time for Results, supported increased choice within the


public sector (Boyd & Kerchner, 1987).
The concept of schools of choice, however, is not a new one.
John Stuart Mill was an early proponent of the idea.
All that has been said of the importance of
individuality of character, and diversity in opinions
and modes of conduct, involves, as of the same
unspeakable importance, diversity of education. A
general State education is a mere contrivance for
molding people to be exactly like one another; and as
the mold in which it casts them is that which pleases
the predominant power in the government--whether
this be a monarch, a priesthood, an aristocracy, or the
majority of the existing generationin proportion as it
is efficient and successful, it establishes a despotism
over the mind, leading by natural tendency to one
over the body. An education established and
controlled by the State should only exist, if it exist at
all, as one among many competing experiments,
carried on for the purpose of example and stimulus to
keep the others up to a certain standard of excellence.
(Mill, 1898).
Milton Friedman, a Nobel laureate economist, is a
contemporary advocate of choice. He first described choice in
some detail in Capitalism and Freedom (1958). In order to
maximize individual choice, minimize inappropriate participation
by the government and encourage greater economic efficiency in
the operation of schools, Friedman proposed that the government
provide families with vouchers redeemable at approved schools
of their choosing. Other modern day free-marketers include E. G.
2


West (1982,1986) who follows the logic of Friedman, and John
Coons (1979) who advocates choice by strengthening the family's
role in education by subsidizing a much wider range of private
choice than is now the case.
Proponents of the market economy model believe that by
giving students and their families a choice of schools, and by
requiring schools to compete for students, massive increases in
educational effectiveness and output would result. If schools had
to compete for students, they would likely be much more
responsive to the particular needs of their clientele (Levin, 1968).
John Chubb and Terry Moe (1986), education advisors to
President Bush, state that if public schools are to develop the
organizational qualities essential for real improvements, it may
be necessary to emulate the system of control that governs the
private schools, where teaching and professional autonomy
flourish. Chubb and Moe insist that in general terms, effective
control over school would have to be transferred from the
government to the market.
These libertarian arguments, which propose the
maximization of individual choice and minimization of
inappropriate participation by the government, and these
economic efficiency arguments, which propose that choice and
competition would increase educational effectiveness, output, and
excellence have led to the development of a variety of choice
3


options in American schools. Mary Ann Raywid (1985) in "Family
Choice Arrangements in Public Schools: A Review of the
Literature" states that there are four possible areas of choice:
curriculum and content, methods, teachers, and schools. She goes
on to classify thirty-six types of plans.
Writing in 1989, Raywid states that demographic data
about schools of choice are relatively scant. There have been
only two national surveys of such schools in the past decade, one
focusing on public alternative high schools (Raywid, 1985) and
the other looking at magnet schools at all grade levels (Blank,
1984). The magnet school survey located 1,019 such programs.
The alternative school survey located 2,500 but estimated that
the actual total might be three or four times that number. Both
these surveys were conducted in 1981 and now appear dated.
Without an up-to-date survey, there is no way of determining the
total number of such schools; but there are reasons to believe
that the number of such schools is now substantially higher. The
spread of the magnet school concept and the renewed interest in
dropout prevention are two reasons for this speculation (Raywid,
1989).
Some researchers feel that this recent rise in number of
choice options in the public schools is not so much a result of
libertarian and economic efficiency arguments as a result of the
"excellence" movement. They feel that the pursuit of excellence
4


has replaced equity as the leading goal of American schooling
(Boyd & Kerchner, 1987). Schools of choice have become one of
the major vehicles to attain this excellence. The list of advocates
promoting this idea is lengthy (Bush,1989, Clinton, 1988, Finn,
1989, Kean, 1989, Lieberman, 1989, Nathan, 1989a &b, Raywid,
1989,). The 21st annual Gallup poll of public attitudes reported
that the public favors, by a 2-1 margin, allowing students and
their parents to choose which public schools in their communities
students will attend (Elam & Gallup, 1989).
This advocacy by educators and desire by the public for
choice is grounded in the belief that choice in the public schools is
central and fundamental to school improvement and excellence
(Nathan, 1989b). Nathan states that schools of choice will bring
about the following changes:
1. Significant improvements in student achievement levels,
accompanied by major reductions in dropout rates.
2. Better student attitudes toward self, school, learning and
educators.
3. Improved educator morale.
4. Greater parental involvement, commitment and
satisfaction with schools.
5. More students of different racial and economic
backgrounds learning with and from each other (Nathan, 1989b).
Unfortunately for choice advocates it is difficult to draw
5


inferences about the consequences of choice plans from the
available evidence. Murnane gives three reasons for this
difficulty. First, choice plans vary a great deal in how they work
and the consequences are sensitive to the details of the individual
plans. Second, there is no reason to believe that the available
evidence accurately portrays the range of programs in operation
in the more than four thousand U.S. school districts that provide
some students with choices of academic programs. Third, all of
the studies of public-sector family-choice plans have one
significant drawback. They provide little direct evidence on the
critical question of how choice affects student achievement
(Murnane, 1986a). An additional reason why it is difficult to
draw inferences about the consequences of choice plans is the
lack of valid and reliable research concerning choice.
Raywid also acknowledges the limitations of the kinds of
evidence available to support student achievement outcomes of
schools of choice. Virtually none of it is experimental; most of the
available evidence comes from correlational studies or from
evaluation of individual programs with no comparisons with
control groups. This lack of experimental studies makes it
difficult to isolate cause-and-effect relationships, for example, to
tell whether academic achievement in a particular school can be
attributed to the school climate, the nature of students, or the
instructional effectiveness and dedication of teachers. Thus,
6


explanations must remain hypotheses (Raywid, 1989).
Despite this lack of empirical research to support the
outcomes proposed by Nathan and others, the growth of choice
options continues. Perhaps the newest option and the one with
the least amount of empirical support is the interdistrict open
enrollment option, often called the Minnesota plan. This plan
allows parents to send their children to public schoolacross
town, in the neighboring suburb, or anywhere in the state. Full
implementation of the plan is to take place in 1991. Five other
statesNebraska, Iowa, Ohio, Idaho, and Arkansashave also
passed laws which will allow the implementation of interdistrict
open enrollment in 1991. It is this choice option that was
investigated in this study.
The interdistrict enrollment option is the plan that most
closely mirrors the economic models which speculate that the
introduction of a market economy into public education will
result in increased efficiency and excellence. These speculations
result from applying the theory of consumer choice (Bettman,
1979) to the interdistrict enrollment option. The basic approach
taken in this theory is to view the consumer as a processor of
information. The consumer is characterized as interacting with
his or her choice environment, seeking and taking in information
from various sources, processing this information, and then
making a selection from among some alternatives.
7


Ultimately, consumers make choices to accomplish goals.
Given some set of goals, the consumer devotes attention to that
information available which is relevant to attaining those goals.
The consumer then interprets this information in light of
previous knowledge and the context in which the information is
obtained (Bettman, 1979).
Once a consumer has obtained and analyzed this
information, then a choice or preference leading to the
accomplishment of the goal is made. Economists use the term
utility to refer to a consumer's preferences, or to the satisfaction
that a person receives from consuming particular units of a
particular good. A basic premise of the theory of consumer
choice is that consumers attempt to maximize their choices. The
decision making process leads to a consumer's maximum
satisfaction position (Miller & Meiners, 1986).
The study of the application of this theory to the inter-
district enrollment choice option required an assumption about
what educational utility parents and students maximize when
they choose schools. For the purposes of this research, it was
assumed that when students and parents exercised the inter-
district enrollment option they were maximizing school quality.
There exist a variety of measures upon which parents and
students can form their perceptions about school quality. In this
study, the measures of school quality were indicated by student
8


achievement as measured by standardized achievement test
scores and school building socioeconomic status as measured by
free and reduced lunch data.
The use of standardized achievement test scores as a
quality school measure is grounded in more than ten years of
Effective Schools Research. Edmonds defined effectiveness as a
highly circumscribed quantitative measure of school
improvement in which students acquisition of basic skills is
measured by recording the annual increase in proportionate
mastery in the lowest social class (Edmonds, 1982). This
measurement of effectiveness was determined by using reading
and math scores on the Stanford Achievement Test and the Iowa
Test of Basic Skills. In the 1978 Search for Effective Schools
Project, the mean math and reading scores for 20 randomly
selected schools in Detroit were compared with citywide norms.
An effective school among the 20 was defined as being at or
above the city average grade equivalent in math and reading. An
ineffective school was defined as one below city average
(Edmonds, 1979).
Although there are methodological problems' with the
Effective Schools Research (the comparison of extreme outliers,
the reliance on case studies, the cross-sectional rather than
longitudinal design, the failure of some studies to control for
confounding variables, and the lack of generalizability),
9


Rosenholtz regards Effective Schools Research as significant.
Several studies describe "turnaround" schools that, because of
changes in organizational conditions, became more successful,
even when controlling for random error. Analysts find that
organizational characteristics account for 32 percent of between-
school variance in student achievement. In addition, Effective
Schools Research has been conducted within a relatively
compressed time frame, not building serially from one study to
the next; yet all studies produce common findings with
remarkable consistency (Rosenholtz, 1985). This acceptance of
Effective Schools Research and its reliance on student
achievement as measured by standardized achievement test
scores helps to validate the use of standardized achievement test
scores as a measure of school quality in this study.
The use of socioeconomic status as a measure of school
. quality was based on the work of Coleman (1966). The finding
that student body composition is an important determinant of
school effectiveness first gained prominence with the publication
in 1966 of the Coleman report. This report emphasized the
positive relationship between the average socioeconomic status of
the students in a school and the academic achievement of
individual students. Subsequent studies either replicated that
finding or found the achievement of individual students to be
related to other characteristics of the student body such as racial
10


composition or the average achievement level of the students in
the class or school.
Purpose of the Study
The purpose of this study was to test the application of the
consumer choice theory in the public education setting by
studying data generated when students cross school district
boundaries to attend school. This study tested the premise that
when consumers make a choice of which school to attend they
maximize school quality. Given the opportunity, parents and
students will move to a school of greater quality as measured by
student achievement and SES. No theoretical research currently
exists to support the claims that interdistrict open enrollment
implementation would create a market economy that would
loosen the monopoly that currently exists in public education.
Yet, proponents of open enrollment continue to argue that by
allowing students to choose their school attendance more
effective schools would survive and less effective schools would
fail (Finn, 1986; Levin, 1968; Lieberman, 1989; West, 1986; ).
Although Colorado has not formally adopted legislation to
promote and facilitate students crossing district boundaries to
attend school, in the 1989-90 school year the Colorado
Department of Education reported that 4,130 students did indeed
attend schools outside of their home district. These data indicate


that students and parents in Colorado are attempting to maximize
utility by making choices about which district to attend school in.
Despite the absence of formal legislation, students are attending
schools of their choice by arranging informal agreements between
districts.
Major researchers in the area of schools of choice
concentrate on the reasons why choice should exist (Elmore,
1987; Nathan, 1989b) or the effects of choice on individual
schools (Nathan, 1989b; Raywid, 1989). Some researchers are
beginning to look at the reasons why students and parents make
educational choices. A report prepared by the research
department of the Minnesota State House investigated the
reasons Minnesota students transferred to other school districts
under the states open enrollment plan. Nearly 40 percent of the
1,234 students who gave a reason for transferring said they did
so either because a school in a neighboring district was closer to
home or because it has a day-care program. Twenty percent said
they were transferring for academic reasons, and six percent said
they changed schools for athletic, extracurricular, or social
reasons. Other students sought a better general school
environment, and still others were continuing in other districts
under agreements that had been reached before open enrollment
came into effect (Education Week, 21 February, 1990).
No researchers, however, have carefully investigated


whether when students and parents exercise choice they
maximize utility and choose a school of greater quality. When
students and parents choose are they acting as rational
maximizers and choosing schools of better quality as measured
by SES and student achievement? It was the purpose of this
study to determine the validity of this argument by testing the
following hypotheses derived from consumer choice theory.
Hypothesis One: When students cross school district
boundaries to attend school, the receiving school has a higher
percentile rank of the mean scores on standardized achievement
tests than the sending school.
Hypothesis Two: When students cross school district
boundaries to attend school, the SES of the receiving school is
higher than the SES of the sending school.
Hypothesis Three: When students cross school district
boundaries to attend school, the receiving school has a higher
percentile rank of the mean score on standardized achievement
tests than the sending school when school populations are
controlled for SES.
Methodology
El Paso County, Colorado, was chosen as the data base for
the purposes of this study. Fifteen different school districts exist
in El Paso County educating approximately 60,000 students.


Urban, suburban and rural districts exist in the county allowing
the representation of three different geographic areas. There has
been a growing trend in El Paso County of increased cross district
enrollment. From 1987 to 1990 there was a 50 percent increase
in the number of students in El Paso County attending school
outside of their home district. In the 1989-90 school year, 446
students crossed district boundaries to attend school. Of these,
362 were non-tuition students and 84 were partial tuition
students. Both the partial tuition and non-tuition students made
up the population of this study. In order to test the three stated
hypotheses, all schools in El Paso County were assigned two
values, one based on student achievement as measured by the
percentile rank of the mean scores on standardized achievement
tests and one based on SES as measured by the percentage of
students in a school receiving free and reduced lunch. The
movement of students across school district boundaries was
traced. Dependent sample t-tests were run on the data to
determine whether to accept or reject hypothesis one and
hypothesis two. A Pearson product-moment correlation was used
to determine the feasibility of testing hypothesis three.
1 4


CHAPTER TWO
LITERATURE REVIEW
The debate about the choices families should have in
deciding how their children should be educated has a long
history in the United States. The current interest and
attention given to the choice issue is merely an extension of
the history of American education. Much has been written
about choice in the recent past. However, very little has been
written on a comprehensive scale to explain the complexities
and varieties of school choice. In addition, little research has
been done to test if the predicted outcomes would occur if
more school choice were allowed.
This literature review begins by providing a brief
overview of the multitude of issues surrounding choicethe
choice options in existence today, the history of choice in the
United States, the libertarian arguments regarding choice, and
the existing empirical research on choice. The intent is to
provide the reader with a sufficient background to better
understand the school choice issue. Once this background is
established, this literature review then moves to provide an


understanding of choice as it affects schools. Both excellence
and economic efficiency arguments are presented. Finally the
economic theory of consumer choice is explained and then
proposed as a framework in which to test the effects of the
interdistrict enrollment option in Colorado.
Current Choice Options Available
In order to assist the reader in understanding the
choice options available today, current options were divided
into two sections, external options and internal options.
Within each section, the varieties of alternatives are
presented.
External Options
External options give parents the opportunities to
educate their children in programs external to the public
school system. Voucher plans, private and parochial schools,
and home schooling are external options, with the voucher
plan constituting the most radical change to the present
system.
Vouchers. Educational vouchers are tuition certificates
that are issued by the government and are redeemable at the
school of the student's choice (Nathan, 1983). A pure voucher


system would include both public and private schools. At
present, there is little evidence to suggest that a voucher
system would improve the quality of education. To date,
there has been only one school district that has implemented
the voucher concept, Alum Rock, California (Lindelow, 1980).
The voucher experiment in Alum Rock only lasted for five
years from 1972-78.
Despite its short term implementation, an evaluation of
the Alum Rock experiment provides some direct evidence
available about schooling under a voucher plan. According to
the evidence, the single most important determinant of
parents' school choices was proximity to the home, even
when free transportation was provided. (Wise & Darling-
Hammond, 1986). Socially advantaged parents were more
aware of the program. Parents with more education were
more likely to choose "innovative" programs; those with less
education were more likely to chose "traditional" programs.
Classroom observation indicated that there was little
diversity in the educational programs of the mini-schools;
variations that did exist were not associated with mini-school
differences, and some announced curricular offerings were
not available to students. When the voucher experiment
ended, the education scene in Alum Rock was little different
from what it had been five years earlier (Salganik, 1981).
More recent efforts to promote the voucher system


come in the form of a law suit filed in federal court in July,
1989, by John Coons and Stephen Sugarman. This law suit
urged the creation of a voucher program to enable black
school children in Kansas city to attend desegregated private
schools at public expense (Snider, 1989). In addition, the
Wisconsin legislature in March, 1990, approved a bill that
will give almost 1,000 Milwaukee public-school students the
option of attending non-sectarian private schools at state
expense. Under the new program, which will last five years,
no more than 1 percent of Milwaukee's 93,00 students will be
permitted to enroll in non-sectarian schools that agree to
accept the state's per-pupil allotment--approximately
$2,500as full reimbursement for tuition costs. The measure
limits eligibility to children from families whose annual
incomes are no greater than 175 percent of the federal
poverty level (Snider,1990).
Home Schooling. A small, but increasing trend is the
movement towards educating children in the home
environment. In Colorado, for example, 835 students out of a
total of the 1987 population of over 560,000 pupils were
enrolled in a home study program approved by the Colorado
Department of Education fFoothold on Future 1989 Progress
Report. 1989). The movement is growing throughout the
country. A key issue is whether state and local funding
should go to the home for such instruction. Thus far, states


have dealt primarily with the curriculum and supervisory
issues without approving new funding mechanisms such as
voucher or tuition payments of any type.
Private-Parochial Education. Close to 10 percent of
America's school age children are enrolled in private schools
numbering approximately 27,690 throughout the country.
These include special education, church-related and non-
church related schools. Most pupils fall into the kindergarten
and elementary school age range.
Parent-Run Private Schools. A small, yet interesting
trend in private school choice is that of parent-run private
schools. Estimates indicate there are 36 such schools in the
United States and the number is growing by 5 or 6 each year
(Covarrubias, 1989). The unique features include extensive
parental involvement in ownership, governance, fund raising,
curriculum planning, facility maintenance and transportation.
Internal Options
Internal options include programs within districts,
schools or classrooms in a public setting. These include open
boundary programs, magnet schools, alternative schools,
post-secondary options, second chance programs, and
schools-within-a-school.


Open Boundaries. The cross-district or open boundary
option provides parents the right to send their children to
any school regardless of their place of residence with a given
state. Currently Minnesota is the only state to implement a
cross-district option state-wide, although Kansas, Nebraska,
Ohio, Idaho, and Iowa have plans for implementation in 1991.
The Minnesota cross-district option program is a
voluntary program that must be adopted by the local school
board. Students can be denied enrollment because of two
factors: 1) lack of space and 2) enrollment would cause
noncompliance with desegregation mandates.
Transportation is provided only within the boundaries
of the resident school district. Parents must provide
transportation to the school district boundary, and then the
resident district transports students enrolled in district
schools. Once a student is enrolled, funds from the resident
district are transferred to the nonresident district.
After three years of implementation in Minnesota the
results are as follows. In the 1987-88 school year, 95 out of
430 school districts participated. The students participating
numbered 137 statewide. For the 1988-89 school year, 153
districts elected the option to participate. The students
numbered 440. For the 1989-90 school year, approximately
1,000 of the state's 711,000 students signed up to participate
in the inter-district enrollment option.
20


Parents surveyed, in 1988, indicated the reasons for
choosing nonresident districts in the first year to be as
follows: 1) better curriculum options, 67 percent; 2) closer to
job or home, 26 percent; 3) student social problems, 21
percent; 4) teachers, 16 percent; 5) specialized help, 14
percent; and 6) ability to complete high school in the district
to which parents were moving, 7 percent (Weedle, 1988).
Post Secondary Enrollment Option. The post-secondary
option is designed to allow high school students to attend
colleges and universities with tax dollars usually paying the
cost of tuition and fees (Snider, 1989). Credit may be earned
towards the high school diploma while attending college
classes. At least six states have adopted specific plans that
provide this form of choiceWashington, Utah, Minnesota,
Maine, Florida and Colorado. In most cases, the programs are
operated on a space available basis and without provisions
for transportation.
Second Chance Programs. Targeting at-risk students or
drop-outs, so called "second chance" programs are designed to
provide students with a wide range of program choice
throughout their state to enable them to continue to progress
towards graduation (Snider,1989). Washington is considering
legislation that would limit choices to within a 50-mile radius
of a student's home district but would also offer
transportation reimbursement for students from low income


families (Snider,1989).
Minnesota's plan allows students to earn a high school
diploma by choosing from either: I) a public high school; 2) a
private, non-sectarian school having a contract with a public
school district to provide services under this program; 3) an
approved public alternative education program or area
learning center; or 4) a college or technical institute under the
state's Post Secondary Enrollment Option Program (Pipho,
1989a).
The Colorado Second Chance program has fostered
alternative programs that attract drop-outs across district
lines. State funds follow the student while the local district is
required to help the student enroll in the school of his or her
choice, provide counseling to the student and parents, and
monitor the progress of the student.
Charter Schools. In 1989, four states considered the
creation of charter schools of choice. This option would give
educators the opportunity to create programs theyve always
dreamed about. Teachers could be given grants to design
innovative educational programs. The Dade County, Florida,
school board in December, 1989, approved five proposals for
new elementary schoolsall developed by local educators
under the Saturn Schools Project (Bradley, 1989).
The educational creativity schools initiative in Colorado
is a public-private partnership organized by the Governor in
22


1989. The initiative seeks to encourage: 1) excellence, 2)
teaching children to think,3) participation by parents,4)
innovations, 5) restructuring education, 6) identifying
barriers to excellences, and 7) involving the total community
as participants in educational quality (Foothold on the Future
1989 Progress Report. 1989).
The South Carolina Board of Education has given an
"unprecedented degree of regulatory relief" to about 100 of
the state's approximately 1,000 schools (Flax, 1989). Schools
with a history of superior academic achievement will be
automatically released from numerous state regulations:
class, time constraints, principal certification; class scheduling
and subject area teacher certification.
Open Enrollment Districts. This option is defined as one
in which "school assignment is based on family choice among
all the schools at a child's grade level within the child's home
district" (Newman, 1989). Primary constraints in this option
include: 1) desired or mandated racial balance goals of the
district must be maintained; 2) students must provide their
own transportation; and 3) enrollment is limited to the
availability of facilities. Many school districts throughout
America already have this provision, and students avail
themselves to it on a regular basis. In Colorado, for example,
without a law on open enrollment, 20,000 students out of
550,000 were in schools in their districts but outside their
23


attendance areas (Foothold on the Future 1989 Progress
Report. 1989).
Controlled Enrollment Options. Controlled choice is a
form of within-district choice that fosters two inter-related
purposes: the voluntary desegregation of a community's
schools and the strengthening of each school by giving its
staff responsibility for improving quality (Newman, 1989). It
has been used since 1981 in Cambridge, Massachusetts, to
resolve desegregation problems. School enrollment is based
on racial ratios within the district with seats in each school
are allocated proportionately to the racial or ethnic groups in
the community. Within this constraint, families may select
the school they wish to have their children attend. Fall River,
Lowell, Lawrence and Boston have adopted this model in
Massachusetts, as have Seattle, Washington, and San Jose,
California.
Magnet Schools. Magnet schools gained momentum
nationally in the early 1970's and are designed to provide
specialized programs to meet the different learning needs
and interests of students within the existing structure of a
school district. Court desegregation orders and federal
funding have often been a factor in establishing magnet
schools (Raywid, 1989). Other reasons given for growth of
magnet schools include school renewal, drop-out prevention
and school reform.
24


Montclair, New Jersey; Montgomery County, Maryland;
and Cambridge and Acton, Massachusetts, are among the
districts which operate elementary schools on a choice basis.
Magnets are usually organized around academic specialty
areas and may encompass a whole school or a program within
a school. Magnets are generally open to all families within a
district on an open enrollment basis. District 4 in East
Harlem, New York, has offered choice programs in their
schools for the past 13 years. Secondary magnet programs
operate in Los Angeles, Denver, Detroit, Houston, Dade County,
Seattle, Chicago and in Buffalo, Mt. Vernon, Newburgh, New
Rochelle, New York City, Rochester, and Syracuse in New York
State.
Alternative Programs. At the beginning of this century,
fewer than 10 percent of all youth in America graduated
from high school. Most schools at the time were oriented to
college preparation and only a small minority of students
would move on to that setting. The development of
alternative programs and schools was a sincere effort to
provide an option for many students who otherwise would
have quit schools entirely (Smith, Burke & Barr, 1974).
Early efforts to begin alternative schools stemmed from
the realization that one mode of schooling could not possibly
reach every student--people learn in diverse ways, settings,
and structure. Communities responded differently. Some
25


saw the need for starting new schools for the talented such as
the Bronx High School of Science in New York City in 1938.
Some provided for drop-outs and at-risk youth such as the
Metropolitan Youth Education Center in Denver in 1964. The
Parkway Plan was developed in Philadelphia in 1969 and
was an attempt to provide options to the traditional
educational program by furnishing learning experiences
throughout the community. The alternative school movement
has included continuation schools (Newport Beach, California),
vocational or career enrichment schools (Westminster School
District No. 50, Colorado), multi-cultural schools (Berkeley,
California), free schools (Berkeley, California), schools without
walls (Madison, Wisconsin), open living schools (Jefferson
County, Colorado), and fundamental schools (Denver,
Colorado).
This brief review of choice options as they exist in
education today indicates that school options are part of the
American educational system. This literature review now
provides a history of school choice in America as a means of
understanding how it is school choice arrived at where it is
today.
The History of School Choice in the United States
From the beginning of American schooling, choice has
26


been an option taking many forms. Joel Spring (1982), in
The Public School Monopoly. (1982) chronicles the evolution
of the political structure of American schooling which has
influenced family choice. He writes that fear of government-
operated schools as an instrument of control and domination
was often expressed by the ideologists of the American
Revolution. Freedom of speech and thought was of primary
concern in the many pamphlets written and distributed in
England and the American colonies before the Revolution. It
was these pamphlets that provided the intellectual
justification for the Revolution and for the first time in
Western history presented the argument that the progress of
society depended upon freedom from government
intervention in the control of ideas and their expression.
Part of this belief in the necessity of freedom of ideas
was a fear that the establishment of government-operated
schools would seriously abridge the free development of
ideas. As this libertarian tradition evolved in the eighteenth
century, concern about freedom of thought and expression
became specifically linked to opposition to government-
operated schools. At the end of the American Revolution
there was no rush to establish government-operated systems
of education. There were proposals for government-operated
schools at the end of the Revolution, but these proposals,
except for those of Thomas Jefferson, emphasized the creation
27


of a political and moral orthodoxy. These proposals reflected
a concern with limiting freedom by controlling individual
conscience. Thomas Jeffersons proposals did not reflect this
moral orthodoxy. He proposed only that all children should
receive three years of free education in reading, writing, and
arithmetic, and that political values were to be formed by the
reading of a free press (Spring, 1982).
Because of these arguments and ideologies, school
choice in the eighteenth century was quite common. James
Guthrie writes that schooling in America was largely private
until the last quarter of the 19th century. Tutoring for fees,
apprenticeships, dame schools, church schools, and
proprietary institutions were all popular (Guthrie, 1983).
Lloyd Jorgenson states that American schools of the colonial
and early national periods were neither wholly public nor
wholly private, as we now use these terms. They were in
large part the product of voluntary efforts, aided sporadically
by government grants. Government financial support for
voluntary (including denominational) schools continued well
into the national period. Far from prohibiting such support,
the early state constitutions and statutes actively encouraged
this policy (Jorgenson, 1987).
It was not until the common school movement begun
in the 1820s and 1830's that family choice in American
schools began to focus on the development of public schools.
28


The campaign to extend educational opportunities in the
United States at the primary-school level and to bring them
increasingly under civil control was set into vigorous motion
during the years 1830-60 and consolidated in the decades
after the Civil War. By 1900, the non-public school share of
primary school enrollments had shrunk to about 8 percent of
the total. According to Jorgenson, this was a striking victory
for the champions of public education.
This growing movement to abolish the inclusion of
private education in family choice came to a head in Oregon
in 1922. The Oregon state legislature, under pressure from
the Klu Klux Klan and other nativistic groups, submitted to
the voters a measure in the general election of 7 November
1922. This law required parents to send their children
between the ages of eight and sixteen to public schools. The
measure was approved by a vote of 115,506 to 103,685.
The law was challenged by the Sisters of the Holy
Names of Jesus and Mary, who maintained a number of
schools in Oregon, and by the Hill Military Academy. On 17
March 1925, the United States Supreme Court declared the
Oregon law unconstitutional. In a clear affirmation of the
rights of family choice, the court declared:
The fundamental theory of liberty upon which all
governments in this Union repose excludes any general
power of the state to standardize its children by forcing
them to accept instruction from public teachers only. The
29


child is not the mere creature of the state; those who
nurture him and direct his destiny have the right coupled
with the high duty to recognize and prepare him for
additional obligations (Pierce vs. Society of Sisters. 1925.)
The court ruling in Pierce vs. the Society of Sisters
insured the rights of families' choice in both the private and
the public educational sector. The history of choice in the
twentieth century, then, is one of creating options in both the
private and public domain.
The private side of choice during the twentieth century
has been allowed to exist without the help of public money.
According to Donald Erickson, the patterns of growth and
decline in private schools vary. At the end of World War I, in
the context of rampant anti-foreign sentiment, many private
schools dropped their ethnic trappings and promptly went
out of business. Amish schools arose at the end of World War
II, when simple public schools in the countryside, where
Amish children could be protected from the influences of
mainstream culture, were replaced with bus rides and
consolidated schools in nearby towns.
In the shadow of the Vietnam War, military schools
became singularly unpopular, but more recent interest in
discipline has reversed that trend. Many private schools
sprang up in the Progressive era, and "free schools" later, in
the heyday of the 1960s emphasis on "doing your own thing."
Black Muslim schools appeared in the context of an emphasis
30


on "Black Power" (Erickson, 1986).
Church supported schools have a similar history during
the twentieth century with disparate dynamics of enrollment
since Pierce vs. the Society of Sisters. Today an estimated
11% of the school-aged population in the U.S. attends non-
public educational institutions (Erickson, 1986). These
institutions encompass an extraordinary range of conditions,
ideologies, and reputations providing for students and
parents who are willing and able to pay the price, a wide
variety of choices.
For the 89% of the school-aged population who attend
public schools, there exists a variety of choices. This variety
of choice has its own history in the twentieth century. In
1958, Milton Friedman advocated family choice in schools in
some detail in Capitalism and Freedom. In order to maximize
individual choice, minimize inappropriate participation by the
government, and encourage greater economic efficiency in
the operation of schools, Friedman proposed that the
government provide families with vouchers redeemable for
services at approved schools of their choosing (Friedman,
1958). According to Laura Salganik, during the 1960s
vouchers seemed to have no opponents. Supporters, along
with the conservative economist Friedman, included liberal
school critic Mario Fantini, and socialist sociologist
Christopher Jencks (Salganik, 1981). Despite this widespread


philosophical support, when the Office of Economic
Opportunity in 1970 offered to fund a voucher system only
Alum Rock, California, agreed to participate. The voucher
system in Alum Rock remained in existence for only six
years.
In early 1980, John Coons and Stephen Sugarman,
lawyers and school finance reform activists, announced a
petition campaign to put vouchers on the California ballot.
Their proposal retained the public system, but gave students
the option of receiving vouchers to attend family choice or
independent public schools, which could charge tuition only
according to the family's ability to pay and accept most
students only by lottery. The campaign was poorly organized
and received little support (Salganik, 1981).
In order to continue providing an understanding of the
school choice issue, this review will now turn to an
explanation of the libertarian arguments concerning school
choice. Many current proponents of school choice rely on
libertarian arguments to justify the expansion of school
choice.
Libertarian Arguments Concerning School Choice
In 1792, Thomas Paine was one of the first Americans
to have consistently urged the distribution of public


\
\
ss to the family to purchase its children's education,
n of speech and thought was of primary concern in
ny pamphlets written and distributed in England and
erican colonies before the Revolution. As part of this
n the necessity of freedom of ideas was a fear that the
hment of government-operated schools would
y abridge this liberty.
\
0 eighteenth century libertarians, freedom of thought
:ech was essential for the general well-being and
1 of society. This meant an absolute and not a
t
i freedom of thought. Knowledge was so complex that
on or groups of persons could determine before the
ation what\ areas of thought should be limited. As
ertarian tradition evolved, concern about freedom of
and expression became specifically linked to
on to government-operated schools (Spring, 1982).
hese libertarian arguments appeared several
ons later in the work of John Stuart Mill. In 1859,
veloped the implications of his commitment to
\m in his celebrated essay On Liberty.
\ has been said of the importance of
\ of character, and diversity in opinions
\conduct, involves, as of the same
\ortance, diversity of education. A
\ation is a mere contrivance for
\>e exactly like one another; and as
\casts them is that which pleases
i
33


the predominant power in the governmentwhether
this be a monarch, a priesthood, an aristocracy, or
the majority of the existing generationin
proportion as it is efficient and successful, it
establishes a despotism over the mind, leading by
natural tendency to one over the body. An
education established and controlled by the State
should only exist, if it exist at all, as one among
many competing experiments, carried on for the
purpose of example and stimulus to keep the others
up to a certain standard of excellence (Mill, 1898).
E. G. West, a contemporary scholar, supports the
libertarian ideas of Mill. West states that a publicly operated
school system results in a diminution of the liberty of
parents, children and taxpayers to a wholly unnecessary
degree. He further states that all the objectives that public
schools are designed to serve are achievable simply by
passing a compulsory education law and then subsidizing
parents so that they could afford to send their children to
private schools. The argument of West ultimately rests on
the notion that diminution of liberty occasioned by the public
schools lowers the general welfare (VanGeel, 1978).
Joel Spring, in analyzing findings regarding political
socialization studies in public schools, states that the results
of these studies demonstrate that public schools are
contributing to the development of an apolitical citizenry.
One of the conclusions of political socialization studies is the
apparent lack of school effect on political attitudes and
knowledge. This does not mean that these attitudes and
34


knowledge are developed in a separate context, but that they
are never developed. This is an important point, for it
supports the idea that the present system of government
control depends for its perpetuation upon a lack of political
sophistication in the general population. Spring sees this as a
leading to a dangerous situation in which the production of an
apolitical citizen is ideal for a political system in which power
resides in the hands of expert managers, not the citizens.
In order to achieve the rebirth of the political person,
Spring feels that the government must get completely out of
the business of education. Competitive educational
institutions are needed that reflect a wide variety of
ideologies. The evolution and progress of U.S. culture and
society depends on the free interplay and development of
ideas not now possible in public schools (Spring, 1982).
Stephen Arons bases his libertarian arguments on the
statement that the present structure of schooling in the
United States taxes dissent by providing free choice for the
rich and compulsory government socialization for the poor. It
is economic discrimination in the distribution of First
Amendment rights to condition the provision of "free"
education upon the acceptance of majority-approved value
inculcation in government schools.
The current structure of education in the United States
is broadly inconsistent with the values advanced by the First
35


Amendment and represented in the system of freedom of
expression. Arons states that this inconsistency between
majoritarian control of schooling and the principles of the
First Amendment becomes more crucial and more painful at a
time when cultural explanations are collapsing and people are
more in need of open communication and room to develop
values than they are of repressive enforcement of orthodox
views. And so the question of educational choice for families
transforms itself into the question of what structure of
education is most likely to survive in a diverse society
undergoing a cultural crisis (Arons & Lawerence, 1982).
This review of the choice options currently available,
the history of choice and the libertarian arguments which
advocate choice provide little empirical data on which to
assess the effects of school choice. Furthermore, researchers
have not provided a large body of data upon which to draw
conclusions regarding the effects of choice. A small research
base does exist and it is to this existing research that this
literature review now turns.
Research Base Surrounding School Choice
Some school choice researchers have focused on the
anticipated increases in student achievement that will occur
as a result of choice. According to Murnane, however, all of
36


the studies of public-sector family-choice plans have one
significant drawback: They provide little direct evidence on
the critical question of how and if choice affects student
achievement (Murnane, 1986a). Raywid also acknowledges
the limitations of the kinds of evidence available to support
student achievement outcomes of schools of choice. Virtually
none of it is experimental; most of the available evidence
comes from correlational studies or from evaluation of
individual programs with no comparisons with control
groups. This lack of experimental studies makes it difficult to
isolate cause-and-effect relationships, for example, to tell
whether academic achievement in a particular school can be
attributed to the school climate, the nature of students, or the
instructional effectiveness and dedication of teachers. Thus,
explanations must remain hypotheses (Raywid, 1989).
Other researchers have focused their studies on
examining the school choice behavior of public school parents.
Nault and Uchitelle (1982) conducted a study of choice
behavior of parents living in a community called Collegeville.
They focused on the choice behavior of parents to learn how
they make their decisions, what factors influenced their
choices, and whether as a result of their decision making they
became more knowledgeable about the schools they selected.
Their study was theoretically based on two assumptions
concerning the effects of choice. 1) If parents are allowed to


select the school that most closely fits their conception of how
schooling should proceed, then there will be greater parent
satisfaction. 2) If parents could select schools, schools would
become more responsive to parent aspirations and be more
likely to offer quality instruction. Nault and Uchetille felt
that scrutiny of parental choice behavior in Collegeville would
usefully inform the largely speculative arguments about what
would happen if school choice were more commonplace.
The major findings of their study were as follows. Less
well-educated and less well-off parents in the population of
the study were less likely to be aware that they had a choice.
Parents of lower socioeconomic status who were aware that
they had a choice were less likely to have investigated the
schools directly before making a decision. In discerning the
factors influencing choice, Nault and Uchitelle found that
parents seemed most concerned about the general
atmosphere in the school, and they seemed to appraise the
atmosphere largely on the basis of the instructional and
managerial styles of the two persons in the schoolthe
principal and their child's prospective teacherwho were
most likely to influence their child's school experiences.
Though not unconcerned about achievement levels, most
parents ranked this measure of school output as
comparatively unimportant. Systematic search by parents
was more common than cursory investigation and parents
38


afforded choice became more knowledgeable about the school
to which they sent their children than did similar parents
assigned to a neighborhood school. Their findings appear to
support the argument that, given a choice, parents will
attempt to maximize utility in the market place and choose
the school they feel is best for their child.
Researchers Moore and Davenport have focused on the
effects of choice on students at risk. Their studies attempted
to discern whether choice would help or hinder school
systems committed to educating students at risk. In studying
the student placement and labeling at the high school level in
New York City, Chicago, Philadelphia, and Boston, Moore and
Davenport found that educational choice was, in the main,
operating to the detriment of students at risk. They found
that a system that was supposed to provide students with
choices has, too frequently, left the typical entering high
school student in big cities out in the cold. The non-selective
neighborhood schools, which 70% to 80% of students
continued to attend, were placed at an increasing
disadvantage by the reallocation of students, staff and
resources to academically selective schools and programs
within schools (Moore & Davenport, 1988).
In a related study, Hammons and Olson (1988)
discovered a very strong relationship between inter-school
transfer and high school graduation. The study examined a


randomly selected sample of Denver Public School eighth
graders in 1980-81 and checked the graduation or
discontinuance of each student five years later. Their findings
showed that 95 percent of the non-movers graduated from
high school. Only 68 percent of those who moved once
successfully completed high school. At two transfers, just
over half (56 percent) graduated. At three moves, only 30
percent did so. These findings have implications for deciding
how much choice is good for students. At some point students
may not be acting as rational maximizers but rather as poor
decision makers.
McPartland (1971) studied the effect of choosing on the
behavior of high school students. They studied student
participation in decision making in fourteen urban high
schools. Of these schools, one in particular, provided students
with an unusual degree of academic choice. Each quarter,
students in this school made specific choices of course and
instructor. McPartland found that these students were more
committed to the school, evidenced higher attitudes of
responsibility, and had more personal and carefully
considered reasons for attending than did students not given
choices. Students given real choice expressed higher
satisfaction with both their teachers and the instruction they
received than did other students. Choice also led to higher
teacher-student trust and concomitantly lower hostility
40


between these role partners. McPartland reported that
students who made decisions did not generally take the easy
way out by selecting only the courses with less stringent
requirements and grading scales.
Nault (1975) studied the differential commitments of
students affiliating with schools by their own choice and
those affiliating through parental compulsion. Using a sample
of freshmen attending six Catholic high schools, Nault found
that students allowed to make their own affiliation decisions
were significantly more committed to their schools on each of
four dimensions: behavioral expectation of the school,
extending oneself on behalf of the formal school organization,
expectations in academic performance, and extracurricular
participation.
Kottkamp (1979) partially replicated Nault's findings in
a public high school setting. Students who made their own
choice among five distinct programs, as opposed to those
whose parents made the choice, behaved better both inside
and outside the classroom. Choice-making students also
produced higher grade-point averages.
Hartman (1980) further replicated Nault's findings in a
study of California public school students. His study findings
showed that voluntarily assigned student have significantly
higher commitment toward formal school, toward compliance
with behavioral expectations, toward compliance with
41


academic expectations and toward achievement of
competency in reading and math.
These studies of high school student choice show that as
consumers high school students often times maximize utility
in the educational market place. Those students, who were
allowed to choose, scored significantly higher on indicators of
commitment and achievement.
At this point in the literature review, the reader should
have a better understanding of some of the attributes and
arguments concerning school choice. Now this literature
review will turn to economic efficiency arguments concerning
school choice. Studies of school choice effects are grounded in
the belief that the harnessing of economic market forces will
create better schools. Proponents of school choice state that
by allowing free market forces to operate better and more
efficient schools will result. These propositions have led to
school choice literature which analyzes the economic
efficiency arguments concerning school choice and the
excellence movement arguments regarding school choice.
Economic Efficiency Arguments Concerning School Choice
Supporters of school choice draw upon a wide variety of
literature in the economic arena. It is from this body of
economic literature that the theory upon which this study


rests is drawn. Some of this literature is very specific to
education and some is more general economic theory. Major
authors in the general economic theory arena include
Borcherding (1977), Boulding (1966), Buchanan (1987, 1968,
1962), Buchanan and Tollison (1984, 1972), Downs (1957),
Hirschman (1970), Savas (1987), and Spann (1977). Major
authors in the educational arena are Bridge (1978), Chubb
and Moe (1988,1986), Coons (1979), Erickson (1976), Finn
(1986), Hanushek (1986), Krashinsky (1986), Levin (1968),
Lieberman (1989), VanGeel (1978), and West (1986).
Advocates of choice in the public finance arena agree
that society chooses to consume many goods and services on
a collective basis. However, they state the fact that just
because goods and services are financed by tax revenues (in
whole or in part) and consumed in some type of collective
fashion does not mean that such goods must be produced by
the government (Spann, 1977).
A new word was coined to explain this growing
sentiment in public policy circlesprivatize. Privatize first
appeared in a dictionary in 1983 and was defined narrowly
as "to make private, especially to change from public to
private control or ownership" (Savas, 1987). According to
Savas, there are a number of forces behind the move to
privatization: pragmatic (better government), ideological (less
government), commercial (more business), and populist


(better society) (Savas, 1987). Spann states that a major
force in privatization is the question of whether the
substitution of a private firm for a government producer
lowers the costs of providing such services. The issue is one
of efficiency (Spann, 1977).
Several empirical studies of the relative costs of .
government enterprises versus private enterprises have been
done. Davies compared the efficiency of two Australian trunk
airlines and found that the data clearly indicated that the
private airline was more efficient than the public airline by
at least 13 percent (Davies, 1971). Ahlbrandt studied the
cost of contracted fire protection services purchased by
Scottsdale, Arizona. The results of his study showed that
were Scottsdale to establish its own fire department the costs
of fire protection would double (Ahlbrandt, 1973). The
Planning Board in Monmouth County, New Jersey, conducted
a survey of garbage collection costs in the communities in
that county. It found that the communities using private
contractors had significantly lower costs than communities
with public sanitation departments (Spann, 1977).
In summarizing these studies Spann states that a
number of government functions can be taken over by
private producers with an attendant reduction in the costs of
government. Public provision or consumption of a good or
service does not imply public production of the good or
44


service (Spann, 1977). This thinking in the public finance
sector carries over to researchers in education who favor
greater choice by the introduction of economic models into
the educational marketplace. Lieberman (1989) is one of the
major proponents of privatization of education. According to
Lieberman, the only ways to improve American education are
to foster private schools that compete with public schools and
among themselves and/or foster for-profit competition
among service providers within the public school system
(Lieberman, 1989).
Lindsay's work (1976) in applying economic theories to
educational choice follows the line of thinking of the public
finance efficiency studies. He states that the publicly
operated system for supplying educational services is
inefficient--it produces a lower quality of service than a
privately operated system with the same resources for two
reasons. The first of these begins with the observation that
in profit-making enterprises the owners of the enterprise can
monitor the performance of the managers of the enterprise
by assessing the contribution the managers make to the
profitability of the enterprise. But since public schools are
not profit-making enterprises, school boards must monitor
the productivity of administrators and teachers differently.
The school board must identify which attributes of the
schools are most important to them and then determine how


much the administrators and teachers have contributed to
the realization of these attributes. The problem is that
certain attributes of the schools are more readily monitored
by school boards than others.
Furthermore, school boards will not fund outputs they
cannot monitor, for to do so would be to encourage the using
up of resources without the boards being able to check on
their proper use. The result of these factors is that public
schools will emphasize the production of certain easily
measured attributes to the neglect of other attributes not so
easily monitored, but perhaps of great importance to children
and parents. While members of the public may complain to
the school board about these deficiencies, they cannot make
themselves as effectively heard in the short run as can the
consumer in the private market shopping around for bargains
and expressing his or her satisfaction with purchases. Thus
public schools will produce a mix of attributes different and
less preferred than the mix of attributes produced by private
schools, which means the public school education will be of
lower perceived quality (Lindsay, 1976).
Alchian and Demsetz also write about the economic
inefficiency of public school systems. They argue that
because the public school system is a virtual monopoly, free
from competition, and is not profit making there exists no
strong incentive for those operating the system to make the
4 6


most efficient use of the available resources. The cost of
inefficient behavior to the managers of the public schools is
considerably less than would be the cost of inefficient
behavior to those operating private schools in a competitive
market (VanGeel, 1978).
Walberg analyzed 3,000 separate studies of educational
productivity for variations in educational practice that yield
greater student achievement. Despite his research, many of
his findings are not common practice. One reason Walberg
gives for this is that improvements in effectiveness are costly
to educators not in money but in required institutional,
technological, staff, and personal changes. Without the
incentives and discipline of the marketplace, the creativity
and energy of educators may not improve productivity but
preserve tradition, emolument, and office (Walberg, 1984).
Other educational researchers draw upon economic
theory to suggest ways to create educational choice. Gamer
and Hannaway draw heavily on Hirschmans Exit. Voice and
Loyalty (1970) in developing a market model for education.
They state that institutions can be controlled or held
accountable through a variety of mechanisms. One is the
mechanism of market organization, in which the control
behavior is exit. In a market of voluntary buyers and sellers,
the consumer or purchaser is free to exit from any firm
whose prices are too high or whose product's quality is


unsatisfactory. The analysis of market behavior in economics
proceeds from a rather strict set of assumptions about ideal
markets in which pure competition prevails. In the case of
schooling, the pure competition model would suppose that
there were a very large number of schools and that parents
were free to choose any school they like. The model would
assume that parents had a good understanding of what they
wanted and what each school provided, leading to an
expectation of a close match between parental preferences
and school offerings. The model would also assume that
parents would not pay a higher tuition than necessary to
purchase a particular type and quality of service, leading to
an expectation that schools with similar offerings would have
similar levels of cost. Schools that did not keep their clients
satisfied would lose clientele to other schools. Although a
pure economic model is not completely satisfactory when
applied to educational institutions, it is still the baseline from
which proponents of the market model for education work
(Garner & Hannaway, 1982).
In addition to school efficiency as an effect of
harnessing economic market forces, proponents of the
introduction of the free market into the educational system
state that better schools will also result. It is to these
excellence arguments that this literature review now turns.
48


Excellence Arguments Concerning School Choice
Worldwide developments, particularly in the economic
sphere, have produced far-flung consequences for
educational policy and politics. Not only have roles and
relationships changed, but there has been a marked shift in
the fundamental values guiding American education (Clark &
Astuto, 1986). According to Kerchner and Boyd, the pursuit
of excellence has replaced equity as the leading goal of
American schooling ( Boyd & Kerchner, 1987). However, the
excellence movement as a policy issue is not just a
phenomena of the 1980s. Levin, writing about choice in
1968, states that competition between schools would induce
innovation and experimentation in that each school would try
to obtain competitive advantages Over the others. Any
measure of competition among schools would lead to
increases in their effectiveness and excellence. Levin points
out, however, that basic education yields two types of
benefitsthose to the student (private benefits) and those to
society-at-large (social benefits)and that proposals for
educational change through choice should be judged on the
basis of both criteria (Levin, 1968).
In terms of private benefits, Levin states that it is
likely to be true that any measure of competition among


schools would lead to increases in their effectiveness. The
motive for successprofit maximizationwould require that
a school meet the need of its students better than its
competitors for any given cost. The fact that existing policies
would have to be re-examined in the light of their
educational contributions would probably engender thorough
changes in the administration of the schools. By increasing
the number of decision-making units, the probability of
schools innovating to gain competitive advantages would be
far greater than under the present system. While many
examples of such change can be envisioned, a notable one
would be the introduction of those new curricula and
instructional aids which showed great promise relative to
their costs.
Another fruit of competition among schools might be
more imaginative recruitment policies for teachers.
Competitive schools would have to hire on the basis of
realities of the market place rather than on the basis of rigid
salary schedules. In addition, competitive schools would be
more likely to adopt a policy of flexible class size depending
upon subject matter, grade level, and type of student, which
is a more sensible goal than maintaining uniform class sizes.
Individual differences among teachers might be utilized as an
asset in the educational process by enabling teachers to
pursue their own teaching styles and approaches in place of
50


the present attempts of the schools to standardize curricula,
syllabi, and pedagogy along narrow guidelines.
Under a competitive market, one could probably expect
that greater educational benefits of excellence would accrue
to students and their families. However, Levin warns that
increases in private benefits may not necessarily yield
similar increases in social benefits. Schools shoulder the
primary burden for satisfying at least two social goals: Those
of imparting minimum levels of literacy, knowledge and the
common values necessary for a stable democracy; and of
decreasing disparities in incomes and opportunities
associated with race and social class. In evaluating choice as
a vehicle to increase school excellence, Levin cautions against
the possibility of imposing a system which would tend to
change the relative distribution of schooling alternatives in
such a way that the present disparities in income and
opportunities among social and racial groups would increase
(Levin, 1968).
In analyzing effective schools in relation to the pursuit
of excellence, Chubb and Moe (1988) state their belief that
schools are largely products of their environments and that
the fundamental determinants of school excellence are to be
found in the larger setting or system in which schools
operate. Until educational reforms move beyond a narrow
focus on schools per se, they are doomed to the treatment of


symptoms rather than causes. Both authors support the
research that suggests that private schools are more effective
than public schools at educating comparable students. Given
this premise, they go on to hypothesize that the key to
educational excellence is in providing choice by modeling
public school organizational features after private school
ones.
Chubb and Moe (1988) state that many differences
between private and public schools can be anticipated by
recognizing that the schools are controlled in fundamentally
different ways: one by politics and the other by markets.
Public schools are governed by democratic authority and thus
are properly controlled by and responsible to the people,
particularly local citizens and their representatives on the
school board. Schools are administrative subordinates in a
very complex system of political authority and control.
About this they have no real choice.
Private schools are not part of this system of
governance. Generally speaking, they are owned and
managed. In the typical private setting, it is much clearer
who is in charge and what the standards areand those who
disagree are free to seek educational services elsewhere. The
ready availability of the exit option is crucial, for when
students and parents are afforded the opportunity to make
choices among schools, a second mechanism of social control


or governance is at work: the market. Public schools are
almost always monopolists in their sectors and are largely
sheltered from market forces. Public schools have a semi-
captive clientele with little choice but to patronize the local
monopoly. Public schools are sheltered even further by the
fact that their funding comes from political authorities via
taxation rather than from parents as a fee for services
rendered. Parents may complain about the quality of
education, but they are not in financial control. According to
Chubb and Moe (1988), these fundamental differences in the
environments of public and private schools due to their
orienting structures of politics and market should have
pervasive implications for the ways in which schools are
organized and operated.
As participants in the "Administrator and Teacher
Survey" used to supplement the "High School and Beyond"
study, Chubb and Moe (1988) came to the conclusion that
private schools have more of the organizational attributes--
such as strong leadership, clear goals, a teamlike atmosphere-
-that are thought to enhance school performance than do
public schools. Their belief is that the organizational and
environmental context of private schools has pervasive
consequences for the organization and operation of all
schools. The key differences between public and private
environments--and thus between public and private schools-
53


-are anchored in their characteristic methods of social control.
Public schools are captives of democratic politics. They are
subordinates in a hierarchic system of control in which
myriad interests and actors use the rules, structures, and
processes of democracy to impose their preferences on the
local school. It is no accident that public schools are lacking
in autonomy, that principals have difficulty leading, and that
school goals are heterogeneous, unclear, and undemanding.
According to Chubb and Moe, the key to public school
excellence may well be to break the bonds of democratic
politics. If public schools are to develop the organizational
qualities that most research indicates are essential for real
improvements in education, it may be necessary to emulate
the system of control that governs the private schools, where
teaching and professional autonomy flourish. There are
various ways in which this might be carried out, but in
general terms, effective control over schools would be
transferred from government to the market. Government
would still set the minimum requirements. It also would
provide funding, but virtually all the important decisions
about policy, organization, and personnel would be taken out
of the hands of politicians and administrators and given over
to schools and their clients: the students and their parents.
In a system requiring competition for students and resources,
schools would have incentives to move toward more efficient
54


and effective forms of organization (Chubb & Moe, 1986).
While Chubb and Moe maintain that the introduction of
choice through the market system will force schools to
reorganize and thus improve, Wagner argues that a system
that permits parental choice would engage the natural
affection that parents have for their own children and
cultivate the responsible interest that they normally take in
their own childs education. This heightened parental interest
would improve the quality of education as parents became
more involved in their childrens educational experiences.
The quality of products offered in education, as elsewhere,
would increase as consumers became interested in them. In
a broad sense, opportunities for choice would surely
strengthen the sense of participation in the life of the
community by giving expression to the natural sentiments
regarding the involvement of one person in the life of
another (Wagner, 1977).
According to Wagner, limiting choice in education
lessens emotional commitments nurtured within the family,
and therefore probably weakens the importance of the family
in our culture. The public school movement socializes
parental responsibility, in this case for the future
development of children. In its impact on the family, the
situation brings to mind the ancient controversy between
Aristotle and Plato, a controversy which has important
55


implications for the present. Plato advocated that parents
should be prevented from knowing which children
biologically were theirs. By doing this, he thought that all
parents would come to feel and act parentally toward all
children. Aristotle, apparently with better understanding of
human nature, noted in his Politics that such a practice would
merely result in all parents acting with equal indifference
toward all children.
Public education without parental choice injects
vestiges of this Platonic ideal into our social order, eroding
the foundation of family life in the process. Restoring
elements of parental choice would reverse some of these
difficulties and thereby strengthen the sense of caring and
community and lead to an improved quality of education
(Wagner, 1977).
According to Boyd and Kerchner (1987), in general
there has been a widespread failure in the educational
establishment to appreciate how the politics of excellence
tends to promote demands for choice that will re-configure
educational politics and management. If the politics of
education are fundamentally ruled by excellence criteria,
then the politics of choice takes on a new character.
Vouchers and tax credits may be judged not according to
equity criteria, where their claims are weak, but according to
satisfaction criteria, where the proponents of broader choice
56


can connect their support for market and quasi-market
mechanisms to both consumer preference and higher
educational achievement. When viewed this way, client
choice is not an abandonment of public education. Rather
client choice becomes viewed as the provision of alternative
means of looking for excellence.
Consumer Choice Theory
As was stated in Chapter One, the purpose of this study
was to analyze and apply the consumer choice theory of
economics to the interdistrict enrollment option. The market
model economy outlined in the previous section of this
literature review was the underlying theory used to apply
basic economic principles to educational choice. Consumer
choice theory provided a framework from which to study
educational consumer behavior in the interdistrict enrollment
option. The basic approach to consumer choice taken in this
paper was to view the educational consumer as a processor of
information. The consumer was characterized as interacting
with his or her choice environment, seeking and taking in
information from various sources, processing this
information, and then making a selection from among some
alternatives. Ultimately, consumers make choices to
accomplish goals. Given some set of goals, the consumer
57


devotes attention to that information available which is
relevant to attaining those goals. The consumer then
interprets this information in light of previous knowledge
and the context in which the information is obtained; that is,
the consumer must decide what a particular piece of
information means (Bettman, 1979).
Several major comprehensive theories of consumer
choice, Engel, Blackwell, and Kollat (1978), Hansen (1972),
Howard and Sheth (1969), and Nicosia (1966), all have the
common focus on choice as a process and the belief that
choice has causes and can hence be explained. In this process
of making choices, consumers are aiming to achieve certain
purposes or to accomplish certain goals. In obtaining these
goals consumers have preferences for certain goods over
others. In applying this theory, economists assume rational
consumer behavior. Individual decision making consists of
comparing preferences with opportunities. The rational
individual arranges his or her affairs to best satisfy given
preferences from available opportunities. The rational
individual chooses so as to maximize his or her utility. The
preference that maximizes a consumers satisfaction is the
optimum for the consumer. A consumer optimum will occur
whenever the consumer maximizes satisfaction, given his or
her budget (Miller & Meiners, 1986).
58


This postulate of individual maximization, in itself,
remains empirically empty until further restrictions are
imposed on the definition of utility. Once this step is taken,
once the "goods" that the individual values are identified, the
way is open for the derivation of hypotheses that can be
tested against observations. The economic model is
predictive in content rather than prescriptive. The actors
who behave as consumers choose "more rather than less,"
with more and less being measured in units of goods that are
independently identified and defined. This becomes a
prediction about behavior in the real world that the
economist carries with him as a working professional scientist
(Buchanan, 1968).
When applying this theory to the interdistrict school
choice plan then, one had to identify the "goods" that parents
and students valued. This identification of individual values
j
then allowed the three hypotheses developed in Chapter One
to be tested against observations. For the purposes of this
study two measures of school value were identified. The first
was student achievement as measured by standardized test
scores and the second was the socioeconomic status of the
school's student body as measured by free and reduced lunch
data.


Student Achievement as Consumer Utility
Research supports the idea that some schools are better
or more effective than others. Researchers continue to
attempt to discover what it is that makes one school more
effective than another. A great deal of attention has been
given to distinguishing the levels of effectiveness and
attempting to determine what it is that causes one school to
be more effective than another. The Effective Schools
Research literature is the richest source of analyzing these
attempts. In this body of research, student achievement as
measured by adjusted standardized test scores was found to
be the best means of discriminating school effectiveness. It
follows then that if differences in standardized test scores are
good measures of effectiveness that when parents and
students choose more effective schools they will choose
schools with higher test scores. In order to more fully
understand the rationale behind the use of school
effectiveness and standardized test scores as consumer
utility, a review of the Effective Schools Research is in order.
The Effective Schools Research of the past decade
concludes that differences among schools do affect students
academic achievement. This literature challenges previous
research that had found unequal academic achievement to be
primarily a function of family background and related
60


variables (Coleman, 1966; Jencks, 1972). Purkey and Smith
(1982), in their "Synthesis of Research on Effective Schools,"
divided the school effectiveness literature into three groups
outlier studies, case studies, and program evaluation studies.
Outlier Studies
One major strategy of school effectiveness research has
been to statistically determine highly effective schools
(positive outliers) and unusually ineffective schools (negative
outliers). Most such studies employ regression analyses of
school mean achievement scores, controlling student body
socioeconomic factors. Based on the regression equation, an
"expected" mean achievement score is calculated for each
school. This "expected" score is subtracted from the actual
achievement level of the school to give a "residual" score for
each school. The researcher then selects the most positive
and the most negative residual scores and labels the schools
they represent an unusually effective or ineffective school.
Characteristics of these two types of schools are then assessed
by surveys or case studies to determine the reason for the
outcomes.
Studies that have adopted this general approach include
three carried out in 1974 by the New York State Department
of Education, a study conducted in 1978 for the Maryland


1974 study of model cities elementary schools and the 1977
study of Delaware schools by Spartz. The results of these
outliers studies were as follows. The most pervasive common
elements were better control or discipline and high staff
expectations for student achievement. Each of these
variables showed up in four of the seven studies for which
there were data. An emphasis on instructional leadership by
the principal or another important staff member was found
to be important in three studies.
Case Studies
Purkey and Smith carefully studied five school case
studies cited in various school effectiveness reviews
(Brookover and others, 1979; Brookover and Lezotte, 1979;
Rutter and others, 1979; Venezky and Winfield, 1979; Weber,
1971) and three other additions to the literature (California
State Department of Education, 1980; Glenn, 1981; Levine and
Stark, 1981). Six case studies in this group looked at urban
elementary schools. Five factors stood out as common to
most, but not all, of the six case studies. These were 1) strong
leadership by the principal or another staff member, 2) high
expectations by staff for student achievement, 3) a clear set
of goals and emphasis for the school, 4) a school-wide
effective staff training program, and 5) a system for the
monitoring of student progress. An emphasis on order and
62


discipline showed up in two of the studies, and a large
number of factors were specific to a single study.
The authors of the other two case studies took a more
complex look at the nature of effective schools than did the
previous six. Brookover and others (1979) observed two
matched pairs of elementary schools. One school in each pair
was high-scoring, the other low-scoring. The researchers
theorized that student achievement was strongly affected by
the school social system, which varied from school to school
even within similar subsamples with SES and racial
composition controlled.
The school social system was said to be composed of
three interrelated variables: 1) social inputs (student body
composition and other personnel inputs), 2) social structure
(such as school size, open or closed classrooms), and 3) social
climate (school culture as the norms, expectations, and
feelings about the school held by staff and students). While
school social inputs affect academic achievement, they are
modified in the processes of interaction with the school social
structure and school social climate.
An effective school was described as characterized by
high evaluations of students, high expectations, high norms of
achievement, with the appropriate patterns of reinforcement
and instruction in which students acquire a sense of control
over their environment and overcome the feelings of futility
63


which characterize the students in many schools (Brookover,
1979).
Program Evaluations
In this category Purkey and Smith looked at six
evaluations that examined school-level variables, Armor and
others (1976), Trisman and others (1976), Doss and Holley
(1982), and three studies carried out by the Michigan
Department of Education (Hunter, 1979). Armor and others
identified the school and classroom policies and other factors
that were most successful in raising the reading scores of
inner-city children who attended schools participating in the
School Preferred Reading Program in the Los Angeles Unified
School District. The Trisman study examined reading
programs in elementary schools throughout the nation. The
researchers surveyed a large number of programs and
carefully studied the characteristics of a few schools that had
especially successful efforts. Doss and Holley summarized
data from an evaluation of Title I programs in Austin, Texas.
The three Michigan studies were conducted from 1973-78 in
an attempt to understand what kinds of schools can carry out
effective compensatory education programs. '
By and large these studies were methodologically
stronger that the preceding two types of research. However,
64


their common findings were remarkably consistent with the
outlier and case studies. Most schools with effective
programs were characterized by 1) high staff expectations
and morale, 2) a considerable degree of control by the staff
over instructional and training decisions in the school, 3)
clear leadership from the principal or other instructional
figure, 4) clear goals for the school, and 5) a sense of order in
the school.
The previously cited studies were all of research in
elementary schools. There is a limited literature on
secondary schools that examines the policies and practices
associated with school effectiveness or school success. Wilson
and Corcoran (1988), in Successful Secondary Schools, single
out five studies for review. They include a study of basic
skills effectiveness in seventeen California high schools
conducted by the California Assembly Office of Research
(1984), the analysis of public and private schools conducted
by Coleman, Hoffer, and Kilgore (1982) using data drawn
from the "High School and Beyond" survey, six case studies of
public and private high schools by Lightfoot (1983), four case
studies of middle schools by Lipsitz (1984) and the study of
twelve London secondary schools by Rutter, Maughan,
Mortimore, Ouston and Smith (1979).
An analysis of the five studies produced twelve
dimensions of school effectiveness at the secondary level.
65


Many of them were very closely aligned with the dimensions
produced in the elementary Effective Schools Research. The
twelve dimensions were as follows: school-site management,
instructional leadership, curriculum articulation and
organization, schoolwide staff development, parental
involvement, schoolwide recognition of academic success,
maximized learning time, district support, collaborative
planning, sense of community, clear goals and high
expectations, and order and discipline.
Careful analysts of the Effective Schools Research find it
not without methodological problems. However, Rosenholtz
(1985) in "Effective Schools: Interpreting the Evidence" cites
three reasons to regard this body of findings as much more
than spurious. First, several studies describe "turnaround"
schools that, because of changes in organizational conditions,
became more successful. Second, even when controlling for
random error, analysts found that organizational
characteristics account for 32 percent of between-school
variance in student achievement. Third, Effective Schools
Research has been conducted within a relatively compressed
time frame, not building serially from one study to the next;
yet all studies produced common findings with remarkable
consistency. According to Rosenholtz, it strains the limits of
credibility that different studies, conducted by different
investigators in different urban areas, could produce
66


strikingly similar findings by chance.
This body of research produced in the past decade has
become widely accepted as a means to evaluate school
effectiveness. According to consumer choice theory, when
parents and students act as rational consumers in the
educational marketplace, they attempt to maximize their
utility and choose to attend schools they perceive are of
higher quality. In this study student achievement measured
by standardized test scores served as a measure of consumer
utility.
Socioeconomic Status as Consumer Utility
The socioeconomic status of the schools student body
was identified as a consumer utility based on the ample
evidence supplied by the work of Coleman (1966) that test
scores are a function to some degree of SES. It was also based
on the supposition that parents attempt to maximize school
utility by choosing to send their students to schools in which
they feel peer influence will be positive. The finding that
student body composition is an important determinant of
school effectiveness first gained prominence with the
publication in 1966 of the Coleman report, which emphasized
the positive relationship between the average socioeconomic
status of the students in a school and the academic
67


achievement of individual students. Subsequent studies
either replicated that finding or found the achievement of
individual students to be related to other characteristics of
the student body such as racial composition or the average
achievement level of the students in the class or school.
According to Murnane (1984a), experts differ in their
explanations of these relationships. However, an important
fact is that parents and school officials know that student
body composition matters, and this knowledge influences
their actions. In particular, when parents choose schools for
their children, they pay attention to who the classmates will
be; and when school officials admit students, they pay
attention to the attributes of those they admit. Using a
subset of data from the 1982 Coleman study, "High School
and Beyond," Murnane tested and found to be true the
hypothesis that the average socioeconomic status of the
student population plays a significant contributing role in
determining students' test scores (Murnane, 1984b).
Summary
Although there are many lines of inquiry when
studying the subject of choice in schools, this study chose to
follow the path of the application of consumer choice theory
to the interdistrict enrollment option. This theory was used
68


to predict the behavior of parents and students when they
chose to cross school district boundaries to attend school. It
was predicted that they would maximize utility as measured
by student achievement and the socioeconomic status of the
student body. The validity of the use of standardized test
scores as a measure of student achievement was based on
Effective Schools Research (Rosenholtz, 1985). The validity of
the use of the socioeconomic composition of the student body
as a measure of consumer utility was based on the work of
Coleman (1966).
69


CHAPTER 3
METHODOLOGY
The theory of consumer choice in the educational settin
provided the framework for this study. The application of this
theory to interdistrict schools of choice generated three
hypotheses.
Hypothesis One: When students cross school district
boundaries to attend school, the receiving school has a higher
percentile rank of the mean score on standardized achievement
tests than the sending school.
Hypothesis Two: When students cross school district
boundaries to attend school, the SES of the receiving school is
higher than the SES of the sending school.
Hypothesis Three: When students cross school district
boundaries to attend school, the receiving school has a higher
percentile rank of the mean score on standardized achievement
tests than the sending school when school populations are
controlled for SES.


A discussion of procedures, instrumentation, and
statistical tests will be developed in this chapter. A statement of
the limitations of the study will conclude the chapter.
Procedures
The population of this study was public school students
in El Paso County, Colorado, who crossed school district boundaries
in the 1989-90 school year to attend school. There were 446
students in the population. Units of analysis were these students
who crossed district boundaries to attend school. Each school was
coded and was given two values, one based on student
achievement as measured by the percentile rank of the mean
score on standardized achievement tests and one based on SES as
measured by the percentage of a schools population receiving
free and reduced lunch. Sources of data included the Colorado
Department of Education and the offices of the superintendents in
each of the individual school districts.
In order to assign the school codes, the following steps
were taken. The percentile rank of the mean scores on
standardized tests reported for each school in mathematics,
71


reading, and language were averaged to arrive at a school quality
value. The percentage of a schools population that received free
or reduced lunch was determined from information provided by
the office of the superintendent in each district. Each school was
then assigned a quality value based on this SES information. Each
of the students who crossed school district boundaries
represented a case and was given a case number. Each student
had four codes assigned to him or her a sending school quality
value based on the percentile rank of the mean score on
standardized achievement tests, a receiving school quality value
based on the percentile rank of the mean score on standardized
achievement tests, a sending school quality value based on the
percentage of students receiving free and reduced lunch and a
receiving school quality value based on the percentage of
students receiving free and reduced lunch. This resulted in the
development of the 446 cases that made up the population of the
study.
In order to test the three stated hypotheses, data were
collected by tracing the movement of students from the sending
school to the receiving school. Dependent sample t-tests were run
on the data to determine whether to accept or reject hypothesis
72


one and two. A Pearson product-moment correlation was run to
determine the feasibility of testing hypothesis three.
Instrumentation
Schools were coded by assigning the percentile rank of
the mean scores reported to the Colorado Department of Education
in each districts 1988-89 test result report. Schools were divided
into three grade-level categories -- elementary, junior high or
middle school, and senior high. Grade configurations were either
K-5 or K-6; 7-9 or 5-8; 9-12, or 10-12. At elementary and
junior high or middle school level, the following tests were used
by districts SRA Survey of Basic Skills (3 districts); California
Achievement Tests (2 districts); Iowa Test of Basic Skills (8
districts); and the Stanford Achievement Test (3 districts). At the
high school level the following tests were used -- SRA Survey of
Basic Skills (3 districts); Comprehensive Tests of Basic Skills (1
district); Iowa Test of Basic Skills (4 districts); Stanford
Achievement Test (1 district); and Tests of Achievement and
Proficiency (4 districts).
Elementary schools were coded for quality by averaging
73


the percentile rank of the mean scores reported by each district
in mathematics, reading, and language for grade four. Junior
highs or middle schools were coded for quality by averaging the
reported grade 7 percentile rank of the mean scores for
mathematics, reading, and language. Senior highs were coded for
quality by averaging the reported percentile rank of the mean
scores for mathematics, reading, and language for grade 10.
In addition, all schools were coded by assigning a
socioeconomic value measured by determining the percentage of
students in the building receiving free or reduced lunch.
Sub-Populations Analyzed
Ten subsets were analyzed for hypotheses one and two.
Subset one was composed of all 446 student cases. Subset two
was made up of student cases in which the same standardized
achievement test was used in both the sending and receiving
school. Subset three contained the student cases in which
different standardized achievement tests were used in the
sending and receiving schools. Subset four was made up of all the
cases in which students paid no tuition. Subset five was
74


composed of all the cases in which partial tuition was paid by
parents. Subset six contained all of the elementary student cases,
subset seven all of the middle school cases, and subset eight all of
the high school cases. Subset nine consisted of all the student
cases for which data disaggregated by individual schools were
available. One district refused to provide data disaggregated by
schools. The student cases from this district made up subset ten.
This district also refused to release the information about
students who attended school in this district but lived in another
district. The Colorado Department of Education reported 19
students in this category. These students were not included in
the study.
Statistical Tests
In order to test the three stated hypotheses, data were
collected by tracing the movement of students from the sending
school to the receiving school. Dependent sample t-tests were run
on the data to determine whether to accept or reject hypothesis
one and two. Dependent sample t-tests were chosen because two
observations were done on the same students. In addition, a
75


dependent sample t-test is more powerful than a simple t-test.
The use of the t-test allowed the determination of the difference
in the means of the pre and post scores. A Pearson product-
moment correlation was used to determine the feasibility of
testing hypothesis three. Mystat, a subset of Systat, was used to
compute the statistical analyses. The analyses were run on a
Zenith Data Systems computer.
Limitations of the Study
The instrumentation used in this study may have caused
some limitations. Standardized achievement test scores are not
the only measure of educational output. Other measures (student
portfolios, grades, measures of student attitudes, degree of
further schooling, graduation rates, attendance rates) could have
been used to indicate school quality; however, these measures
are not easily standardized nor are they readily available. The
use of a variety of test instruments may have made the results
less reliable. However, the same test is not used consistently
throughout the county.
A selectivity bias may have existed in the population.
76


Because Colorado is not an open enrollment state, the cases
studied were those students and parents who were able to
somehow negotiate an interdistrict transfer. A follow-up study in
open enrollment state such as Minnesota would be a valid
reliability check of this studys findings.
The selection of El Paso County, Colorado, based on access
to data was a limitation. Generalizability beyond El Paso County,
Colorado, was limited because of the small geographic area
represented. This was the study of the behavior patterns of
parents and students when they chose to cross district boundaries
to attend school. It was not a study of the reasons why they
chose to do so; therefore, no conclusions could be drawn as to the
reasons for the choice behavior.
77


CHAPTER 4
DATA ANALYSIS
The purpose of this chapter is to report and analyze the
data collected from the fifteen school districts in El Paso County,
Colorado, regarding the hypotheses set forth in Chapter One.
Hypothesis One
Hypothesis One states: When students cross school
district boundaries to attend school, the receiving school has a
higher percentile rank of the mean score on standardized
achievement tests than the sending school.
Dependent sample t-tests were run on the data about
446 students who crossed district boundaries in the fifteen
school districts in El Paso County during the 1989-90 school
year. Sending and receiving schools for each student were coded
by averaging the percentile rank of the mean scores reported in


mathematics, reading, and language for grades four, seven, and
eleven. Each case was assigned a sending school value based on
the percentile rank of the mean score on standardized
achievement tests and a value based on the percentage of
students receiving free and reduced lunch. For instance, if
student A lived in the school X attendance area of District 1, but
attended school Y in District 2 the percentile rank of the mean
score on standardized achievement tests for school X became the
sending school score used and the percentile rank of the mean
score on standardized achievement tests for school Y became the
receiving school score used. The same procedure was done using
free and reduced lunch data.
Ten subsets of data were createdall student cases,
student cases in which the same standardized achievement test
was used in both the sending and receiving school, student cases
in which the sending and receiving schools used different
standardized achievement tests, student cases in which no
tuition was paid, student cases in which partial tuition was paid,
elementary student cases, middle school student cases, high
school student cases, student cases in which building level data
were available, and student cases in which only district level
data were available.
79


Mean percentile rank scores on standardized
achievement tests were determined for the sending and
receiving schools in each subset. The difference was computed
and then dependent t-tests were run to determine whether to
accept or reject the hypothesis. The following tables provide
data for each subset.
Table 4.1. Comparison of Sending and Receiving Schools When
Student Cases Were Used.
Mean Test Score Mean Test Score Difference Degrees t Probability
Sending School Receiving School of freedom
58.03 65.42 -7.40 445 7.37 .000
A dependent sample t-test of all student cases (N=446,
t=7.37) was used to test the null hypothesis of no difference
between sending (58.03) and receiving (65.42) school means
against the one-tail alternative hypothesis that the percentile
rank of the mean scores on standardized achievement tests were
higher at the receiving school than at the sending school. As is
indicated in the above table the null hypothesis was rejected
(pc.000). Hence, it was concluded that the receiving schools had
a higher percentile rank of the mean score than the sending
schools.
80


The dependent sample t-test indicated that the 7.4
increase from the mean test scores of the sending schools to the
mean test scores of the receiving schools was statistically
significant. Although it is difficult to provide a precise way for
the consumer to judge the educational significance of this 7.4
percentile rank increase to the consumer, it may be useful to
consider what a 7.4 percentile rank increase means in this
subset of all student cases.
At grade 5 on one of the frequently used standardized
achievement tests, a percentile rank score of 58.0 corresponds to
a grade equivalent score of 6.0. A percentile rank score of 65.4
corresponds to a grade equivalent of 6.2. Thus, the 7.4
percentile rank difference in this subset would correspond to an
expected increase in student achievement of an additional two
months of school. This 7.4 percentile rank difference fairly
consistently represents a two month differential in student
achievement at all grade levels represented in this study. For
instance, at grade 10 a percentile rank score of 58 corresponds
to a grade equivalent score of 10.1. A percentile rank score of
65 corresponds to a grade equivalent of 10.3. Thus, a 7.0
percentile rank difference in this subset would correspond to an
increase in achievement expected of an additional two months.
81


This differential in student achievement means that
students in the receiving school achieved twenty percent more
on standardized achievement tests than did students in the
sending school. The question this study addressed was whether
parents and students maximized school quality as measured by
student achievement. It is conceivable that a two month
differential in student achievement is enough for students and
parents to recognize even if they do not compare test scores. In
elementary schools, for instance, parents might notice that
students are at different places in basal readers and that
students are making progress more quickly in mathematics. At
middle level and high schools similar indicators might be the
reading difficulty of text books, the difficulty of mathematics
courses, and the percentage of students pursuing further
education.
Thus it seems that along with being statistically
significant, the difference between the sending and receiving
school test scores is also apparent and educationally significant
to parents and students when they act as consumers. It also
appears that students and parents attend to perceived school
quality when they choose a school.
82


Table 4.2. Comparison of Sending and Receiving Schools When
Same Achievement Tests Were Used in Both Schools.
Mean Test Score Mean Test Score Difference Degrees t Probability
Sending School Receiving School of freedom
58.71 64.60 -5.97 171 3.98 .000
A dependent sample t-test of student cases (N=172,
t=3.98) was used to test the null hypothesis of no difference
between sending (58.71) and receiving (64.6) school means
against the one-tail alternative hypothesis that the percentile
rank of the mean scores on standardized achievement tests were
higher at the receiving school than at the sending school. As is
indicated in the above table the null hypothesis was rejected
(pc.000). Hence, it was concluded that the receiving schools had
a higher percentile rank of the mean score than the sending
schools when schools using the same test were compared.
One of the concerns regarding instrumentation stated in
Chapter Three was the use of a variety of standardized
achievement tests to measure student achievement. For this
reason, a data subset was created in which the sending and the
receiving school used the same standardized achievement test.
This created some control over the problem of multiple test use.
Although the difference between the means was not quite as
83


large in this subset (5.97) as it was in the all cases subset (7.4),
the difference was still both statistically and educationally
significant.
Table 4.3. Comparison of Sending and Receiving Schools When
Different Achievement Tests Were Used.
Mean Test Score Mean Test Score Difference Degrees t Probability
Sending School Receivine School of freedom
57.61 65.96 -8.34 274 6.22 .000
A dependent sample t-test of student cases (N=275,
t=6.22) was used to test the null hypothesis of no difference
between sending (57.61) and receiving (65.96) school means
against the one-tail alternative hypothesis that the percentile
rank of the mean scores on standardized achievement tests were
higher at the receiving school than at the sending school. As is
indicated in the above table the null hypothesis was rejected
(pc.OOO). Hence, it was concluded that the receiving schools had
a higher percentile rank of the mean score than the sending
schools when different tests were compared.
The mean difference in this subset (8.35) was larger than
the previous subset (5.96) indicating that the concern for
instrumentation validity was valid. The larger differential in the
84


mixed test subset may be a result of the variance measurement
between the different tests. However, as with previous data
subsets, the difference was both statistically and educationally
significant.
As data were collected for this study, one superintendent
discussed the possibility of a different behavior pattern being
exhibited by those students and parents who paid no tuition and
those students and parents who paid at least part of the tuition.
His speculation was that those paying for the opportunity to
maximize school quality would be more inclined to maximize to
a greater degree. For this reason the following two data subsets
were creatednon-tuition and partial tuition.
Table 4.4. Comparison of Sending and Receiving Schools When
No Tuition Was Paid.
Mean Test Score Mean Test Score Difference Degrees t Probability
Sendine School Receiving School of freedom
57.98 65.96 -8.02 361 6.73 .000
A dependent sample t-test of student cases (N=362,
t=6.73) was used to test the null hypothesis of no difference
between sending (57.98) and receiving (65.96) school means
against the one-tail alternative hypothesis that the percentile
85


rank of the mean scores on standardized achievement tests were
higher at the receiving school than at the sending school. As is
indicated in the above table the null hypothesis was rejected
(pc.OOO). Hence, it was concluded that the receiving schools had
a higher mean test score than the sending schools when non-
tuition cases were compared.
Table 4.5. Comparison of Sending and Receiving Schools When
Partial Tuition Was Paid.
Mean Test Score Mean Test Score Difference Degrees t Probability
Sendine School Receiving School of freedom
58.21 63.12 -4.90 83 3.27 .000
A dependent sample t-test of student cases (N=84,
t=3.27) was used to test the null hypothesis of no difference
between sending (58.21) and receiving (63.12) school means
against the one-tail alternative hypothesis that the percentile
rank of the mean scores on standardized achievement tests were
higher at the receiving school than at the sending school. As is
indicated in the above table the null hypothesis was rejected
(p<.000). Hence, it was concluded that the receiving schools had
a higher percentile rank of the mean score than the sending
schools when partial tuition cases were compared.
86


Although one might speculate that when students and
parents had to pay money to maximize the utility of student
achievement, they would exhibit a behavior pattern in which the
difference between the mean of the sending and receiving
schools would be greater in the partial tuition subset than the
non-tuition subset, this was not the case. At first glance this
appears to be a finding contradictory to consumer choice theory.
However, investigation of the two data subsets helps to provide
an explanation. In both the non-tuition and the partial tuition
data subsets parents were faced with a bounded choice option.
Districts, not parents and students, controlled whether
enrollment was free or whether there was a partial tuition
charge. This created circumstances in which consumers were
not in control of the type of choice options available. These
district policies regarding tuition versus non-tuition caused the
uneven choice distributions in these two data subsets. This
could be an explanation of what appear to be contradictory
findings. Further research needs to be done regarding tuition
versus non-tuition choices using populations in which the choice
was not bounded. It is very difficult to state with any certainty
from this research study whether payment of tuition influenced
the choice behavior of parents and students.
87


Table 4.6. Comparison of Sending and Receiving School When
Elementary Student Cases Were Used.
Mean Test Score Mean Test Score Difference Degrees t Probability
Sending School Receiving School of freedom
55.11 62.07 -7.02 183 3.54 .000
A dependent sample t-test of student cases (N=184,
t=3.54) was used to test the null hypothesis of no difference
between sending (55.11) and receiving (62.07) school means
against the one-tail alternative hypothesis that the percentile
rank of the mean scores on standardized achievement tests were
higher at the receiving school than at the sending school. As is
indicated in the above table the null hypothesis was rejected
(p<.000). Hence, it was concluded that the receiving elementary
schools had a higher percentile rank of the mean score than the
sending elementary schools.
Table 4.7. Comparison of Sending and Receiving Schools When
Middle School Student Cases Were Used.
Mean Test Score Mean Test Score Difference Degrees l Probability
Sending School Receiving School of freedom
57.65 66.11 -8.46 92 4.56 .000
A dependent sample t-test of student cases (N=93,
88


t=3.54) was used to test the null hypothesis of no difference
between sending (55.11) and receiving (62.07) school means
against the one-tail alternative hypothesis that the percentile
rank of the mean scores on standardized achievement tests were
higher at the receiving school than at the sending school. As is
indicated in the above table the null hypothesis was rejected
(pc.001). Hence, it was concluded that the receiving middle
schools had a higher percentile rank of the mean score than the
sending middle schools.
Table 4.8. Comparison of Sending and Receiving Schools When
High School Student Cases Were Used.
.Mean Test Score Mean Test Score Difference Degrees t Probability
Sending School Receiving School of freedom
61.39 68.71 -7.32 170 6.18 .000
A dependent sample t-test of student cases
(N=171, t=6.18) was used to test the null hypothesis of no
difference between sending (61.39) and receiving (68.71) school
means against the one-tail alternative hypothesis that the
percentile rank of the mean scores on standardized achievement
tests were higher at the receiving school than at the sending
school. As is indicated in the above table the null hypothesis
89


was rejected (pc.000). Hence, it was concluded that the receiving
high schools had a higher percentile rank of the mean score than
the sending high schools.
Table 4.9. Comparison of Sending and Receiving Schools When
Only Building Level Student Achievement Test Data Were Used.
Mean Test Score Sending School Mean Test Score Receiving School Difference Degrees of freedom t Probability
61.47 66.03 -3.98 367 3.74 .000
A dependent sample t-test of student cases (N=368,
t=3.74) was used to test the null hypothesis of no difference
between sending (61.47) and receiving (66.03) school means
against the one-tail alternative hypothesis that the percentile
rank of the mean scores on standardized achievement tests were
higher at the receiving school than at the sending school. As is
indicated in the above table the null hypothesis was rejected
(pc.000). Hence, it was concluded that the receiving schools had
a higher percentile rank of the mean score than the sending
schools when only building level standardized achievement test
data were compared.
90