University of Southern California Dissertations and Theses

Private school choice at the state level: diversity and success

(USC Thesis Other)

Private school choice at the state level: diversity and success

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content PRIVATE SCHOOL CHOICE AT THE STATE LEVEL:
DIVERSITY AND SUCCESS

by

Matthew Daniel Jones

________________________________________________________________________

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(POLITICAL SCIENCE)

December 2010

Copyright 2010 Matthew Daniel Jones
ii
ACKNOWLEDGEMENTS

This project would never have seen completion without the involvement of a
number of people working to gently – and sometimes not so gently – help shape my often
chaotic impulses, ideas, and writings into something with an intelligible and relatively
focused analytical purpose. Any remaining shortcomings only indicate the limits of their
incredible efforts and should in no way be associated with anyone listed here.
In all honesty, the entire Political Science department at USC as earned my
heartfelt gratitude and respect. Every person I worked with, from professor to staff, has
provided nothing but support for my work and fond memories of my graduate experience.
I would especially like to thank my advisor and friend Jeb Barnes, who, even though his
patience was constantly tested, never failed to make time to listen, provide invaluable
advice, and to advocate on my behalf. Thanks as well belong to Janelle Wong, who was a
source of encouragement as well as an incredibly enjoyable and helpful person to work
and talk with. This project would not have been possible without Ann Crigler’s
willingness to support its author on several levels and to seek funding support in the face
of budget pressure. I would certainly like to thank Eliz Sanasarian, Allison Renteln, and
Mark Kann, each of whom turned out to be wonderful people to work with as a TA and
wonderful people to get know in general. Finally, I would like to thank Howard Gillman
for his ability to challenge my thinking, research, and teaching in helpful ways and for
agreeing to serve on my original committee despite transitioning to a new position.
The support and encouragement I received from family and friends was humbling
to say the least. My parents, Bill and Betty Jones, and my sister Christy, never faltered in
iii
their desire to step in and help with short-notice child care, meals, and the innumerable
small gifts of time and spirit that only family can provide. I can only say that you are
appreciated more then I was ever able to tell you. To my wonderful in-laws Rich and
Joan, thank you for the interest and encouragement you demonstrated for my studies,
even when it seemed your daughter may have married someone who prefers school to
work. To the rest of my in-laws, Kim, Christine, the Ousey and Struck clans, and even
Tracy, I say thank you for your welcoming friendship, which provided an enjoyable
reprieve from the long years of graduate school. To my dear friends and pastors Paul and
Colleen Kuzma, your efforts to help me maintain the proper perspective and motivations
for my academic pursuits were valuable and often necessary. Finally, to my un-official
editor Katie, without whom this manuscript would contain an embarrassing number of
extra words that did not add anything of value, I am in your debt for the time and
encouragement you and your husband Josh so generously offered.
Last but certainly not least, I have no words that can adequately communicate the
depth of love and respect I have developed during this long process for my exceptional
wife, whose steadfast love and commitment went beyond what any husband could
reasonably ask for or imagine he should receive.

iv
TABLE OF CONTENTS

Acknowledgements ii

List of Tables vi

List of Figures vii

Abstract viii

Chapter One: Introduction 1
Defining Private School Choice 3
Existing Scholarship Relevant to School Choice Politics 7
Overview of the Analysis 17

Chapter Two: Making Sense of School Choice 22
Subsidy Mechanisms 24
Qualifications and Restrictions 29
Putting the Pieces Together 47
Chapter Two Conclusion 57

Chapter Three: Mapping Legislative Activity 58
State and Regional Distribution of Activity 59
Legislative Activity and the Distribution of State Characteristics 65
Chapter Three Conclusion 84

Chapter Four: Competing Explanations for Proposal Success 86
Political Science, Policy Chance, and Families of Explanation 87
Public Problem and Resource Explanations 91
Political Power Explanations 98
Content and Framing Explanations 107
Other Possible Explanations 115
Chapter Four Conclusion 121

Chapter Five: The Determinants of Success 123
Data Collection and Measures 124
Logit Results 128
Causation and Political Battles over Private School Choice Bills 134
Interpreting the Results 139
Chapter Five Conclusion 147

v
Chapter Six: Conclusion 149

Bibliography 157

Appendix A: Legislative Activity of Specific States from 1997 to 2007 163

Appendix B: Indicators of Public School Performance 165

vi
LIST OF TABLES

Table 2.1: Four Key Dimensions of Private School Choice Bill Design 24

Table 2.2: Distribution of Bills across Subsidy Mechanisms 27

Table 2.3: Legislative Progress by Subsidy Mechanism 28

Table 2.4: Distribution of Bills across Subsidy Value Qualifications 33

Table 2.5: Distribution of Bills across Population Targets 37

Table 2.6: Distribution of Bills across Cost/Participation Caps 44

Table 2.7: Major Policy Dimension Combinations 48

Table 2.8: Distribution of Subsidy Values across Subsidy Mechanism 50

Table 3.1: State Demographics across Bill Introduction Categories 67

Table 3.2: State Demographics across Bill Attention Categories 72

Table 3.3: State Public School Figures across Bill Introduction Categories 74

Table 3.4: State Public School Figures across Bill Attention Categories 76

Table 3.5: State Politics across Bill Introduction Categories 78

Table 3.6: State Politics across Bill Attention Categories 80

Table 3.7: Describing High/Low Bill Introduction and Bill Attention States 82

Table 4.1: Summary of Explanatory Families, Hypothesis, and Measures 90

Table 5.1: Private School Choice Bill Success: Logit Results 128

Table A.1: Individual Bill Activity by State 163

vii
LIST OF FIGURES

Figure 2.1: Distribution of Proposal Policy Impact 54

Figure 3.1: Private School Choice Bill Introduction Activity by State 61

Figure 3.2: Private School Choice Bill Attention Activity by State 63

Figure 5.1: Private School Choice Bill Outcomes by State 124

viii
ABSTRACT

The idea of using publicly funded vouchers to subsidize parents’ ability to pick
and choose education providers has been part of the contemporary policy debate for
decades. Beginning in the 1990s, this private school choice idea began to be translated
into specific legislative proposals showing up on the agenda of state legislatures with
increasing frequency, with a few states going so far as to enact some form of private
school choice law. Yet, scholarship on the issue has remained focused on debating the
substance, advocates, and constituencies of the general private school choice idea and
little attention has been directed toward understanding the diverse landscape of actual
legislation driving attempts to introduce private school choice reform into state education
policies.
To what extent does different legislation present policy designs with distinct
implications for changing existing public school policy arrangements and for the chances
of a bill becoming enacted into law?
This work uses an originally dataset of individual pieces of private school choice
legislation introduced into state legislatures from 1997 to 2007. Thus allowing it to
capture and analyze distinctions in private school choice policy design, the frequency of
individual bill introductions across state legislative venues, and the influence of both
policy design and venue context on individual bill enactment chances.
The analysis reveals that the private school choice idea has been incorporated into
legislation in a diverse range of different policy designs, and that bills were introduced
ix
into and considered by state legislatures in very uneven patterns. Logit estimation
indicated that the legislation most likely to get enacted into law tended to be designed to
restrict eligibility to students with a learning disability, or designed with specific and
restrictive funding or participation caps, and was introduced into states with
comparatively weaker public school systems and Republican controlled governments.
Overall the analysis suggests legislative policy design matters to bill enactment even
more then partisan politics but that legislative dynamics are so complex and idiosyncratic
that they defy easy explanation at the quantitative level.
Although the analysis largely confirmed a conventional understanding of policy
change as being complex, incremental, and path dependent, I argue that a systematic
study of individual pieces of private school choice legislation was also able to usefully: a)
identify certain bill content designs that more effectively communicate the limited and
supplemental nature of a reform proposal and influence bill success chances at the
margins and; b) reveal patterns of legislative activity across policy venues that defy
conventional wisdom, point to new puzzles of private school choice reform advocacy,
and identify more promising state cases for comparative analysis.

1
CHAPTER ONE:
INTRODUCTION

"If an unfriendly foreign power had attempted to impose on America the mediocre
educational performance that exists today, we might well have viewed it as an act of war.
As it stands, we have allowed this to happen to ourselves."

On April 26
th
, 1983, the Reagan administration released that sharp indictment of
America’s public school policies with a blue-ribbon panel report two years in the making.
The commissioned report, A Nation at Risk, seemed to document inadequate academic
performance throughout the public school system and recommended policy changes that
emphasized academic standards and school accountability. Coinciding with the report’s
release, President Reagan launched a public campaign for his education reform agenda,
which included a strong push for school vouchers, tuition tax credits, and other free-
market oriented reform policies.
Although he would ultimately be unable to get Congress to sign off on any school
voucher bill, Reagan’s criticism of public education and persistent advocacy for free-
market reform arguably energized a school choice movement that had remained on the
fringe of educational policy discussion since Milton Friedman initially championed the
use of publicly subsidized school vouchers to increase competition and individual choice
in education in the 1950s.
1
By the end of the 20
th
century, the push for school choice
policy reforms was supported by a multiplicity of politicians, think tanks, pressure

1
See Friedman’s 1955 article on “The Role of Government in Education”.
2
groups, and scholars at both the national and state level; between 1997 and 2007, at least
568 voucher or tuition tax subsidy proposals were attempted in almost every state
legislature in the nation with very mixed success.
This contemporary battle for school choice reform is a substantively and
politically important case of policy change advocacy. School choice has significant
implications for state budget and policy dynamics. Most states devote between one-third
and one-half of their overall budget appropriations to public education, and school choice
reforms deliberately seek to change the rules under which this vast pot of money gets
distributed to schools. An energized school choice movement’s ability to succeed, even in
limited ways, has the potential to start a process in which educational policy decisions
will be controlled to a greater extent by personal decisions and market forces instead of
democratic processes and administrative procedures. School choice reform also threatens
to redistribute large sums of money and control away from public school interests, such
as teachers and school boards, and toward private and religious institutions; therefore the
politics of debating school choice reform involve powerful and long entrenched interests
as well as issues of religious and moral diversity.
In order to better understand school choice, we must first understand the robust
diversity contained within the school choice policy family, the ways it presents itself to
decision-makers, and how and why it gets adopted. This project utilizes an original
dataset to systematically identify and analyze private school choice reforms at the state
level in an effort to address these issues and make progress in our understanding of
school choice politics. Along the way, the project also connects with broader issues
3
related to policy change and the importance of policy content and design in achieving
legislative success.
The rest of this introductory chapter sets up the analysis by clarifying the projects
focus on private school choice over public school choice; describing the way past
scholarship has approached the study of school choice politics by focusing on its mass
and elite level support; highlighting the scholarships’ empirical deficiencies in terms of
defining school choice as a policy issue and explaining why school choice legislation gets
adopted; and finally by pointing out the projects intentions to fill in the empirical gaps by
systematically analyzing individual private school choice legislation and its chances of
enactment.

Defining Private School Choice

The first step in understanding the politics of school choice is defining it. That
task is more complex than one might think, since the concept of school choice takes
shape in a diverse array of actual policy options advanced, proposed, and implemented
since Milton Friedman argued in 1955 that government-financed vouchers would
engender less expensive and more effective education than the largely government-
administered system currently in place. These policy options include reforms as limited
as inter-district public school open enrollment and as far reaching as statewide tuition
voucher programs financed by the state. Despite the range of policies that can be
associated with the school choice movement, they can be conceptually grouped into one
4
of two distinct categories of school choice reform options: public school choice and
private school choice. This project’s primary interest focuses on the latter type of school
choice policy options.
Public school choice involves policy options intended to create more educational
choices for parents within the existing framework of governmentally administered or
sponsored schools and includes inter/intra district public school enrollment and charter
school policies. These forms of school choice maintain a relatively direct connection
between public financing of schools and public administration of schools. For example,
laws that allow students to apply to any public school in their district, as opposed to
limiting them to attend the public school to which they are assigned, are an example of
public school choice through intra-district enrollment. Inter-district enrollment laws, on
the other hand, are a bit more expansive, allowing students to apply to public schools
outside of their local district. Both of these laws allow parents an increased choice in
education providers, since parents do not need to demonstrate a particular need to change
the public school their child is attending, though the choice is still limited to only public
schools that will prioritize students who would normally be assigned to the school.
Charter schools are operated by independent administrators under the supervision of a
government agency – usually a school board or division of the state’s education
department – and exist at the regulated discretion of a chartering authority.
Private school choice involves policy options intended to expand the educational
choices available to parents to include privately administered and sponsored schools.
Such options nearly always use government financing to directly or indirectly subsidize
5
private school tuition costs, and they generally include some version of publicly funded
scholarship programs (i.e., vouchers), tuition tax credits, or tax credits for donations to
privately funded scholarship programs. These forms of school choice are generally
considered more radical than their public school choice counterparts because they more
clearly weaken the connection between public financing of schools and public
administration of schools. The most well publicized private school choice law is the 1995
enacted Cleveland Voucher Program. This voucher law provided taxpayer funded
scholarships worth up to $3,000 to low income students assigned to Cleveland’s troubled
public schools. The voucher could be redeemed at any other public school in Cleveland
and participating suburbs and, more controversially, the voucher could also be redeemed
at private and parochial schools. The law was the subject of the U.S. Supreme Court’s
2002 Zelman vs. Simon-Harris case, in which the Court decided that state sponsored
scholarships subsidizing tuition at religiously affiliated schools does not violate the
Establishment Clause of the First Amendment as long as the public money is funneled
through parents making a free choice to send their child to the school.
Although both forms of school choice are part of the overall school choice
movement, there are clear differences in the policy and political implications associated
with the two policy types. Public school choice options like open enrollment and charter
schools have been steadily gaining acceptance and implementation by state governments
since the 1980s and can be considered the more incremental and less politically divisive
way to introduce a measure of school choice into an existing state education system. The
more mainstream status of public school choice is illustrated by the fact that only 11
6
states did not have some form of charter school enabling law in place by 2007, as well as
the fact that Democratic presidents such as Bill Clinton and Barack Obama have been
willing to publicly endorse charter schools while maintaining their opposition to private
school choice policies. Advocates for increased parental choice in education of all
ideological stripes usually support public school choice policy options; it is private school
choice options whose support is more likely to be restricted to those organizations,
philanthropists, and political leaders most ideologically committed to free-market
principles playing a larger role in policy outputs, a commitment conventionally
associated with political conservatism.
Politically conservative supporters tend to prefer private school choice options
because they are perceived as having greater potential to reduce political interference and
bureaucratic regulation in the provision of education and increase the power of private
decision making and producer competition. A private school choice reform that is limited
in its scope, cost, and impact on policy outputs still introduces the critical policy core
dynamic of public education resources being used, or allowed to be used, to subsidize the
individual choice of parents to send their child to a private education provider, something
even a robust charter school law does not accomplish. Such a reform allows public
money to be used to pay for an education and to support a curriculum that is not under
public control. This dynamic – if accepted and if expanded – has the potential to
fundamentally alter the governance of education and tilt it toward a default position of
private service providers, personal choices, and individual market dynamics instead of
government providers and public control.
7

Existing Scholarship Relevant to School Choice Politics

A considerable amount of energy, scholarly and otherwise, has been directed
toward evaluating the various moral, economic, and legal implications of school choice
as an education policy reform. In comparison, a significantly smaller amount of attention
has been directed at studying efforts to actually get a private school choice program
through the policy process and approved by political decision-makers. Accordingly, the
literature on school choice provides substantial discussion – though little consensus –
about the potential costs and benefits of adopting a school voucher program, but not a lot
of discussion about how legislatures adopt such a program in the first place. This is
somewhat disappointing, considering the substantive importance of the issue and its
utility as an example of major policy change in the face of entrenched and powerful
interests.

School Choice Scholarship
Although there is some general agreement that successful private school choice
proposals will likely result in private choices, market forces, and religious institutions
playing a larger role in the provision of primary and secondary education, there is no such
agreement about whether this will positively or negatively impact the provision of
education. Some scholars see the increase in private and market choices as injecting
healthy competition into the educational system, increasing school efficiency, and
8
empowering parents by reducing the over-bureaucratization that has resulted from
democratic bargaining (e.g., Chubb and Moe 1990). Others see them as a catalyst for
increased segregation and elitism and undermining the role public debate and community
discussion should play in resolving educational problems (e.g., Henig 1994; and Smith
and Meier 1995).
2
Add to the pile a growing number of research studies, commissioned
or conducted by advocacy organizations, think tanks, and other school choice partisans,
both inside and outside academia. Indeed, the policy differences over school choice seem
significant enough that most research focuses on the normative and empirical policy
implications of changing the established system, increasing the influence of market
forces, or reducing the influence of democratic and bureaucratic mechanisms, for the
public provision of education on school performance.
A smaller subset of school choice scholarship concern’s itself with understanding
the political dynamics involved in advancing school choice program. This scholarship has
primarily utilized surveys, interviews, and case studies in order to get a handle on the
elite and mass actors, constituencies, as well as coalitions that rally around the choice
reform flag and the tactics that have been employed to promote it.
At the national level, Terry Moe (2001) surveys 5,000 respondents in an effort to
capture a national cross-section of opinion on vouchers and provide a thorough analysis
of the demographics of voucher support. Gokcekus, Phillips, and Tower (2004) analyze
the constituencies and backgrounds of Congressional representatives who voted for an

2
Considering these intended changes implicate critical value priorities about how to best achieve
educational and societal goals, I believe these characteristics of school choice reform in general, and
especially private school choice reform specifically, satisfy Sabatier and Jenkins-Smith’s (1993 and 1999)
criteria for major policy change – i.e., changes in the policy core aspects of a governmental program.
9
amendment to the No Child Left Behind Act of 2001 that would have allowed money
from the Title IV program for disadvantaged students to go toward a private school
voucher program.
At the state level, Brunner, Sonstelie, and Thayer (2001) analyzed precinct vote
totals in Los Angeles County for the voucher program proposed in California’s 1993
referendum vote, while Brunner and Sonstelie (2003) used survey response data to
analyze voter support for the voucher referendum put before California voters in 2000.
Morken and Formicola (1999) interview an impressive number of individuals both inside
and outside of elected office, who are committed to promoting school choice policies in
more than a dozen cities and states. The result is an impressive collection of details about
real reformers, policy battles, and strategies. Michael Mintrom (2000) focuses his
analysis on the importance of school choice policy entrepreneurs to the chances of all
forms of school choice reform adoption, and uses a mail survey sent to people involved in
state educational policy communities in 48 states to identify key policy entrepreneurs
with which to conduct targeted interviews. These interviews are combined with in-depth
case studies of school choice battles in Oregon, Michigan, and Nebraska to illustrate the
qualities of an effective policy entrepreneur. Finally, Lance Fusarelli (2003) merges a
series of Texas policymaker interviews with legislative testimony, committee hearings,
floor debates, and newspaper stories, among other sources from various other states, in
order to provide specific examples of how theories of interest group dynamics and
advocacy coalitions have played out in school choice conflicts.
10
A few studies moved beyond case studies and interviews and sought to
understand why school choice policies were considered and adopted more generally.
Lawrence Kenny (2005) summarizes the various state voucher bills and ballot measures
enacted through 2003 and discusses the different political, demographic, and policy
design characteristics associated with them. Michael Mintrom’s (1997 and 2000)
provides a more robust and systematic analysis of school choice policy adoption in the
forty-eight continental states by utilizing event history analysis, with state consideration
of school choice and state approval of school choice as the dependent variables. In both
studies, Mintrom includes both political and socioeconomic measures in his effort to
analyze the influence of school choice policy entrepreneurs on the states chances of
adoption.
This scholarship has certainly added to the overall pool of knowledge about the
dynamics school choice politics, especially in terms of the individuals and groups both
inside and outside of political office that advocate for school choice.
School choice advocacy can involve a confusing and shifting set of
constituencies with different motivations for their support. Free-market conservatives
believe the private sector will produce better and more cost efficient schools (Fusarelli
2003, Brunner and Sostelie 2003); religious conservatives seek alternatives to secular
education (Moe 2001); inner-city activists and families see choice as a way to improve or
opt out of failing urban schools (Moe 2001; Fusarelli 2003); low income minorities and
African-Americans often view the issue in terms of educational equity and the chance to
opt out of failing schools (Gokcekus, Phillips, and Tower 2004). Republicans at both the
11
mass and elite levels are consistently inclined to support all forms of school choice (Moe
2001; Brunner, Sostelie, and Thayer 2001; Brunner and Sostelie 2003; Gokcekus,
Phillips, and Tower 2004; and Kenny 2005).
The importance of individual policy entrepreneurs both inside and outside elected
or administrative office for mobilizing these constituencies and placing school choice on
the policymaking agenda is compellingly highlighted in both narrative (Morken and
Formicola 1999) and systematic form (Mintrom 1997 and 2000). The micro-level tactics
used by school choice policy entrepreneurs tend to be very case specific and designed to
cobble together disparate pockets of reform supporters into a winning coalition for a
particular school choice policy (Morken and Formicola 1999).
Although the Democratic party has not characteristically been a supporter of
school choice policies, individual Democrats inside and outside of elected office have
been known to be vigorous and vocal school choice advocates (Morken and Formicola
1999; Mintrom 2000), with certain core Democrat constituencies comprising part of
school choice reform coalitions (see inner-city activists and African-Americans, above).
The most consistent, attentive, and important opponents of school choice are teachers
unions and public school administration, who collectively benefit from existing public
school arrangements more than any other set of interests (Mintrom 2000; Brunner,
Sostelie, and Thayer 2001; Gokcekus, Phillips, and Tower 2004). Other potential
opponents of school choice include suburban voters and voters located in areas attached
to high performing public schools with high test scores, who worry about the effect
12
school choice (particularly vouchers) will have on their neighborhood school and area
property values (Brunner and Sostelie 2003; Kenny 2005).

The Limits of Existing Scholarship for Understanding Private School Choice Success
The fledgling scholarship studying school choice political battles has indeed
collected considerable information about school choice politics; however, our
understanding of the reasons behind successful private school choice legislation remains
rather partial. This is in significant part because the current scholarship does not
adequately account for private school choice as a distinct sub-population of school choice
policy proposals, nor does it adequately account for the complex range of design
characteristics presented by different private school choice proposals. It tends to
conceptualize school choice too generally, ignore its internal policy distinction, and
approach its diversity unsystematically. In terms of overbroad school choice
conceptualization, the studies conducted by Morken and Formicola (1999), Mintrom
(1997 and 2000), and Fusarelli (2003) include the public school choice reform activity of
primarily charter school policies and public school open enrollment, along with the
private school choice reform activity of vouchers and tax credits. This approach ends up
capturing too broad a population of school choice policy, blurring very important policy
and political distinctions between differing school choice designs. The studies provided
by Moe (2001), Brunner, Sonstelie, and Thayer (2001), Brunner and Sonstelie (2003),
and Kenny (2005) do consider private school choice proposals independently, but focus
only on voucher policies as a generic stand-in for generalizing about all private school
13
choice reform. This approach does not adequately account for the subtle but significant
distinctions between the type, size, and scope of different private school choice policies
(as will be further detailed in Chapter Two).
Another gap in the existing literature’s analysis of school choice politics is that,
with the exception of Mintrom’s (1997 and 2000) studies of the significance of policy
entrepreneurs to state policy innovation, the existing studies have not been designed as
systematic attempts to explain the sporadic pattern of success in state adoption of private
school choice policies. Most are surveys and case studies that focus their analysis on
specific reform attempts – such as California’s referenda, the No Child Left Behind Act,
or school choice as an abstract reform idea – and have been designed primarily to map
the sources of school choice support among mass and elite actors, along with strategies
and tactics employed by reformers. With the exception of Mintrom (1997 and 2000) and,
to a lesser extent Kenny (2005) – though Kenny’s analysis is more of a hybrid literature
review using some general descriptive statistics as support – current studies are not
designed to explain why some form of private school choice legislation was attempted in
almost every state, some more vigorously than others, but only enacted in a few.
There is, however, a relevant set of policy change scholarship that does concern
itself with policy innovation and policy diffusion among states. Although this scholarship
does not focus on school choice reform in particular, it does set out to understand why
states adopt policy innovations, of which school choice reform can certainly be
14
considered an example.
3
Indeed, Mintrom’s work actually represents an application of
this type of scholarship to the issue of school choice. These types of state policy studies
usually emphasize the economic, social, and political characteristics of states in
explaining variation in policy innovation (e.g., Dye 1966; Hofferbert 1974). As Gray
(1994) points out and Miller (2004) summarizes, these studies “typically focus on factors
such as state fiscal capacity, short-term economic conditions, election cycles, inter-party
competition, party control, legislative professionalism, education, political culture,
religion, and urbanism”. The problem is that although state policy innovation studies can
inform us about potential state socioeconomic and political characteristics that encourage
innovation in general, and Mintrom’s work can inform us about how some of those
characteristics are relevant to school choice innovation more specifically, neither can
inform us about how the variation in the way school choice legislation is designed can
relate to how and why it passes.
In fairness, none of the scholarship on school choice politics or state policy
innovation has needed to attempt an accurate and systematic consideration of variation
and complexity in the private school choice proposal population to accomplish their
authors’ intended goals. The primary thrust of the scholarship on school choice politics
has been to try and get a handle on the dizzying array of political and policy actors,
organizations, constituencies, coalitions (both stable and those of convenience), and
tactics involved; meanwhile, the focus of analysis for studies of policy change has

3
A policy innovation was first defined by Walker as a “program or policy which is new to the states
adopting it, no matter how old the program may be or how many other states may have adopted it” (Walker
1969, pg 881).
15
typically been the decision-making unit choosing to adopt or reject a change in policy,
such as state governments, congress, or courts. While these pursuits have contributed
much to our understanding of school choice political dynamics, they remain curbed in
their ability to explain private school choice policy change because private school choice
proposals, firstly, present a distinct, and more strongly free-market set of policy reforms
than public school choice policies and, secondly, contain a robust range of design
features that can be incorporated in different combinations within a particular policy
proposal.

Implications that Private School Choice Policy Design is Important
School choice policy distinctions and diversities significance on the consideration
and passage of reform proposals is an issue of inquiry where school choice scholarship
has merely scratched the surface, but it has scratched the surface. The various
constituencies and their associated motivations identified by the existing scholarship
indicates one way in which policy design variation can be significant. Terry Moe’s
(2001) survey analysis indicates that different arguments were favored by different
constituencies. The free-market arguments of increased competition and efficiency
generated favorable responses among ideologically conservative supporters but not
among low-income and inner-city constituencies. The latter groups were more likely to
respond favorably to arguments about inequality, unaccountability, and regulation in
public schools. Although this analysis does not tackle school choice design features
directly, when connected with the work of Levin (1987, 1999, and 2002) its implications
16
become clearer. Levin has given considerable thought to the different ways school
vouchers can be framed around different educational goals as well as the broad design
characteristics that can affect these goals’ success. He argues that the debate over school
choice and vouchers can be considered in part as a conflict between proponents and
opponents over four educational goals: a) freedom of choice, b) productive efficiency, c)
equity, and d) social cohesion. Voucher advocates tend to prioritize “a” over “d”, whereas
opponents prioritize in the reverse, and perceiving the effect of vouchers on “b” and “c”
differently. He argues that different private school choice programs are designed with
provisions that will promote some of these objectives while undermining others, therefore
no program can appeal to all constituencies at once. Levin’s work intends to reveal how
framing and designing school choice policy reform proposals can encompass important
trade-offs in educational goals and implicate different public policy outcomes, not to
connect the dots to school choice politics or adoption. However, his work highlights how
including different provisions in different private school choice proposals can affect the
educational goals any given proposal can be effectively claim to be contributing toward,
which, in light of Moe’s conclusions, may in turn help mobilize potential constituencies
in support.
Other analyses that implicate the broader significance of school choice policy
design in school choice politics and adoption include the work of Morken and Formicola
(1999), who discuss the ways in which key local advocates connect reform coalitions to
particular school choice policies, and that of Mintrom (2000), who finds that school
choice entrepreneurs can be more effective when they are willing to accept public school
17
choice options (such as public school open enrollment) as a fallback position when more
ambitious reforms stall. Although Kenny’s (2005) analysis is a bit limited to draw any
strong conclusions, he provides data that suggests a proposed voucher law’s features
matter for its passage; in particular, provisions to prevent participation of upper income
and existing private school families, as well as to target large, struggling schools, stand
out as promising features for success.
In order to more firmly grasp school choice politics, a study specifically designed
to systematically account for the population of private school choice policy options
available and analyze what relationship, if any, variation in individual bill design may
have on explaining the patterns of private school choice adoption seems warranted.

Overview of the Analysis

Intended Contributions
This project is based on the assumption that variation in proposal content matters
for decisions to adopt or reject policy change in ways that, rather than idiosyncratic, can
be discerned as relatively stable patterns significant to interpreting the reasons behind
policy change in an issue area. Under this reasoning, the project asks the following
questions:
• What is the range of forms adopted by private school choice as it is
incarnated and reincarnated into specific legislative proposals in different,
states, legislatures, and legislative sessions?
18
• Which forms does it seem to favor more than others, if any?
• What types of states and legislatures does it appear in more often?
• Which forms in which types of states and legislatures are more likely to
win the support of an eclectic set of decision-makers in a fractious
legislative process?
In the course of answering these questions, this project utilizes an original dataset
of 568 individual private school choice bills introduced into state legislatures from 1997
to 2007 that will allow it to: a) provide a systematic and detailed description of
contemporary private school choice legislation and its distribution across state policy
venues; and b) include policy design variation in an attempt to explain why state
lawmakers have chosen to adopt or reject private school choice reform.
This project’s approach to the politics of school choice reform explicitly
recognizes private school choice policies as a substantively and theoretically distinct
subset of school choice and focuses the analysis on the more ambitious free-market
education reforms they present. It is more comprehensive and specific in its consideration
of private school choice because it systematically distinguishes between available
features and provisions, and combinations of those features and provisions, in its attempt
to explain private school choice bill adoption.
By collecting all instances of private school choice bills explicitly introduced into
state legislature,
4
the project paints a clearer picture of what the population of private

4
That is, the bill was introduced as an independent proposal sent to a substantive committee for
consideration, as opposed to an amendment to an existing bill under floor consideration or a subtle attempt
19
school choice proposals has actually looked like in recent history, including its major and
minor families of characteristics and important sub-groupings, and explains how that
population has been distributed in varying concentrations across states. This type of
description has not been attempted before and promises to illuminate some internal
dynamics relevant to understanding the advocacy activity of a major policy change issue
that shows no signs of going away any time soon. This is especially important in terms of
understanding the extent to which private school choice reform design ambitiously or
weakly challenges existing public school dynamics.
To unlock the empirical complexities within private school choice in a useful
manner for explaining state level decisions to adopt private school choice reform, this
project’s analysis shifts focus from people to content. Individual policy reform proposals,
rather than the people deciding upon those proposals, has become the unit of analysis in
order to properly consider whether proposal content and policy design variation is
important when it comes to explaining state policy innovation on the issue of private
school choice reform. This project systematically accounts for the range of potential
proposal design features, shifting the focus of analysis from the decision-makers who
choose whether a policy is passed to the individual policy reform proposals themselves,
thus illuminating the relationship between previously ignored variation and the proposal’s
chance of passage.

to create new policy through the budget process. Bills originally introduced as independent proposals but
subsequently converted to amendments to either another bill or to the budget also comprise collected data.
20
Overview of the Chapters
This chapter has defined the boundaries between public and private school choice
and clarified the project’s focus on the politics of private school choice policy reforms. It
has also laid out the contributions to our understanding of private school choice politics
made by existing literature on school choice political dynamics, demonstrated the limits
of this scholarship, and highlighted the added value of approaching the robust diversity of
private school choice policy design more systematically and studying school choice as a
state policy innovation through an analysis of individual proposals instead of personal
stories of decision-making units.
Chapter Two takes on the task of mapping out the diverse landscape of private
school choice proposals and laying the groundwork for the significance of policy content
to explain why some legislative proposals are adopted and others are not. Drawing from
the collected dataset of 568 bills introduced between 1997 and 2007, the chapter details
the major ways design particulars can vary across proposals in terms of their effect on the
type, size, scope, and cost of the private school choice funding offered and outlines the
more common ways these features combine to create different private school choice
outcomes and implications.
Chapter Three shifts gears from mapping out the landscape of what kinds of
private school choice policies were proposed to mapping which states and legislatures the
proposals were proposed in, providing a description of high-activity versus low-activity
states and a discussion about which state characteristics seem to attract more private
school choice bill activity.
21
Chapter Four reviews the plausible alternative explanations for the variation in
private school choice policy adoption across states and lays out the specific set of
hypotheses that strive to interpret private school choice policy adoption patterns. The
evidence for these hypotheses is drawn from the descriptions of policy diversity and state
bill activity in previous chapters, as well as from the literature on policy change and
comparative state policy innovation.
Chapter Five will utilize logit analysis to test the hypotheses’ ability to explain
why a few bills were enacted into law while so many others were not. The main focus of
the chapter will be to discuss, explain, and elaborate on the results of the logit analysis.
Chapter Six kits together the lessons and conclusions from previous into general
conclusions about the politics of private school choice policy change, the political
significance of private school choice policy design, and the political and strategic
significance of policy design for major policy change more broadly.

22
CHAPTER TWO:
MAKING SENSE OF SCHOOL CHOICE

From January 1997 to December 2007, at least 568 individual bills proposing a
form of private school choice were introduced into state legislatures. These individual
policy proposals comprised a spectrum of policy approaches that varied along four key
dimensions: subsidy mechanism, value qualifications, target populations, and budget and
participation caps. (See Table 2.1 for a summary.)
These dimensions can be combined in a multitude of ways, resulting in a rich
landscape of policy proposals, ranging from modest bills, like Missouri House Bill 345 to
very ambitious bills, like Arkansas House Bill 2275. The private school choice program
proposed in the 2003 Missouri bill would have facilitated private school choice only
indirectly through a tax credit worth 50 percent of a donation to a private non-profit
organization that would provide tuition scholarships for students from low-income
families. The bill limited funding for the tax credit program to $5 million out of a state
budget of almost $8 billion, and also specified that the non-profit could not offer
scholarships worth any more than $3,400, or roughly 45 percent of what the state
provides public schools per student. By contrast, Arkansas’ 1999 private school choice
bill would have offered every student in the state a voucher scholarship directly funded
from the state budget with no restrictions on its funding and worth the entire cost of
private school tuition up to the full amount the state would spend to send a student to a
public school.
23
The difference between these two bills is staggering in terms of the number of
school-age children able to attend private schools with public support, and in terms of the
threat it poses to the existing public education system and interests. Indeed, the space
between these bills is comprised of a bewildering set of possible policy combinations. In
order to impose some order on this diversity, this chapter discusses and identifies the
outer boundaries and internal partitions of the private school choice landscape as found in
recent legislation. It first discusses each of the four policy dimensions and their policy
significance individually; it than brings these separate pieces together to summarize the
range of policy proposals attempted at the state level during the 11 years between 1997
and 2007. The picture of private school choice activity that emerges from this process is a
policy reform effort that contains both very aggressive and very meek policy proposals
but, more often than not, pursues policy combinations that would introduce free-market
policy dynamics in more modest and incremental steps. The majority of legislative
proposals were limited in their potential policy impact and seem more likely to be
designed as a supplement, rather than a challenge, to existing public school policy
arrangements.

24
Table 2.1: Legislative Progress of All Private School Choice Bills

Subsidy Mechanism

Value Qualifications Target Populations Budget and
Participation Caps
• Voucher
• Direct Tax Credit
• Indirect Tax Credit
• Percent of State Per
Pupil Funding
• Specific Dollar
Amount
• Students with
Disabilities
• Low Income
Students
• Urban School
Students
• Underperforming
School Students
• Overcrowded
School students
• Hard Caps on
Spending
• Soft Caps on
Spending
• Hard Caps on
Participation
• Soft Caps on
Participation

Subsidy Mechanisms

The first and most fundamental distinction between private school choice policies
involves the method in which a proposal uses public funds to subsidize private school
expenses. Any given proposal may or may not contain value qualifications, target
populations, or budget/participation caps but all private school choice proposals build
upon one of three basic private school expense subsidy mechanisms:

1. Voucher Subsidies are state-funded scholarship programs that allow some
portion of the state’s school age population to attend private schools by paying
some portion of the tuition costs.
2. Direct Tax Subsidies allow parents to decrease some portion of their private
school education expenses (not always including tuition) from their state tax
burden, either through a tax credit or a tax deduction.
25
3. Indirect Tax Subsidies allow individuals and/or corporations to reduce their state
tax burden through contributions to privately-funded scholarship programs that
pay for private school tuition costs. Since charitable donations are generally
already eligible for a tax deduction, this mechanism is almost always a tax credit.

The first two types of private school cost subsidies – vouchers and direct tax
subsidies – seek to use public funds to directly subsidize the choice of parents to send
their children to a private school. This is in contrast to indirect tax subsidies, where
public funds must first be funneled through charitable donations to private scholarship
organizations, creating a more indirect use of tax money to subsidize private school
expenses.
Although vouchers and direct tax subsidies are often considered two sides of the
same coin, or just slightly different ways to achieve direct public subsidization, there are
significant policy implications for choosing one mechanism over another. The most
obvious and policy-significant distinction between the two mechanisms is that tax credits
utilize taxation dynamics for subsidization and apply after school expenses have already
occurred, while vouchers typically reduce the immediate expenses incurred by parents.
The potential policy consequences of this distinction are not lost on school choice
advocates. Some advocates argue that by retaining a greater degree of parental financial
responsibility, expense credits are more likely to increase school responsiveness to
families, reduce fraud, and reduce the ability of government administrators to impose
politically-motivated restrictions on public subsidies. Others argue that by funneling
26
subsidization through taxation dynamics, direct tax subsidies are less likely to be utilized
by the low-income families who could most use them but cannot afford the upfront costs.
They are also critiqued as restrictive in their potential application because they must be
attached to specific tax revenue streams and as an improper use of the tax code for
purposes of social engineering.
5

Despite the meaningful distinctions, both policy options still propose direct
subsidization, unlike an indirect tax subsidy. Arguably, the direct use of public funds for
private school expenses facilitates more privatization and, if parents choose parochial
schools, creates visible links between taxpayer money and religious instruction. This
implicates vouchers and expense credits as the more politically and ideologically
controversial private school choice policy option when compared to donation credits.
Direct subsidization policy options may also have a greater impact on a state’s
established education system and education budget. Practically speaking, the number of
people willing to incur little to no upfront cost in sending their children to private school
likely surpasses the amount of people willing to be reimbursed for the expense months
later, and both of these groups are probably larger than the amount of people and
businesses willing and able to receive delayed reimbursements for donating to subsidize
the private school expenses of other peoples’ children.

5
For a more thorough explanation of the different approaches to private school choice see Andrew
Coulson, Forging Consensus: Can the School Choice Community Come Together on an Explicit Goal and
a Plan for Achieving It?, Mackinac Center for Public Policy, 2004
27
Distribution of Subsidy Mechanisms among Bills Introduced
The data displayed in Table 2.2 describes the frequency with which each type of
private school choice subsidy mechanism shows up among the total private school choice
bills introduced during observation. Voucher proposals make up about half of all the
private school choice bills collected, even though they are arguably more likely to
generate controversy and opposition than the other two basic types of subsidy
mechanisms. Proposals that reimburse private school expenses already incurred by
parents (i.e., direct tax credits) made up a little over half the non-voucher bills introduced,
with donation reimbursements composing the remaining portion.

Table 2.2: Distribution of Bills across Subsidy Mechanisms

Type of School
Choice Policy Option
# of Bills
including
provision
% of bills
including
provision
# of states with at
least one bill that
includes
provision
% of states with
at least one bill
that includes
provision

Voucher

281

49.5%

42

85.7%
Expense Credit

160 28.1% 39 79.6%
Donation Credit

127 22.4% 37 75.5%
Total 568 100.0% 49

100.0%

Table 2.3 goes beyond introduction and includes data on the number of bills
associated with each basic type of subsidy mechanism able to advance through the
different decision points of the legislative agenda, from committee to enactment. Voucher
proposals seem to increase their significance in the private school choice battle by also
having the largest presence on the legislative agenda. They make up a larger portion of
28
all private school choice bills reported out of committee than their proportion of all
private school choice bills introduced. Both voucher and indirect tax subsidy proposals
pass through legislative decision points at higher rates of success than the total of all
private school choice bills. Direct tax subsidy proposals, in comparison, struggled to
make progress through state legislative processes, with both vouchers and donation
credits gaining momentum at their expense. Direct tax subsidies seem to have the most
difficulty relative to all other private school choice bills when it comes to getting reported
out of committee. Since legislative committees do the work of deciding what the chamber
as a whole will take seriously, direct tax subsidies appear to have some difficulty getting
legislative agenda attention compared to other private school choice subsidy mechanisms.

Table 2.3: Legislative Progress by Subsidy Mechanism

Type of
Policy
Option

#bills
Intro-
duced
# bills
reported
% bills
reported
# bills
pass
one/
both
chamber
% bills
pass
one/
both
chamber
# bills
signed
into law
% bills
signed
into law

Voucher

281

63

52.6%

35

55.6%

11

57.9%
% passing

21.3% 12.4% 3.9%
Expense 160 22 19.3% 9 14.3% 3 15.8%
% passing

13.7% 5.6% 1.8%
Donation 127 32 28.1% 19 30.2% 5 26.3%
% passing

25.3% 15.0% 3.9%
Total 568 117 100% 63 100% 19 100%
%passing

20.6% 11.0% 3.3%

Despite their more controversial nature, it seems that policy options offering
direct public subsidization statistically dominate the private school choice advocacy
29
activity of state legislatures, and vouchers in particular, are the most pursued type of
private school choice subsidy mechanism. The dominant role of vouchers in private
school choice reform is evident both in terms of the number of reform legislation
attempts or bills introduced, and in terms of the number of policy options given serious
attention by the state legislative process overall. On the other end of the scale, the indirect
subsidization approach of tax credits for donations appears to be the least attempted type
of private school choice subsidy mechanism by advocates but the one most likely to be
given agenda attention by state legislatures relative to its rate of bill introduction.
Although vouchers are not the only viable option for achieving progress in the effort to
create free-market policy change in education, they are the one most pursued by private
school choice advocates and legislative sponsors.
Each of these distinct subsidy mechanisms can be proposed with different
combinations of qualifying and/or restricting provisions that can further alter the final
school choice reform product. Thus, the relative policy and political implications between
the different private school subsidy mechanisms can vary in ways not fully captured by
distinctions between private school choice subsidy mechanisms. These types of
provisions are explained and outlined below.

Private School Choice Qualifications and Restrictions

In their most ambitious form, each type of private school choice subsidy
mechanism could hypothetically have an unpredictable and potentially radical impact on
30
a state’s budget and existing education system. A public scholarship or direct tax subsidy
program that covers a 100 percent of tuition costs for every student in the state could
hypothetically cause substantial disruptions in public school enrollments and funding if a
large portion of families in the school district choose to remove their children from their
local public school and instead send them to high-tuition private schools in a given year.
An indirect tax subsidy program that covers 100 percent of an individual or corporate
donation has the potential to substantially impact both state tax revenue and public school
funding if too much of the state’s tax revenue gets diverted into donations too quickly
and if too many students in a school district take advantage of the increase in private
school scholarships generated by the new program.
Without wading into debates over the potential negative or positive policy
consequences of school choice, it can be readily assumed that opponents predict that
private school choice policy changes would engender severe and negative disruptions of
the existing public school system and education spending, while proponents predict that
any resulting disruptions would be significantly less drastic and problematic. For the
purposes of this project’s analysis, the merits of these predictions are unimportant in
comparison to the uncertainty they can communicate to elected policymakers. In response
to this uncertainty, lawmakers frequently adopt a number of design characteristics when
introducing or considering a particular private school choice policy proposal, with the
ostensible intention of mitigating adverse impacts.
The various and often idiosyncratic nature of the features intended to qualify or
mitigate policy uncertainty can be overwhelming. Taking a step back and considering
31
their general impact, they essentially end up being various ways of qualifying the value
of the subsidy offered in relation to the states funding of public schools, reducing the
scope of the subsidy by targeting eligibility to a particular student subgroup, or placing
restrictions on the subsidy’s budget funding or student participation levels. The inclusion
or exclusion of different types of mitigating provisions creates significant distinctions
between individual voucher, education credit, and donation credit policy options, which
allow two different proposals of the same basic type to have significantly different policy
implications for the state’s education system. Beginning with those provisions that
qualify the value of the subsidy offered, the range of mitigating design characteristics are
further explained and their presence in the collected private school choice bill population
analyzed.

Value Qualifications
Most voucher and indirect tax subsidy proposals determine a maximum annual
dollar amount per student the state will contribute toward the cost of attending private.
6

They generally do this by capping the value of the voucher or tax credit for each student
at either a fixed dollar amount or a percentage of the state’s average Per Pupil Allotment
(PPA).
7
Those proposals that use a fixed dollar amount to cap the state’s obligation per

6
Note: Donation credits do not involve the state directly subsidizing education costs and so they most often
do not include percent of PPA provisions, though some bills do limit the value of scholarships awarded by
the receiving organization to a percent of PPA if the organization received tax subsidized donations.
Voucher and expense credit bills without a PPA provision were typically killed very early in the process
before a PPA limit was decided upon or, in a few cases, they were worth the full cost of tuition.

7
Per Pupil Allotment (PPA) is the average amount of money per year a given school or school district
within the state receives from the state’s general fund for each student enrolled. This funding is often
32
student tend to be established with the state’s PPA in mind, and rarely are they set higher
than the average per pupil funding available to public schools. In this manner, every
voucher and education credit proposal can be characterized in part by the percentage of
PPA at which it caps the value of the voucher or education credit being offered. Different
proposals set this limitation at different percentages. Analyzed caps range from 3 percent
PPA to as much as 110 percent PPA. This naturally creates very significant disparities
between proposals in terms of their utility and desirability to parents seeking to transfer
their children from public to private schools. Proposals capped at higher percentages PPA
cover more of the cost of choosing a private school and are naturally more sought after
than those capped at lower percentages, which in turn increases the probable impact of
the intended private school choice policy on current public school enrollment, and thus
public school funding, and state education budgets.
According to Table 2.4 private school choice bills commonly impose a wide
range of maximum values on the subsidies they propose, with a full quarter of the bills
collected containing no value determining provision. Of those bills that included value
qualifications, the low, middle, and high value categories saw the most use, with bills that
restricted subsidy values to either PPA percentages or to specific dollar amounts (less
than 20 percent of state PPA) having the largest “across state” presence, showing up in at
least one bill in 37 state legislatures.

weighted to favor certain schools and school districts over others; often schools in urban areas or with
higher percentages of low-income or English-as-a-second-language students are weighted to receive more
state resources.
33

Table 2.4: Distribution of Bills across Subsidy Value Qualifications

Subsidy Value in
Relation to Per Pupil
Funding
# of Bills
including
provision
% of bills
including
provision
# of states with at
least one bill that
includes
provision
% of states with
at least one bill
that includes
provision

1-20% PPA

103 18.1% 37 75.5%
21-40% PPA

57 10.0% 25 51.0%
41-60% PPA

104 18.3% 31 63.3%
61-80% PPA

59 10.4% 19 38.8%
Over 80% PPA

102 18.0% 29 59.2%
no PPA provision

143 25.2% 36 73.5%
Total 568 100.0% 49

100.0%

Rhode Island House Bill 5717, introduced in 2001, represents a bill on the
extreme low end of the value qualification scale. It proposed a tax credit for educational
expenses, excluding tuition, worth 20 percent of the first $500 spent per child up to $500
per family, which amounted to an individual student subsidy of less than 1 percent of
state average PPA and a maximum family subsidy of roughly 3 percent of state average
PPA. This proposal is also an example of a value qualification based on a fixed dollar
amount rather than explicitly attached to a percentage of the state’s average per pupil
funding. Indeed, every value qualification that either limited subsidy values to below 10
percent or set them at over a 100 percent of current state PPA were fixed dollar amounts
of this type instead of percentages, though fixed dollar qualifications span the range of
PPA percentages. For example, Arizona House Bill 2362, introduced in 1997, proposed a
34
voucher pilot program worth up to $5,000 for up to 1,000 students eligible for the federal
free-lunch program, which amounted to almost a 110 percent state average PPA for that
year. A bill representing a more middling fixed dollar amount demonstrating the way in
which value qualifications can vary within a particular proposal, is Indiana Senate Bill
470. Introduced in 1998, the bill proposed a tax credit for tuition costs worth up to
$2,000, which amounted to roughly 32 percent of the average state PPA for 1998. The
value of the subsidy went down to $1,000, or 16 percent of PPA, if annual family income
was higher than $35,000.
A good example of a high value qualification that is not a fixed dollar amount is
found in Tennessee House Bill 2675 and its companion Senate Bill 2662, introduced in
2005, which together proposed a publicly funded tuition scholarship worth 100 percent of
the average state per pupil funding. This proposal also represents one of the relatively
few bills that included a provision requiring all public and private schools accepting the
scholarship to accept it as full tuition. These provisions were only found in bills that
capped subsidy values at 100 percent of PPA, otherwise public schools receiving them
would be required to take a funding reduction for the voucher recipient. Representing the
low end of value qualifications based on a PPA percentage is the 1999 New Hampshire
House Bill 701, which proposed a voucher subsidy worth 10 percent of PPA for the
student’s assigned district.
Finally, 2001 enacted Mississippi House Bill 1643 serves as an example of the
way indirect tax credit proposals incorporate value qualifications. It proposed a tax credit
for donations to private scholarship granting organizations worth 50 percent of the
35
donated amount, but only donations that went to organizations granting scholarships
worth no more than 90 percent of state average PPA would be eligible for the credit.
Overall, value qualification variation in private school choice bills ranges widely
with low, middle, and high value proposals being almost equally represented among the
introduced bills collected. Roughly 46 percent of all the collected private school choice
bills restrict subsidy values to 60 percent of state per pupil funding or lower, indicating a
substantial though not overwhelming trend to mitigate the threat of uncertain policy
impacts by maintaining a significant gap between public funding and private subsidies.

Population Target Eligibility Restrictions
Private school choice proposals can be significantly distinct in terms of the type
and scope of the student population they are intended to benefit. Some proposals are
specifically designed to favor, or be restricted to, a particular segment of the student
population that is arguably at a disadvantage in the state’s public school system. These
types of proposals are designed with a more narrow scope of application than proposals
designed so that almost any student or family can choose to take advantage of them (i.e.,
state-wide programs). Proposals targeted to particular categories of public school students
can be quite distinct in terms of which segment of the student population the proposal is
designed to apply. The five most common categories of students associated with targeted
provisions in recent private school choice proposals are disability students, low-income
students, students in urban schools, students in poorly performing schools, and students
36
in overcrowded schools.
8
A given proposal may include one or more of these provisions,
or none of them, significantly altering the potential scope of participation in the intended
program.
The data displayed in Table 2.5 provides the distribution of each target population
provision across private school choice bills. With 268 collected private school choice
bills including one or more target student populations, roughly 47 percent of all the bills
introduced over 11 years were specifically targeted to a disadvantaged student subgroup.
These bills collectively had a total of 346 targeted provisions, with 72 bills doubling or
tripling up on such provisions. Of the 346 targeted population provisions found in those
268 bills, provisions that specifically target or restrict public subsidies to qualified low-
income students and families feature in more than a quarter of all 568 bills collected,
making them by far the most popular method for advocates and legislatures to direct a
private school subsidy toward a in-need student subgroup to frame the bill and mitigate
the policy impact of a private school choice bill. Provisions that target overcrowded
schools were only found in 14 proposals, which amounts to 2.5 percent of the total bill
population and represents the least utilized target population provision, followed by
provisions targeted to students with disabilities, which were found in only 43 (7.6
percent) of bills. Each common type of targeted provision is discussed more specifically
and includes sample legislation below.

8
Several of the proposals observed were designed for, or restricted to, more idiosyncratic student
populations such as students with a militarily deployed parent or students in foster care. Since these targets
were always far more restrictive than the more common ones specifically listed, and since no provisions
were included in more than one or two different proposals at most, they are grouped in the “Other Cost or
Participation Limitations” category below instead of this one.
37
Table 2.5: Distribution of Bills across Population Targets

Type of Eligibility
Restricting Provision
# of Bills
including
provision
% of bills
including
provision
# of states with at
least one bill that
includes
provision
% of states with
at least one bill
that includes
provision
Population Targets

0 student subgroups
targeted

301 53.0% 48 98.0%
1 student subgroups
targeted

196 34.5% 38 77.6%
2 or 3 student
subgroups targeted

72 12.7% 26 53.1%
Total 568 100.0% 49

100.0%
Student SubGroup
Targeted

disability

43 7.6% 24 49.0%
low income

174 30.6% 38 77.6%
urban school

55 9.7% 19 38.8%
low performing
school

60 10.6% 24 49.0%
overcrowded school

14 2.5% 7 14.3%
Total 346 100.0% 49

100.0%

Disability Student Provisions are targeted to students with a registered disability
or classified by the state as special needs. Many of these types of proposals define
disability eligibility to include the more commonly diagnosed learning disabilities such as
Attention Deficit Disorder. The most famous example is Florida’s McKay voucher
scholarship program, which allows any student with an individualized education plan
who has been in public school for at least a year to receive a publicly funded scholarship
worth up to 100 percent of state PPA for the student’s assigned district. Another example
38
of this type of more liberal disability classification is Vermont House Bill 57, introduced
in 2003, which proposed a publicly funded scholarship worth up to 100 percent of
average state PPA for students who had not met two or more goals for their
individualized education plan. Some proposals are more restrictive, such as Indiana
House Bill 1005, introduced in 2005, which proposed a publicly funded tuition voucher
worth 100 percent of average state PPA, but only autistic children would have been
eligible; or Mississippi Senate Bill 2332, introduced in 2007,which proposed a publicly
funded tuition scholarship in which only students recognized as possessing special needs
that have a multiplier effect on state per pupil funding are eligible, the value of the
subsidy would have been 100 percent of state PPA, with disability multiplier effect
incorporated.
Low-Income Student Provisions are targeted to students determined to be part of
low-income families, with different proposals outlining different ways for the low-
income determination to be made. The majority of proposals targeted to low-income
students made the low-income determination by linking scholarship eligibility to either a
student family’s income eligibility for federal lunch programs or to federal poverty
guidelines. These proposals either set the eligibility directly at the federal limits, as
represented by Washington State House Bill 1670 and Senate Bill 5949, both introduced
in 1999, which proposed a publicly funded tuition scholarship targeted only to students
whose families had incomes no higher than 100 percent of the maximum allowed to
qualify for the reduced-price lunch program or set eligibility at some higher percentage
based on the federal income limits, as represented by New Jersey Assembly Bill 257,
39
introduced in 2006, which proposed a tax credit for donations to private scholarship
granting organizations that was only allowable if the receiving organization provided 95
percent of its scholarships to families earning no more than 250 percent of the federal
poverty level. None of the collected proposals established an eligibility determination that
was more restrictive than 100 percent of the federal limits. Some proposals put in a
specific maximum income dollar amount instead, such as Oklahoma House Bill 1881.
Introduced in 2003, it proposed a tax credit for tuition costs worth up to $1,000 per child
for families whose taxable income is no higher than $50,000 annually.
Although provisions that limit scholarship eligibility by family income usually do
so with the intent to frame and target the reform to low-income families that do not have
the resources to choose private schools or move to alternative public school districts on
their own, a small number of bills set income eligibility limits at levels that stretch the
limits of what can realistically be considered low-income. For example, West Virginia
Senate Bill 582, introduced in 2003, proposed a tax credit for tuition expenses worth 100
percent of tuition cost up to $500 for families earning no more than $100,000. This
suggests that some income eligibility provisions are more about preventing upper-income
citizens from using the subsidy than it is about focusing on the needs of low-income
citizens in particular.
Urban School Provisions are targeted to students assigned to schools that are part
of larger urban school districts. These are usually designed specifically to apply to a
particular type or set of school districts, or to apply to school districts located in a
particular type or set of cities or counties. Quite often, these types of eligibility provisions
40
are introduced and designed with a particular city or county in mind, or even a particular
set of cities or counties. For example, Florida House Bill 1177, introduced in 1997,
proposed a publicly funded tuition scholarship worth up to 50 percent of state PPA, but
only for students attending schools located in Clay, Dade, or Orange County. The cities
or counties specified are commonly more populated or urban and exhibit demonstrated
history of cost, management, and performance problems. Almost as often, provisions
target subsidy eligibility to school districts, cities, or counties that meet a specified set of
growth, population, or size criteria. These criteria also tend to target large and urbanized
areas, such as Texas House Bill 2366, introduced in 1999, which proposed a publicly
funded tuition scholarship worth up to a 100 percent of PPA for students attending
schools in the six school districts in the state with the largest school membership, or
Oregon House Bill 3010,introduced 2007, which proposed a publicly funded tuition
scholarship worth 100 percent of PPA for students attending schools in a district with an
average daily attendance of 40,000 or more. Even when a particular school district or city
isn’t specifically targeted, the criteria established for which urban school students qualify
are sometimes set up in such a way as to only apply to a particular city or county without
having to name it directly. Missouri Senate Bill 497, introduced in 2003, is representative
of this type of provision, as it proposed a tax credit for donations to private scholarship
organizations as long as the receiving organizations used the donated funds to offer
scholarships to students in districts with a population of at least 400,000 which would
only apply to Kansas City and St. Louis.
41
Underperforming School Provisions are targeted to students assigned to schools
determined to be underperforming or failing. This determination is most often based on a
school or district being designated as underperforming by some education authority,
usually either the Department of Education, the Education Commissioner, a board of
education, an independent state board, or even by the local school district. A good
example of this type of provision is Colorado House Bill 1180, introduced in 2001, which
proposed a tax credit worth up to $3,000 toward tuition for students attending schools
designated with “low” or “unsatisfactory” ratings by the state Board of Education. Other
proposals containing this type of determination mechanism are similar, just switching the
terms of designation and the designating authority. Somewhat less often, the
determination of underperforming is based on a school’s comparative rank, either within
the state or nationally, or based on a school or district’s standardized test scores. Georgia
Senate Bill 68, introduced in 1999, is representative of a comparative rank determination.
It proposed a publicly funded tuition scholarship worth up to 90 percent of tuition for
students attending schools performing in the lowest 40 percent of all schools nationally.
Typifying a determination based on standardized test scores, Minnesota House File 2586,
introduced in 2002, proposed a publicly funded tuition scholarship worth up to 100
percent of the states average PPA for students attending public schools that score below
50 percent on state assessment exams for three consecutive years.
Overcrowded School Provisions are targeted to students assigned to schools
and/or school districts determined to be overcrowded. This determination is commonly
based on state imposed guidelines as found in Florida House Bill 303, introduced in
42
2001, which proposed a publicly funded tuition scholarship worth up to $3,000 with
eligibility limited to those students in schools at 120 percent or greater of determined
optimal capacity or determinations made by an educational authority, such as with Maine
House Bill 915, which was introduced in 2003 and proposed a voucher program for
students in districts the school commissioner declared to be “high growth”. As a unique
aspect of this type of targeted provision, the proposal allows the local school district or
school board to decide whether or not a private school choice subsidy is needed to reduce
overcrowding. For example, Utah Senate Bill 61, introduced in 1997, would have
allowed local school boards to issue vouchers to students in the district attending schools
targeted for “growth restrictions” if they so chose thereby reducing class sizes.

Cost and Participation Caps
All cost and participation limiting provisions, despite their often unique design
particulars, reduce the scope and application of the proposed policy change through two
general and related mechanisms: firstly by limiting the amount of state funds available to
the proposed choice program, and/or secondly by limiting the amount of students who
can participate in it. Thus, regardless of the particular limitation adopted or included, all
of these different provisions have a similar intent and impact on the particular policy
change under consideration.
9
The more important question is one of elasticity – i.e.,

9
If the funding of a proposed program is restricted than the number of subsidies that can be generated will
decrease along with the number of families or students who can participate in the program. If the restriction
is placed on participation instead than the number of subsidies that need to be funded will decrease along
with the program’s needed funding from the budget. In this way both funding and participation restrictions
have mitigating effects on budget costs and policy impacts.
43
whether the provision creates hard and specific limits or soft and general limits on the
program’s scope of cost and participation.
Hard limitations create specific, restrictive guidelines on how much money a
program will receive or how many vouchers or tax credits the program can offer in a
given year of operation. Soft limitations create more general, flexible, and less restrictive
guidelines on how much money a program will receive or how many vouchers or tax
credits the program can offer. While a hard limit provision may dictate that eligibility for
a given program is capped at 10,000 students statewide, a soft limit provision may dictate
that only students in kindergarten through fifth grade are eligible, allowing the number of
students eligible for participation to fluctuate with elementary school enrollment over
time. Proposals with hard limitations are arguably intended to create programs that are
smaller in size and scope than proposals with soft limitations, regardless of the population
targeted or the value of the voucher/credit.
As demonstrated in Table 2.6, provisions that specifically restrict the funds
available to a proposed private school choice program or the number of students allowed
to participate in it are more common than provisions that impose more flexible
restrictions. However, neither type of provision shows up very often in the collected
private school choice bills, with only about a quarter of all bills containing either type of
restriction.

44
Table 2.6: Distribution of Bills across Cost/Participation Caps

Type of Eligibility
Restricting Provision
# of Bills
including
provision
% of bills
including
provision
# of states with at
least one bill that
includes
provision
% of states with
at least one bill
that includes
provision
Hard Restrictions

Yes

83 14.6% 32 65.3%
No

485 85.4% 49 100.0%
Total 568 100.0% 49

100.0%
Soft Restrictions

Yes

56 9.9% 16 32.3%
No

512 90.1% 49 100.0%
Total 568 100.0% 49

100.0%

An example of a provision that places a rather low hard cap on the budget
appropriations for the proposed private school choice program is found in Illinois House
Bill 991 and Senate Bill 565, both introduced in 1997. These bills would have created a
pilot voucher program for Chicago schools that was worth up to $2,500. The funding cap
placed on the program was only $5 million, which given the size of the Chicago school
system is a relatively restrictive amount of money from which to both administer the
program and fund voucher subsidies. Proposals not restricted to a particular school
system tended to include higher caps. Arizona Senate Bill 1263, introduced in 2003,
represents a hard cap that started out relatively small but included automatic yearly
expansions. It would have a tax credit for corporate contributions to private scholarship
granting organizations worth 100 percent of the donated amount. The cap limited tax
credits to only $10 million the first year but included annual increases until 2007, at
45
which point it would be capped at $50 million in tax credits per year. The restrictive
nature of hard caps on funding levels are more or less restrictive depending on the budget
and public school funding characteristics of the state they are proposed in, making it
somewhat more difficult to identify the range from more to less restrictive.
Looking at hard caps on participation, Connecticut House Bill 5037, introduced in
2004, proposed one of the most restrictive hard caps found in the collected bills. It
proposed a publicly funded tuition scholarship worth up to $4,000 but limited to 500
students in “priority” school districts chosen by annual lottery. The largest participation
allowance found in a hard capped bill was in California Assembly Bill 25, which was
introduced in 1997 and proposed a voucher subsidy worth up to 90 percent of state
average PPA for up to 250,000 low-income students chosen by lottery.
There are two things to note about these types of caps. The first is that, in similar
fashion to hard caps on funding, the restrictive quality of hard caps on participation is
relative to the public school populations of the state or district to which the subsidy is
attached. The second is that hard caps on participation levels tend to include a lottery
mechanism to decide who receives a subsidy and who does not when they are included
with direct subsidy mechanisms.
Soft cap proposals can restrict funding or participation levels in far more creative
or idiosyncratic ways. A number of bills contain soft cap provisions that are repeated in
any form in other bills or states. A good example of a unique soft cap is found in
Washington Senate Bill 6283, which was introduced in 1998 which would have created a
tax credit worth about 10 percent of state PPA but only allowed tax payers to claim the
46
credit if they had submitted a notarized statement with their county treasurer claiming a
“conscientious objection” to their child attending public school. Another unique example,
and one of the only examples of a soft cap on funding levels rather than participation
levels, is Vermont Senate Bill 265, introduced in 2007, which would have required
districts that approve budgets exceeding projected costs to use the excess funds to
provide voucher subsidies worth up to 80 percent of the district PPA to students chosen
by lottery.
The two most common types of soft cap provisions are those that restrict
participation to a particular set of grade levels and those that specify a percentage instead
of a fixed number of students allowed to participate. Provisions that restrict participation
for students in a particular grade can range from specifying a wide range of grade levels
to only one grade level. New York Senate Bill 1939, introduced in 2005, is very narrow
in its grade specification, only allowing the parents of students in kindergarten to
participate in the proposed tax credit program that would have reimbursed up to $3,000
of incurred private education expenses. Those bills that specified a range of grade levels
usually directed benefits to only those grades associated with high school or to all grades
but those associated with high school. For example, Missouri Senate Bill 601, which was
originally introduced in 1998, would have allowed only the parents of students in grades
9 through 12 to apply for its proposed direct tax subsidy, worth up to $2,500. Those types
of provisions that specified a particular percentage of students allowed to participate
almost always linked the percentage cap to district student populations instead of state
student populations, with assigned percentage caps ranging from 1 percent to 15 percent
47
of district enrollment. Two examples of a moderately high enrollment cap, then, are
Nevada Assembly Bill 507 and Senate Bill 385, both introduced in 1999. They proposed
a publicly funded tuition scholarship worth up to 100 percent of state average PPA with a
maximum of 10 percent of any district’s school-aged population allowed to participate,
chosen by lottery if necessary.

Putting the Pieces Together

The key policy dimensions outlined above can be combined in various ways to
create an array of different policy proposals. Since each policy dimension includes
options that are more or less ambitious in their policy impact, the particular combination
contained within a bill can suggest a wide range of potential policy impacts, from very
limited reforms that, if enacted, would only facilitate private school choice for a very
small fraction of school age students, to very ambitious reforms that, if enacted, would
potentially facilitate private school choice for almost any family that wished to take
advantage of the new program. To get a handle on this variation, Table 2.7 displays the
distribution of the collected private school choice bills according to the major possible
policy combinations.
10

10
Even with the generalized method different content dimensions are categorized above, the number of
possible combinations reaches into the several hundreds. Therefore they are further simplified and
generalized for the summary table to create a more manageable yet still meaningful set of 24 possible
combinations.
48
Table 2.7: Major Policy Dimension Combinations

No Target Populations

1,2, or 3 Target Populations

No Soft or
Hard Caps

Either Soft or
Hard Capped

No Soft or
Hard Caps

Either Soft or
Hard capped

Voucher

50%+ PPA

77

23

58

34

1 – 49% PPA

23

7

30

29

Direct Tax
Subsidy

50%+ PPA

31

1

11

1

1 – 49% PPA

79

8

24

4

Indirect Tax
Subsidy

50%+ PPA

67

15

20

17

1 – 49% PPA

4

2

2

1
Note: 50%+ PPA category includes bills that contained no provisions qualifying subsidy values

The distribution of bills across the simplified policy combinations reveals two
things beyond the wide variety of policy forms private school choice can adopt. The first,
and most obvious, is that bills do not distribute evenly across the different policy
combinations. The second is that subsidy mechanisms seem to have an inclination toward
certain types of subsidy values, and different subsidy mechanisms also seem to have
different predilections for including target populations or for imposing funding and
participation caps. For example, the displayed data reveal that voucher proposals are far
more likely to have attached provisions limiting their size and scope – i.e. target
populations and funding and participation caps – than either direct or indirect tax
49
subsidies. With a hundred bills and 64.4 percent of voucher proposals containing a
limiting provision, they were the only subsidy mechanism to have more limited eligibility
bills than high eligibility bills considered by legislatures. By contrast, both types of tax
subsidy proposals were more likely to avoid the inclusion of size and scope limitations,
with only 39.8 percent of direct tax subsidies (49 bills) and 44.5 percent of indirect tax
subsidies (57 bills) attached to either target population or funding and participation cap
provisions.
Looking at how subsidy value distributes across the three basic types of private
school choice subsidy mechanisms, it becomes clear that benefit values load differently
on different options. To explore this variation more accurately, Table 2.8 displays the
distribution of value qualifications across subsidy mechanisms. Voucher bills tend to
allow for much higher benefit values than either direct or indirect tax subsidy bills.
Roughly three-fourths of voucher bills contain value determinations close to half of
current PPA spending or higher, and value maximums equal to state PPA spending load
heaviest on these types of bills. Direct tax subsidy bills most often include very low
subsidy value limitations, with more than half of all such bills containing provisions
limiting assistance to below 20 percent of PPA spending. This is most likely due to the
fact that many of the introduced expense credit bills deliberately and explicitly excluded
tuition costs from reimbursement eligibility. Since tuition is the lion’s share of private
school expenses, this naturally lowers the needed value of the tax credits offered. Tax
credits for donations to private organizations do not involve the state directly subsidizing
education costs, and because of this, they generally do not include percent of PPA
50
provisions, as is made clear by the fact that 83 percent of them contain no provision
specifying a value maximum. Indeed, it is indirect tax subsidy proposals that inflate the
overall number of bills in which no percent of PPA spending provision was identified.
The roughly 17 percent of indirect tax subsidy proposals that did contain value
determinations either limited the value of any scholarship granted using eligible
donations to the specified maximum PPA value or required any private scholarship
organization wishing to receive tax credit eligible donations to not grant scholarships of a
value higher than the specified maximum PPA value.

Table 2.8: Distribution of Subsidy Values across Subsidy Mechanism

Subsidy Value in relation
to Per Pupil Funding

Voucher

Education Credit

Donation Credit

Total
1-20%
Of PPA

13
4.2%

87
54.7%

3
2.4%

103

21-40%
Of PPA

37
13.0%

22
12.6%

0
0.0%

59

41-60%
Of PPA

79
27.8%

14
8.8%

11
8.8%

104

61-80%
Of PPA

50
17.6%

5
3.1%

2
1.6%

57

Over 80%
Of PPA

85
29.9%

12
7.5%

5
4.0%

102

No PPA
Provision

20
7.0%

19
11.9%

104
83.2%

143

Total 284
100%
159
100%
125
100%
568
100%

51
Based on the distribution of policy combinations across subsidy mechanisms,
three patterns of bill content are more common than others over the 568 introduced bills.
The most common policy combination proposed a voucher subsidy that was high in value
but was directly qualified or limited in size and scope through either target populations or
funding and participation caps. This type of policy proposal comprised 40.9 percent of all
introduced voucher subsidy bills and 20.2 percent of all introduced private school choice
bills. Florida’s original A+ Opportunity Scholarship Program, signed into law by
Governor Jeb Bush in 1999, is representative of a successful bill enactment based on this
type of private school choice bill. The program is a voucher subsidy that targets
eligibility to a disadvantaged portion of the state’s student population and offers
scholarships worth the full amount of tuition, up to 100 percent of state PPA. The
Opportunity Scholarship Program targets program eligibility to students attending
schools that have received a “Failing” grade on the Florida Comprehensive Assessment
Test (FCAT) two years in a row. In 2005, 740 opportunity scholarships were awarded to
Florida students. Ohio’s Educational Choice Scholarship Program was signed into law by
Governor Bob Taft in 2005 and represents another successfully enacted bill based off a
different version of this private school choice policy formula. Ohio’s voucher subsidy is
not targeted to any particular type of student or set of schools but is limited to a
participation level of roughly 14,000 slots chosen by lottery.
The second most common combination proposed a direct tax subsidy that
combined low subsidy values with a statewide program not directly limited by either
target populations or caps of any type. These policy proposals comprised 49.6 percent of
52
all introduced direct tax subsidy bills and 13.9 percent of all introduced private school
choice bills. This type of private school choice policy also gained an incarnation in
existing state law when Illinois Governor George Ryan signed the Individual Tuition Tax
Credit, into law in 1999. This direct tax subsidy allows parents to deduct 25 percent of all
incurred private school expenses, including tuition, up to $500 per family per year. The
program was unrestricted by any eligibility targets, funding caps, or participation caps,
but only amounted to a subsidy worth a small fraction of state funding for public school
students. The program generated $67 million in tax savings for 193,000 state taxpayers in
2003, while the state’s overall budget expenditures were roughly $25 billion.
Finally, indirect tax subsidy bills most often offered an indirect subsidy
mechanism that was accompanied by no restrictions or qualifications on subsidy value,
eligibility, or funding and participation caps. This unrestricted but indirect policy
constituted 52.3 percent of the indirect tax subsidy bills and 11.8 percent of all private
school choice bills introduced into state legislatures over the 11 year timeframe.
Arizona’s Individual Tax Credit for Scholarship Donations, signed by Republican
Governor Jane Dee Hull in 1999, emerges as probably the best example of this type of
policy formula in existing state law. The program created by the law is an indirect tax
subsidy in which any individual taxpayer to claim a dollar-for-dollar tax credit up to $500
for donations to non-profit scholarship granting organizations. There are no statutory
limits to the value of the scholarship offered, though most organizations offered
scholarships worth a smaller proportion of PPA in order to be able to offer more
scholarships, and there were no limits placed on the number or dollar amount of tax
53
credits available per year. In 2005 the program generated $42.1 million in tax credits
compared to a state budget of more than $10 billion.
These are the content combinations that tend to be favored in private school
choice policy proposals. However, even these proposal types only represent about half of
their respective subsidy mechanism populations and altogether only represent about 46
percent of all collected private school choice bill introductions, which again speaks to the
diversity of policies contained within private school choice. Still, these particular content
formulas present a set of private school choice programs that are blunted in their potential
impact on both existing public school enrollments and on state budgets. Those proposals
that present the most attractive type of subsidy (i.e., vouchers) tend to be the most
directly restricted in size and scope, while those proposals that are the least restricted in
size and scope either offer low subsidy values making their offered subsidy less attractive
or else offer an indirect subsidy mechanism. The question than becomes this: To what
extent do the other 54 percent of private school choice bills offer policies that are more or
less ambitious in their suggested impacts. This is the issue to which we now turn.
Figure 2.1 captures the overall distribution of private school choice policy
proposals by organizing them according to their policy impact.

54
Figure 2.1: Distribution of Proposal Policy Impact
Large Direct Subsidies
of High Value, 19%
Common
Combinations, 46%
Miscellaneous
Combinations, 16%
Small Direct Subsidies
of Low Value, 18%
Small Indirect
Subsidies of Low
Value, 1%

One of the more crucial points this graph illustrates is that only a minority of the
private school choice proposals introduced into state legislatures were designed or
developed to create the types of school choice programs that combine the right
combination of policy dimensions to conceivably challenge, and perhaps eventually
replace, a state’s existing system of government funded and administered schools. Data
suggests that only a small minority of the bills making their way through state legislatures
over those 11 years would have turned a state’s education policy into something that
would substantially satisfy the ideal of a state school system largely driven by private
service providers, individual choices, and market dynamics, at least in the foreseeable
future. Arguably, to come close to satisfying that ideal, a proposed private school choice
program should present an accessible, attractive, and unrestricted subsidy for private
school costs in order to eventually generate a critical mass of parents able to viably take
advantage of the subsidy and inject a hefty enough measure of market choice into the
55
system to impose market discipline on existing public schools. This would suggest a bill
designed to directly subsidize (i.e., a voucher or direct tax subsidy) education costs to
private providers with no target population provisions or budget/participation limits (i.e.,
have state-wide eligibility) and with subsidies of at least 50 percent of what public
providers receive. However, only 108 bills, just 19 percent of all recently considered
private school choice legislation satisfies these requirements. This is a comparatively
small share of the overall private school choice reform efforts.
On the other end of the impact scale, an even lower proportion of introduced
proposals would create private school choice programs that were so restrictive, meek, and
unthreatening in their policy impact that they would have almost no meaningful impact
on existing education policy dynamics. The most toothless private school choice reforms
would be those that indirectly subsidize private school expenses through tax credits for
donations to private-scholarship-granting organizations, require the receiving scholarship
organization to offer scholarships worth less than 50 percent of what a public schools
receive per student, and than ensure the program is narrow and small by only allowing
the funded scholarships to go to a small subset of the student population and by placing a
restrictive cap on the amount of credits allowed in a given year. Only one proposed bill
fits this description Missouri House Bill 345, highlighted at the beginning of the chapter
so we turn to the next most unchallenging type of private school choice reform: those
that offer direct subsidies of low value to a narrow portion of the state’s student
population with limited funding or strict participation caps. The total for both types of
private school choice proposal only amounts to 34 bills and 6 percent of all collected
56
private school choice legislation. Somewhat less meager in their policy impact are those
bills that provide or require low-value scholarships that are either targeted to a narrow
subset of the student population or have their funding or participation levels capped.
These types of proposals are more prevalent than the lowest impact bills but still only
account for 73 bills, or about 13 percent of the total collected bill population.
The majority of proposals introduced combined the various policy dimensions in
such a way as to create a modest but not insignificant policy reform. These types of bills
make up almost 63 percent of all collected private school choice proposals and are
divided into two categories: those proposals that combine policy dimensions in one of the
three common formulas, and those proposals that have a moderate impact but combine
the policy dimensions in various other ways. Considering that the majority of attempted
school choice legislation ends up being designed either initially or during legislative
consideration to supplement instead of compete directly with public education, there are
some indications that private school choice advocates themselves understand the political
and policy implications of their preferred policy alternative and are quite willing to
sacrifice short term policy impact for a longer term strategy of getting particular free-
market policy dynamics established into a policy subsystem.

Chapter Two Conclusion

There are lots of different ways a private school choice bill can be structured. The
choice of subsidy mechanism and subsidy value, as well as the choice of whether or not
57
to target the subsidy to a particular student population and the choice of whether or not to
place funding or participation limitations on the subsidy program, can all affect what the
final policy will like including the political challenges and policy uncertainties it brings
to the table. The range and variation contained within these choices is not just an abstract
possibility but a demonstrated reality, the private school choice bills introduced into state
legislatures over an 11 year period were so eclectic in the combination of choices they
incorporated that even when the choices were significantly generalized no particular
formula was able to capture more than 20 percent of the overall bill population and the
three most common formulas only amounted to 46 percent that population. Clearly, the
political and policy implications involved in private school choice reform are quite
complex and there is more to understanding school choice politics than distinguishing
between public and private school choice policies. Given the ability of different policy
design choices to create vastly different policy impacts, it would seem that this policy
content variation should matter for understanding why private school choice does or does
not pass. The extent to which particular types of policy design decisions, common
combinations are actually more likely to gain success in the legislative process will be
examined in Chapter Four and Five. The next chapter will continue the discussion of
individual bill introduction activity by examining the way those bills were distributed
across different state legislatures and the extent to which different states and different
legislatures can be considered to have invited more or less activity.
58
CHAPTER THREE:
MAPPING LEGISLATIVE ACTIVITY

The struggle over private school choice involves a complex range of policy
options as detailed in the previous chapter, and has played out on a wide range of
legislative venues. Following Reagan’s exit from office, school choice reformers
continued to work toward realizing his goal of enacting voucher or tax credit programs at
both the state and national level. The state level push launched in earnest in 1995 and
1996 when 14 states saw voucher legislation introduced, and when Michigan and
Missouri saw constitutional amendments proposed that would have expressly permitted
the creation of voucher programs. During the 11 years following those efforts, free-
market education reform advocacy would result in almost every state in the nation
experiencing at least one private school choice reform proposal, and in the case of one
state as many as 32 separate proposals.
This chapter is focused on adding to an understanding of contemporary school
choice politics, and further setting up an analysis of why and how private school choice
passes in state legislatures, by analyzing the uneven distribution of private school choice
legislative activity across states and state legislatures. It describes the types of states that
have experienced both relatively high and low levels of legislative activity, and describes
which type of states more likely to provide some agenda attention to private school
choice legislation once it is introduced. More specifically, it maps out the overall
distribution of collected individual private school choice bills across all 50 states,
differentiates between high, average and low-activity states, and considers how these
59
state categories can vary according to key demographic, public school system, and
political characteristics.

State and Regional Distribution of Private School Choice Activity

The average number of private school choice bills introduced into a given state
from January 1997 to December 2007 is 11.36, or roughly one bill introduced per year
over 11 years. On the low end of activity, only Wyoming experienced no private school
choice bill introductions; five states saw only one or two bills introduced; and 13 states
averaged an introduction rate of one bill each alternating year or less. On the high-
activity end, Louisiana tops the range with a total of 32 bills introduced, and 12 states
averaged more than one and a half bill introductions per year.
Looking at the legislative consideration these bills received after introduction, the
average number of bills favorably reported out of committee for all states is 2.34. Just shy
of half, 22 state legislatures completely ignored the private school choice proposals
submitted, neglecting to report a single bill out of its assigned substantive committee.
Five other states reported only a single bill out of substantive committee, which
subsequently died before being scheduled for a floor vote. Effectively ignoring private
school choice legislation comprises slight majority position among state legislatures. Of
the remaining 23 states that did pay reform bills some attention, 10 enacted at least one
proposal into law. Arizona’s legislature proved to be the most attentive and friendly
environment for private school choice, with 18 bills successfully reported out of
60
committee and 10 of those passed at least one legislative chamber, three bills were
eventually signed into law. The Florida state legislature also demonstrated a willingness
to engage free-market education policies. It passed 14 bills out of committee, nine
through at least one legislative chamber, and three bills achieved enactment.
Map 2.1, below, geographically illustrates the “by state” variation in bill
introduction activity by categorizing states according to the number of private school
proposals introduced into their legislature as compared to the overall state average 11.36
bills between 1997 and 2007, roughly equating to about one bill introduction per year.
11

State legislatures that experienced six or less bill introductions a rate of, at most, one
private school choice bill every other year are categorized as “low-activity” states. Those
states that saw 18 or more bill introductions a rate of at least one-and-a-half bills every
year are classified as “high-activity.” States that experienced a rate of bill introduction
activity that is comparatively closer to the average of one bill a year, between 7 and 17
bill introductions over 11 years, are classified as “average-activity.”

11
A specific list of the number of bills introduced into each state legislature is included in the appendix.
61
Figure 3.1: Private School Choice Bill Introduction Activity by State

Low-activity (0 – 6 bills introduced)
Medium Activity (7 – 17 bills introduced)
High-activity (18+ bills introduced)

The geographic distribution of private school choice legislation indicates
identifiable regional fluctuation in activity. Southwest and New England encountered
more bill introductions relative to other states, with the Southeastern and Midwestern
states seeing mixed levels of activity and Pacific and Northwestern states receiving
relatively low levels of activity. More specifically, New England and Mid-Atlantic states
62
averaged 13.88 bill introductions, and those states located in the Mountain and Western
South (i.e. southern states west of the Mississippi River) averaged 14.5 bill introductions
from 1997 to 2007, whereas all other regions averaged between 7 and 11 bill
introductions during the same period of time.
However, if we turn to the geographic distribution of bill attention, or the number
of proposals that state legislatures opted to favorably report out of their initial committee
assignment, we see a somewhat different regional pattern. Map 3.2 illustrates the “by
state” variation in bill attention activity by also categorizing states according to the
overall state average. States are categorized as either high, average, or low in the
legislative attention they provided introduced bills, based on the number of bills
successfully passed through their substantive committee assignment as compared to the
overall state average of 2.34. States whose legislatures did not allow any private school
choice bills to pass through substantive committee were, naturally, classified as “low
consideration” or “low attention.” States whose legislatures saw four or more introduced
bills receive and succeed past a committee vote which nearly double the overall average
were classified as “high consideration” or “high attention.”

63
Figure 3.2: Private School Choice Bill Attention Activity by State

Low Attention (0 bills reported out of committee)
Medium Attention (1 – 3 bills reported out of committee)
High Attention (4+ bills reported out of committee)

Comparing the regional distribution of bill attention across states to the
distribution of bill introduction, we see that New England and Northeastern states go
from being highly active in terms of introductions to minimally active in terms of
committee passage. Indeed, the Northeastern states actually report the lowest average
number of bills passing out of committee, with an average of less than one bill (0.78 bills
64
reported) per state over 11 years. The Mountain and West South states, on the other hand,
follow up their robust showing in bill introduction activity with the highest average rate
of successful committee reporting, as with 4.81 bills per state over 11 years.
Assuming that the uneven distribution of legislative activity in terms of the
number of bills both introduced and provided legislative attention, is at least partly
related to the fact that different states present different social, economic, political, and
public school environments. Examining the relationship that key state characteristics and
high, average, and low legislative activity provides some insight into the geographical
distribution patterns displayed and this, in turn, provides initial indications as to what
types of state legislative venues may be inclined more favorably to consider private
school choice reforms.
In order to more effectively and efficiently compare the state characteristics of
high, average, and low legislative activity states, data draws from existing scholarship on
school choice politics. This literature has highlighted the relevance of public
constituencies particularly religious conservatives (Moe 2001), low-income and urban
communities (Moe 2001; Fusarelli 2003), and ethnic minorities (Gokcekus, Phillips, and
Tower 2004) indicating the importance of a state’s racial, class, and religious
demographics. This scholarship has also been consistent in its identification of
Republican political leaders and conservative political constituencies (Brunner, Sostelie,
and Thayer 2001; Brunner and Sostelie 2003; Gokcekus, Phillips, and Tower 2004;
Kenny 2005) as the natural political home of private school choice reform, indicating the
importance of partisan control and political ideology. Finally, a state’s existing public
65
school dynamic is naturally relevant to the issue of private school choice. Authors who
pay attention to private school choice often highlight that proposed reforms and their
advocates tend to argue and frame the issue as a remedy to perceived deficiencies in
existing public school systems. The importance of public school performance for school
choice politics has been directly implicated by those studies that have observed higher
levels of opposition to school choice reforms among suburban voters and residents in
neighborhoods with higher performing public schools (Kenny 2005). Accordingly, it will
be along these dimensions that we will compare states, beginning with their
demographics and than turning to their public school system and their political
environment.

Legislative Activity and the Distribution of State Characteristics

State Population Demographics
The relationship between state population characteristics and state legislative
activity categories is the first stop on our road to describing the types of states more or
less likely to experience more legislative activity. Table 3.1 and Table 3.2 use the
following seven demographic measures, each of which represents a particular
demographic snapshot taken from a year close to the middle of the 1997 – 2007
timeframe:

66
1) State total population in millions of people
2) Percent of state population that self-identifies as white non-Hispanic
3) Percent of state population that self-identifies as black
4) Percent of state population with household incomes below the federal poverty
guideline
5) Percent of state population that self-identifies as evangelical Christian
6) Percent of state population that self-identifies as Catholic Christian
7) Percent of state population that attends church services at least once a week
67
Table 3.1: State Demographics across Bill Introduction Categories

All States (50) High-activity
States (12)
Average-activity
States (24)
Low-activity
States (14)

Population
(in millions)

Mean 5.6 m 7.5 m 6.9 m 3.4 m
Min – Max 500 k – 33.9 m 600 k – 20.9 m 780k – 33.9 m 500 k – 12.3 m
Std. Deviation 6.21m 7.03m 6.9m 3.33m

% White Pop

Mean 73.1% 66.1% 72.1% 80.6%
Min – Max 24.9% – 95.3% 41.7% – 95.2% 24.9% – 95.3% 65.7% – 89.6%
Std. Deviation 15.2 15.5% 16.7% 8.2%

% Black Pop

Mean 9.9% 12.5% 10.1% 7.3%
Min – Max .3% – 37.2% .5% – 37.2% .6% – 28.9% .3% – 26.0%
Std. Deviation 9.6% 11.6% 9.1% 8.6%

% Poverty Level

Mean 12.4% 13.9% 11.3% 12.8%
Min – Max 7.7% – 20.3% 8.1% – 20.3% 7.7% – 18.5% 9.7% – 17.4%
Std. Deviation 3.2% 4.17% 2.8% 2.5%

% Evangelical

Mean 26.4% 25.9% 23.8% 31.4%
Min – Max 7% – 53% 10% – 47% 7% – 53% 21% – 51%
Std. Deviation 12.4% 11.4% 13.2% 11.0%

% Catholic

Mean 22.2% 25.3% 23.0% 18.1%
Min – Max 5% – 43% 9% – 43% 5% – 43% 6% – 31%
Std. Deviation 10.1% 9.2% 11.1% 8.0%

% Weekly Church
Attendees

Mean 38.8% 38.4% 38.0% 40.6%
Min – Max 22% – 60% 23% – 60% 23% – 57% 22% – 52%
Std. Deviation 9.6% 11.1% 9.4% 8.9%

Based on the demographic information presented, state activity categories differ
primarily along the demographic characteristics of population size, racial diversity, and,
to a lesser degree, Catholic citizenry. The proportion of low-income, evangelical, and
68
weekly church-attending citizens in states did not display mean results in a consistent
pattern across the bill activity categories that create confidence in their descriptive value.
Population demographics were distributed across bill activity categories so that
mean population size increased with higher bill introduction categories, though the
largest disparity between the means is between average and low-activity states, whose
means differ by 3.5 million, with the means of high and average population states only
differing by 0.6 million. It seems more accurate to say that low-activity states tended to
have smaller populations than to say that high-activity states tended to have larger
populations. However, in considering these numbers, it is important to acknowledge that
some states disproportionately high in population undermine the descriptive accuracy of
the mean population scores for the high and average-activity states. This is especially the
case with the over 33 million residents of California, a distinctive 13 million more
populous than the next two most populous states, Texas (20.9m) and New York (19.5m).
California is an average-activity state with nine bill introductions; the next highest
population state in that category is Illinois, with 12.4 million people. This is a large
disparity, and one that does not exist to nearly the same extent in the high and low state
activity categories. To see how deep an impact California’s extreme population has on
the average-activity state mean, we see that without California’s population the mean
drops from 6.9 to 4.8 and the standard deviation drops from 6.9 to 3.7. This suggests that
the mean population for average-activity states and perhaps, to a lesser extent, for high-
activity states is probably somewhat overstated. If we look at population from a different
angle – the number of states in each category that had populations of less than 2 million –
69
low-activity states are still clearly described as possessing comparatively lower
population sizes than other states. This is because 50 percent of low-activity states had
populations below two million whereas only 29 percent of average-activity states and 16
percent of high-activity states had such a low population size.
The percentage of a state’s population that self-identify as white is distributed
across state activity categories a bit more evenly, with high-activity states having the
lowest mean proportion of white citizens, average-activity states at about the overall state
mean for white population, and low-activity states possessing the highest mean
percentage of white citizens. The association of smaller white majorities and larger
minority populations with high bill activity states is further supported by the fact that
people identifying as black were conversely distributed across state activity categories.
The only outlier state in terms of white population demographics is Hawaii, with a
disproportionately low white population of 24.9 percent; the next lowest white population
state is New York, with a white population of 41.3 percent. Excluding Hawaii increases
the mean for average-activity states to 74.9 percent, which is a significant change but not
one that alters the clear association of high-activity states with larger minority
populations and of low-activity states with smaller minority populations.
Although high-activity states had a higher percentage of citizens whose income
fell under poverty guidelines, the difference between state activity categories was not
substantial. Added to this, low-activity states which we would anticipate to have a below
average percentage of low-income citizens instead had an above average proportion of
their citizens under the poverty level. Both factors do not create confidence in the utility
70
of describing state activity categories by their mean percentage of low-income citizens at
this point.
The only religion measure to show a consistent pattern of distribution across state
activity categories was Catholicism. Low-activity states had a smaller mean percentage of
Catholics than those states that experienced at least seven private school choice bill
introductions. These states, however, are only four percentage points below average for
state Catholic populations, and the total disparity between the percentage of Catholics in
high-activity states and the percentage of Catholics in low-activity states is only about six
percentage points. Still, the weighting of Catholics toward low-activity states is
straightforward, and no state presented an outlier situation on either end of the scale.
Despite religious conservatives being considered a natural constituency for private school
choice and evangelicals conventionally falling into the religious conservative description,
it was actually low-activity states that contained the largest percentage of evangelical
citizens. The descriptive significance of this result, despite its intriguing counterintuitive
implications, is undermined somewhat by the inconsistency of high-activity states
possessing the next largest percentage of evangelical citizens. It seems that even if
religious conservatives can be considered part of the pro-private school choice reform
coalition that relationship precludes does not readily help describe states that experience
higher or lower private school choice legislative activity, at least according to the
admittedly general data presented here.
The demographic patterns demonstrated for state legislative attention categories
or bills favorably reported out of substantive committee, as displayed in Table 3.2, reveal
71
that the state population characteristics which were descriptively suggestive of bill
introduction categories also tend to be descriptively suggestive of bill attention
categories. Low bill attention states were significantly below average in population and
also below average in terms of racial diversity, while higher and average bill attention
states were both more populous and diverse. The proportion of Catholic citizens in a state
is the only demographic characteristic that does not follow suit with its results for bill
introduction activity. Low-activity states may have had the lowest proportion of Catholic
citizens but low attention states have the highest proportion of Catholic citizens. As with
the bill activity results, evangelical citizens and weekly church attendees do not differ
across bill-attention categories in ways that inspire any confidence in their significance as
descriptive factors for high or low-attention states.
72
Table 3.2: State Demographics across Bill Attention Categories

All States (50) High Attent.
States (12)
Average Attent.
States (16)
Low Attent.
States (22)

Population
(in millions)

Mean 5.6m 6.8m 6.8m 4.1m
Min – Max 500 k – 33.9 m 1.2 m – 20.9m 750k – 33.9m 500 k – 19.5m
Std. Deviation 6.21m 6.2m 8.0m 4.4m

% White Pop

Mean 73.1% 66.8% 73.4% 76.2%
Min – Max 24.9% – 95.3% 41.7% – 93.1% 24.9% – 90.3% 57.7% – 95.3%
Std. Deviation 15.2 15.6% 18.1% 10.2%

% Black Pop

Mean 9.9% 9.2% 8.8% 11.1%
Min – Max .3% – 37.2% .8% – 35.5% .3% – 28.9% .5% – 37.2%
Std. Deviation 9.6% 8.9% 9.7% 10.2%

% Poverty Level

Mean 12.4% 13.1% 11.9% 12.3%
Min – Max 7.7% – 20.3% 7.7% – 20.3% 7.8% – 14.2% 8.1% – 19.9%
Std. Deviation 3.2% 3.78% 2.0% 3.6%

% Evangelical

Mean 26.4% 23.5% 28.1% 26.8%
Min – Max 7% – 53% 7% – 37% 18% – 51% 10% – 53%
Std. Deviation 12.4% 9.6% 9.3% 15.5%

% Catholic

Mean 22.2% 23.5% 20.1% 23.0%
Min – Max 5% – 43% 10% – 32% 7% – 31% 5% – 43%
Std. Deviation 10.1% 6.1% 7.7% 13.1%

% Weekly Church
Attendees

Mean 38.8% 38.8% 39.3% 38.5%
Min – Max 22% – 60% 23% – 57% 31% – 54% 22% – 60%
Std. Deviation 9.6% 10.4% 6.9% 11.0%

State Public School System
It makes sense that the legislative activity of policy reforms designed, framed, and
argued to be a solution to the problems facing public school systems would be, to a
73
certain extent, dependent upon the performance of state public school systems. Tables 3.3
and Table 3.4 use the following public school measures to help get a handle on the extent
to which this dynamic is actually present in state legislative activity. These measures are
also snapshots taken toward the middle of the bill collection timeframe:
1) Overall score for math for state 8
th
grade students on the Department of
Education’s National Assessment of Education Progress (NAEP) exam
2) Overall score for reading for state 8
th
grade students on Department of
Education’s National Assessment of Education Progress (NAEP) exam
3) Percentage of state public school students graduating from high school
4) Total number of state public schools located in a large city
5) Overall percentage of state public schools located in a medium or large city
74
Table 3.3: State Public School Figures across Bill Introduction Categories

All States High-activity
States (12)
Average-Activity
States (24)
Low-activity
States (14)

NAEP Score
(8
th
grd math)

Mean 277.4 275.2 278.9 277.5
Min – Max 261 – 291 263 – 286 266 – 291 261 – 287
Std. Deviation 7.4 7.3 7.4 7.6

NAEP Score
(8
th
grd reading)

Mean 262.9 260.8 263.1 263.9
Min – Max 251 – 273 252 – 271 251 – 270 253 – 273
Std. Deviation 5.9 6.9 6.1 5.2

Graduation Rate

Mean 73.0% 68.7% 74.0% 74.6%
Min – Max 54% – 93% 58% – 81% 54% – 93% 62% – 88%
Std. Deviation 8.4% 7.5% 10.4% 6.6%

# Urban Schools

Mean 239.6 396.3 253.8 144.0
Min – Max 0 – 1891 0 – 1891 0 – 1768 0 – 1538
Std. Deviation 417.5 522.5 434.4 324.4

% Urban Schools

Mean 22.1% 28.7% 22.1% 18.5%
Min – Max 3% – 48% 12% – 48% 7% – 40% 3% – 38%
Std. Deviation 9.6% 10.1% 8.1% 8.7%

All of the state education characteristic patterns displayed were consistent across
state activity categories and suggest that high, average and low-activity states are
descriptively distinct when compared by their public education outputs and urban school
character. Looking at public school output measures, it seems that although the overall
differences between public school outputs across state activity categories were not very
dramatic, they were consistent. Public school students in high-activity states scored lower
on national assessment exams and were less likely to graduate from high school relative
75
to the scores and graduation rates of students and schools in average and low-activity
states.
The distinction between high-activity states and non-high-activity states is much
more dramatic when compared by their urban school numbers. High-activity states are
significantly more likely to have more public schools located in a large city and to have a
larger proportion of their total public schools located in an urban area. As with state
population numbers, the disparity between the urban school numbers of the most
populous states – California (1768 schools), Texas (1891), and New York (1538) – and
the other 47 states is large enough to skew the means to a degree that needs to be
addressed. These three states have more than double the number of urban schools of the
states with the fourth and fifth largest number of urban schools, Arizona (734) and
Illinois (613). If we cap the urban school numbers of these three states to 1,000 in an
effort to mitigate their impact on the means, high-activity states have a mean of 345.8
urban schools and average-activity states have a mean of 160.1. These numbers still serve
to highlight the relationship between high-activity states and urban schools. The
proportion of states’ total public schools that can be classified as urban also points to the
validity of describing high-activity states as characterized by a more urbanized school
system; the difference between high and average-activity states is more than twice the
difference between average and low-activity states.
Turning to state bill attention categories, we see that, similar to high-activity
states, states with high rates of bill attention possessed the lowest mean scores on national
assessment exams for middle school students in both math and reading, and along with
76
the lowest mean percentage of students graduating high school. We also see that the
public school performance differences between categories is not very large, especially as
compared to the differences between categories displayed between high bill-attention
states and other states when it comes to urban schools. It seems that urban schools aptly
described those states whose legislatures provided private school choice bills more
attention as well as those states that experienced high levels of bill introductions.

Table 3.4: State Public School Figures across Bill Attention Categories

All States High Attent.
States (12)
Average Attent.
States (16)
Low Attent.
States (22)

NAEP Score
(8
th
grd math)

Mean 277.4 275.1 277.5 279.2
Min – Max 261 – 291 261 – 286 266 – 291 262 – 287
Std. Deviation 7.4 8.4 7.1 7.1

NAEP Score
(8
th
grd reading)

Mean 262.9 261.2 263.0 264.0
Min – Max 251 – 273 252 – 271 251 – 273 253 – 270
Std. Deviation 5.9 6.7 6.0 5.2

Graduation Rate

Mean 73.0% 68.9% 73.8% 75.1%
Min – Max 54% – 93% 59% – 84% 54% – 93% 60% – 88%
Std. Deviation 8.4% 7.3% 8.2% 9.0%

# Urban Schools

Mean 239.6 464.9 202.8 109.8
Min – Max 0 – 1891 0 – 1891 0 – 1768 0 – 330
Std. Deviation 417.5 623.3 374.9 110.5

% Urban Schools

Mean 22.1% 26.7% 21.5% 19.3%
Min – Max 3% – 48% 3% – 48% 7% – 40% 7% – 30%
Std. Deviation 9.6% 12.4% 8.6% 7.4%

77
State Political Character
The ideological nature of private school choice reform which implicates small
government and free-market preferences, convention and school choice scholarship
would suggest that states with stronger measures of political conservatism and
Republican partisanship will be more legislatively active. Table 3.5 and Table 3.6
examine this dynamic, using the following four measures of state politics:
1) Number of times the state’s Electoral College votes went for the Republican
candidate for president in the 1996, 2000, 2004, and 2008 elections
2) State citizen liberal ideology score on a 100-point scale
12

3) Total number of years both chambers of the state legislature had Republican
majorities between 1997 and 2007
4) Total number of years the state governor’s office was occupied by a Republican
between 1997 and 2007

12
Scale and measures developed by Berry et. al.(1998) and updated through 2006.
78
Table 3.5: State Politics across Bill Introduction Categories

All States High-activity
States (12)
Average-activity
States (24)
Low-activity
States (14)

GOP Prez Nominee
Vote Pattern

Mean 2.04 2.2 1.5 2.9
Min – Max 0 – 4 0 – 4 0 – 4 0 – 4
Std. Deviation 1.5 1.5 1.7 1.6

Citizen Liberal
Ideology Score
(0 – 100)

Mean 49.4 50.7 53.5 41.3
Min – Max 22.7 – 81.5 33.6 – 80.7 26.6 – 81.5 22.7 – 60.1
Std. Deviation 14.7 14.4 14.4 12.9

GOP Control
(Legislature)

Mean 4.1 yrs 3.2 yrs 3.5 yrs 5.3 yrs
Min – Max 0 – 11 yrs 0 – 11 yrs 0 – 11 yrs 0 – 11 yrs
Std. Deviation 4.5 3.9 4.4 4.8

GOP Control
(Governor)

Mean 6.2 yrs 7.2 yrs 5.9 yrs 6.4 yrs
Min – Max 0 – 11 yrs 3 – 11 yrs 0 – 11 yrs 0 – 11 yrs
Std. Deviation 3.4 2.8 2.9 3.7

The data in Table 3.5 offers some rather counterintuitive results for the idea that
increased private school choice bill activity should occur more often in politically
conservative states with conservative legislative leaders, at least for the period of
legislative activity that occurred between 1997 and 2007. It was not high-activity states,
which would have been expected, but rather low-activity states that actually had the
highest measure of political conservatism for three of the four measured political
characteristics. The voters of low-activity states gave the state’s Electoral College votes
to the Republican presidential candidate most often, with a mean of 2.9 out of four
presidential election cycles, and the citizens of low-activity states are noticeably more
79
conservative in their mean ideology score than those of average and high-activity states.
Even more striking is that low-activity states also seemed more prone to elect
Republican- controlled legislatures than other states.
The one political characteristic that does not run counter to conventional
assumptions of private school choice politics is partisan control of the state’s executive
office. States that experienced the most bill activity were also the states more likely to see
a Republican in their governor’s office for a majority of the 11 year observation period.
However, high-activity states are only slightly above the overall state mean, with low-
activity states only slightly below the overall state mean, and average-activity states
displaying a Republican governor mean only slightly below, though even it was not
significantly below the overall mean. All of these results suggest that, when it comes to
identifying states that are more likely to experience higher levels of private school choice
bill, activity politically conservative voters and Republican governments may not be very
important.
The results for the distribution of state political characteristics across state bill
attention categories looks quite different from their distribution across state bill activity
categories (see Table 3.6). When it comes to identifying state legislatures that are more
inclined to provide some policy process attention to private school choice bills political
conservatism is more significant. More politically conservative voters switch from being
found in greater overall proportions in low-activity states to occupying a greater
proportion of high-activity state populations. High attention states had a mean Republican
presidential nominee voting score of 2.4, as compared to a mean of 2.0 for low-activity
80
states and 1.8 for average-activity states. These states also had a mean liberalism score of
44.7, as compared to 49.1 for average-activity states and 52.1 for low-activity states.
Perhaps the most dramatic shift between the two results tables occurs in the measure of
partisan control of the legislature. Low-attention states go from having by far the largest
mean number of years of total Republican control to having by far the lowest mean years
of Republican total control, while high-attention states have a mean somewhat higher
than the overall state average but below the mean of average-attention states. It would
seem that lack of Republican legislative control is more of a bane to bill attention than the
presence of Republican legislative control is a boon.

Table 3.6: State Politics across Bill Attention Categories

All States High-Attent.
States (12)
Average-Attent.
States (16)
Low-Attent.
States (22)

GOP Prez Nominee
Vote Pattern

Mean 2.04 2.4 1.8 2.0
Min - Max 0 – 4 0 – 4 0 – 4 0 – 4
Std. Deviation 1.5 1.2 1.8 1.9

Citizen Liberal
Ideology Score
(0 – 100)

Mean 49.4 44.7 49.1 52.1
Min - Max 22.7 – 81.5 31.8 – 57.9 22.7 – 75.0 26.6 – 81.5
Std. Deviation 14.7 7.2 11.0 19.2

GOP Control
(Legislature)

Mean 4.1 yrs 4.8 yrs 5.4 yrs 2.4 yrs
Min - Max 0 – 11 yrs 0 – 11 yrs 0 – 11 yrs 0 – 11 yrs
Std. Deviation 4.5 3.9 4.4 4.8

GOP Control
(Governor)

Mean 6.2 yrs 6.8 yrs 6.1 yrs 6.3 yrs
Min - Max 0 – 11 yrs 3 – 11 yrs 0 – 11 yrs 0 – 11 yrs
Std. Deviation 3.4 2.8 2.9 3.7
81

Putting the Descriptive Patterns Together
Putting this data together, we get a suggestive description of what high and low
legislative activity states tended to look like between 1997 and 2007, relative to both each
other and overall state averages, as illustrated by state characteristics relevant to the
school choice issue. Given that there is not a large amount of distance between the mean
scores of the different state activity categories on any given measure, and that there is a
considerable amount of overlap in the range between each categories minimum and
maximum characteristic scores, these descriptions should not be overstated. The state
characteristics associated with high bill introduction states and/or high bill attention states
should be considered general tendencies and not defining traits.
82
Table 3.7: Describing High/Low Bill Introduction and Bill Attention States
High Bill
Introduction States
Low Bill
Introduction States
High Bill
Attention States
Low Bill
Attention States

average to large
population sizes

larger minority
populations

larger Catholic
populations

lower indicators of
student performance

higher numbers of
urban public schools

less Republican
controlled legislatures

average to small
population sizes

smaller minority
populations

smaller Catholic
populations

larger evangelical
populations

higher indicators of
student performance

lower numbers of
urban public schools

more politically
conservative
populations

more Republican
controlled legislatures

average to large
population sizes

larger minority
populations

lower indicators of
student performance

higher numbers of
urban public schools

more politically
conservative
populations

average to small
population sizes

smaller minority
populations

less politically
conservative
populations

less Republican
controlled legislatures

According to the descriptions presented, the demographic and public school
characteristics of states can be relevant factors for both state bill-introduction and state
bill-attention activity, though they appear somewhat more relevant for introduction
activity than attention activity. State political characteristics, on the other hand, seem to
have a dichotomous and partly counterintuitive relationship with legislative activity,
share both low-bill introduction and high-bill attention state categories tend to have
relatively more conservative citizens and more Republican legislatures. Looking a bit
83
more closely, it becomes clear that the prima facie counterintuitive relationship is
probably due to the fact that a lot of smaller, less urban states tend to have center-right
politics. Perhaps the best example of this is Wyoming, which is a deep red state with a
long history of Republican legislative dominance that is also the lowest-populace, least-
urbanized state in the nation. Indeed, the 17 states below average in population and urban
schools, and above average in standardized exam scores, had a mean Republican control
of legislature score of 6.9 years, well above the overall state mean of 4.1 years. Those
states that were larger in population and urban-school numbers had lower student exam
scores and were prone to center-right politics, had both a high number of bill
introductions and a high number of those bills receiving some legislative attention.
Arizona and Florida are both examples of this mix of state characteristics, and both states
also have a very high number of bills introduced into their legislature and a very high
number of bills passing out of committee. The more appropriate conclusion is that state
political characteristics are not as relevant as state demographic and public school
characteristics to describing state bill-introduction activity patterns, though they are quite
relevant to describing state bill-attention patterns.
Along these lines, the ideal state for private school choice legislative activity
would be one that had a moderate to large, ethnically-diverse population whose students
were often assigned to urban public schools, yet that was also sympathetic to free-market
ideology and prone to elect Republican legislatures. This explains why Southwestern
states, which tend to have moderate to large populations and large cities as well as more
conservative or free-market sympathies, have such high rates of bill introduction as well
84
as a high proportion of bills surviving the winnowing process of substantive committees.
It also explains why Northeastern and New England states, which tend to have more
average populations and large urban centers for their geographic size but less sympathies
for conservative or free-market politics, have higher rates of bill introduction activity but
very low numbers of bills surviving the committee process.

Chapter Three Conclusion

Private school choice legislation had a widely varying and somewhat complex
presence in state legislatures between 1997 and 2007. In some states it’s presence has
been very strong, appearing multiple times in each legislative session and with multiple
proposals acquiring at least some agenda attention – a scarce resource in the American
legislative process – while in other states it’s presence has been very weak, appearing
only a few times or less over the 11 year observation period and largely being ignored
once introduced. The complexity of private school choice legislative activity shows up in
those state policy venues where a large number of bills are introduced and none of them
make any appreciable progress on the agenda. There were even a few states in which
private school choice reform was only introduced a handful of times but was able to gain
a place on the legislative agenda and pass favorably out of committee.
By analyzing the distribution of collected private school choice bills across all
fifty states and analyzing where those bills received at least some measure of legislative
attention, this chapter has provided clear indications that states where private school
85
choice legislation had a larger presence tended to possess present different social,
educational, and political environments from states where private school choice’s
presence was minimal. In particular, it has suggested that states with average to large
populations, with more racial minority citizens, with comparatively lower performing
public schools, and with a lot of urban area public schools are more likely to see state
legislators attempting to initiate private school choice legislation and more likely to see
their legislative bodies indicate a willingness to at least consider the issue. It has also
suggested that states with more conservative populations and with more Republican
controlled legislatures are more likely to see private school choice bills make some
progress through the legislative process once they are introduced, though these
characteristics are apparently not indicative of actually seeing more bills introduced in the
first place.
Now that we have identified the major variations and complexities in the private
school choice landscape both in terms of the policy content of what was introduced into
state legislatures and in terms of the policy venues in which that content was introduced,
the next chapter begins the process of putting them to the test and setting up the logit
analysis of private school choice bill success.
86
CHAPTER FOUR:
COMPETING EXPLANATIONS FOR PROPOSAL SUCCESS

The analyses conducted in the previous two chapters have described a private
school choice reform struggle that is both diverse and complex. Chapter Two illustrated
that the policy arsenal of private school choice proponents has involved several major
design dimensions that can and have been molded and combined to create a wide range
of policy designs with distinct implications for the size, scope, and potential impact of
any resulting policy program. Chapter Three illustrated how this diverse selection of bills
was attempted in a large number of different states, each of which presented a range of
different socioeconomic and political characteristics that in turn created potentially more
or less favorable environments for legislative activity associated with private school
choice reform.
This chapter lays out the possible explanations for the complex landscape of
policy change described. As discussed below, the literature provides three general
families of explanations: public problems and resources for reform; political power for
reform; reform content and framing. Each family and how it is measured is discussed
separately, beginning with public problems and resources, than turning to political power,
content and framing.

87
Political Science, Policy Change, and Families of Explanation

Theodore Lowi (1964 and 1979) described the policymaking process as a system
of limited participation in which administrators, Congressional committee leaders, and
economic producers would work together to develop and implement policy decisions in a
particular issue area with minimal input from the broader public. This became known as
the “subgovernment” or “subsystem” model of policymaking (Baumgartner and Leech
1999). This model suggested that once a particular set of policy arrangements became
accepted as the status quo, policymaking on the issue would result in a slow, deliberate,
and incremental rate of policy change that usually demonstrated a preference for the
status quo over new ideas (Lindblom 1959 and 1977; and Wildavsky 1984).
13

Since the identification of subgovernment politics and subsystem policymaking, a
number of scholars have directed their energies toward explaining the occurrence of
policy change in the face of these obstacles. Perhaps the most famous explanation comes
from John Kingdon (1985), who made the argument that successful policy change
involves effectively attaching preferred policy solutions to recognized public problems in
the right political environment. Sabatier and Jenkins-Smith’s (1993 and 1999) “Advocacy
Coalition Framework” presents a similar take on policy change within the context of
explaining struggles over policy ideas between competing advocacy coalitions. They

13
Lowi, Lindblom, and Wildavsky were primarily concerned with the power of economic producers, but
the policy issue area of public education has come to closely resemble their description of a subsystem,
with public schools and public school supporters now playing the role of an entrenched and powerful
producer group whose representatives (primarily teachers’ unions and public school administrators) work
with state education agencies and legislative committees to protect existing public school policy
arrangements from outside reform efforts.

88
hypothesize that in any given issue area, policy change is unlikely to occur without
“changes in socioeconomic conditions, public opinion, system-wide governing coalitions,
or policy outputs from other subsystems.”
I suggest that the following three points can be reasonably derived, either directly
or indirectly, from the fairly straightforward formulations about policy change of
Kingdon and of Sabatier and Jenkin. Explaining successful policy change involves
accounting for:

1) The socioeconomic context those ideas are presented in – i.e., the pool of possible or
recognized public problems available, and the social and governmental resources for
innovation available;

2) Who holds political power, and who their constituents are – i.e. the political will to
recognize and/or prioritize certain public problems over others and the inclination to view
a particular new idea as a possible solution;

3) The content and framing of policy arguments associated with new ideas – i.e. the way
in which alternative policy solutions are attached to public problems and presented to
decision-makers.

From these three points stem three primary families for organizing key
explanations of individual private school choice bill enactment. Point One indicates that
89
the socioeconomic contexts in which new ideas and policy alternatives are presented
matter in terms of creating public problem opportunities and organizational resources that
policy change advocates can use to leverage new ideas into political action for new or
revised policy programs. From this is derived the Problem/Resource family of
explanations which includes the socioeconomic explanations regarding state wealth,
urbanization, racial demographics, and education performance. Point Two indicates that
the political views, orientations, and constituent pressures of elected officials and their
appointees are a key force in changing the conditions surrounding a policy subsystem so
that a reform idea can gain legislative attention and support. From this is derived the
Political Power family of explanations which includes the political explanations
regarding partisan control of government, electoral competition, professionalism of the
legislature, and voter ideology. Point Three indicates that the arguments of advocates
matter in that they can effectively present developed ideas as specific and viable solutions
to one or more recognized public problems. Although policy change scholarship has
mostly considered framing and idea content dynamics in a macro-policy change context
involving general reform issues, longer timeframes, and multiple policymaking stages, it
is still on this basis that we derive the Content and Framing family of explanations which
includes explanations regarding the four key policy dimensions discussed in Chapter
Two: subsidy mechanism, value qualification, target populations, and funding and
participation caps.
Each of these families of explanation are further discussed and explained by
building a more specific grounding in the political science literature for each particular
90
family and specifying the set of hypotheses and measures associated with it, beginning
with problem/resource explanations and moving, in turn, through political power and
content and framing explanations. Table 4.1 summarizes which hypotheses, measures,
and expected results belong with each family of explanation.

Table 4.1: Summary of Explanatory Families, Hypothesis, and Measures
Explanation
Family

Hypothesis Expected Outcome Measure
Public
Problems &
Resources
State Wealth Wealthier states increase enactment
chance

State gross domestic product for
2003
Urbanization More urbanized states increase
enactment chance

Number of schools located in a
“large city” in 2002
Racial
Demographics
Larger white majorities decrease
enactment chance

Percent of state population self-
identifying as “white”
Public school
competitiveness
Lower public school performance
increases enactment chance

“Education State Ranking” scores
for the 2005 – 2006 school year
Political
Power
Partisan Control Republican control of govt.
increases enactment chance

GOP majorities in both chambers (0,
1) & GOP gov. (0, 1)
Inter-party
competition
Lower inter-party competition
increases enactment chance
Average inter-party balance of seats
held by majority party during
observation

Professional
legislature
More professional legislatures
increase enactment chance

Part-time leg = (0), Hybrid leg = (1),
Full-time leg = (2)
Citizen Ideology Less conservative populations
decrease enactment chance

Average liberalism score for citizens
from 1997 to 2006
Content
& Framing
Subsidy
Mechanism
Voucher subsidy bills decrease
enactment chance

Bill establishes a voucher subsidy
(0, 1)
Subsidy Value
Qualification
High subsidy values decrease
enactment chance

Subsidy values of 50% of Per Pupil
Amount or higher (0, 1)
Target Populations Targeting benefits to one or more
student pop. increases enactment
chance
Bill includes 0, 1, 2, or 3 provisions
targeting benefits to a student group
&; bill includes provision targeting
benefits to disability stu. (0, 1), low-
income stu. (0, 1), urban sch. (0, 1),
low perf or overcrowded sch. (0, 1)

91
Table 4.1 Summary of Explanatory Families, Hypothesis, and Measures
(Continued)

Funding & partic.
caps
Hard/specific maximums on subsidy
funding or participation increases
enactment chance
Bill includes a provision specifying
maximum funding/participation (0,
1)

Control/Other Expansion Bill Expanding on previously enacted
programs increase enactment chance
Bills that only add to the scope,
funding, size, participation, etc. of
previously enacted laws (0, 1)

Federal Policy
Changes
Introduction after the passage of
NCLB or the Zelman ruling increase
enactment chance

Bills introduced after 2002 (0, 1)
State Const.
Language
Introduction into states with a
Blaine amend. or Compelled
Support clause decrease enactment
chance

Blaine Amendment (0, 1)
Compelled Support (0, 1)
Charter School
Law
States with existing Charter School
law decrease enactment chance
State contain some form of charter
school law (0, 1)

Public Problem and Resource Explanations

The socioeconomic aspect of policy change is a primary concern for comparative
state innovation studies, the policy change scholarship that analyzes and compares policy
variation and policy innovation across the American states. These types of state policy
studies usually emphasize the economic, social, and political characteristics of states in
explaining variation in policy innovation (see for example Dye 1966 and Hofferbert
1974). Some scholars writing in this genre have conceived of states with certain
important characteristics as having general, non-specific inclinations to adopt new
policies. Operating under this conception, states that are larger and wealthier (Walker
1969), or with professional legislatures (Sigelman and Smith 1980), or that have more
moralistic cultures (Elazar 1984), or that are dealing with the social problems that come
from more urbanization and industrialization (Hofferbert 1966), may be more likely to
92
adopt new policy alternatives in general. Others have focused more tightly on issue-
sensitive factors for explaining variation in innovation. For example, when it comes to
morally-significant policy changes, the level of religious fundamentalism in state
populations appears to play a role in the choice to innovate (Mooney and Lee 1995; Berry
and Berry 1990). Accordingly, analysis of why private school choice bills achieved
enactment should account for the possibility that the socioeconomic environment of some
states creates both a set of public problems relevant to public education reform and a set
of available resources for private school choice innovation.
The description of state legislative activity in Chapter Three identified several
socioeconomic characteristics as having a prima facie relationship with the introduction
and consideration of private school choice legislation. State population levels, racial
demographics, urbanization, and existing educational outputs were all characteristics in
which high bill activity and high bill attention states demonstrated significant differences.
High-activity and high-attention states were more likely to contain larger, more racially-
diverse populations with more urban schools and lower scores on national assessment
exams. The following hypothesis builds off of these characteristics in generating
explanations more specific to private school choice bill enactment.

Public Problem & Resources Hypothesis: State Wealth
An early conclusion of state policy innovation research was that larger, wealthier,
and more urbanized states were more likely to innovate in general (Walker 1969,
Hofferbert 1966, Gray 1973). These types of states were theorized to have more
93
resources to draw from, as well as a larger organizational capacity to facilitate the
adoption of new policies. This certainly makes sense with regard to school choice issues,
as higher populace states are more likely to have both larger tax revenues and larger
budgets from which to fund private school choice experiments. The analysis of private
school choice activity suggested that the majority of reform bills introduced into
legislatures tend to be limited programs, states with larger revenues and budgets would be
more likely to allow for a limited program to be adequately funded while keeping its
impact on the total state budget comparatively small. The description of high and low
legislative activity states suggested that states with low levels of legislative activity were
more likely to have below-average population and tax revenue figures.
Although high-population states usually have a larger tax base and thus larger tax
revenues and budgets than low-population states, aggregate tax revenues do not
necessarily, in and of themselves, inform a state’s wealth. Low-population states with
higher-income populations are generate more resources and revenue relative to their
overall population than large states with larger total budgets. Since wealth and population
do not necessarily go together, a state’s available resources for innovation can be, in part,
attributed to general wealth independently of its population size.

Public Problem & Resources Hypothesis 1: Private school choice bills are more
likely to get signed by the governor and enacted into law in wealthier states.

A state’s overall wealth is measured using state gross domestic product data per
capita collected by the Census Bureau for 2003.
94

Public Problem & Resources Hypothesis: Urbanization and Urban Schools
States with larger populations also tend to have more and/or larger urban centers,
which can influence the perceived public problems associated with education policy.
Large urban school districts tend to be the most costly, under-funded, overcrowded, and
difficult to organize and administer public school systems in a state: a situation that gets
amplified when families with the will and resources to engage more constructively with
education providers choose instead to opt out of struggling urban schools and relocate to
better performing suburban districts, which than increases the likelihood that the
remaining parents and students in these districts will become “trapped” in low performing
schools without the same options to remedy their situation. These issues tend to be
readily and consistently highlighted and publicized by urban community leaders, who
also are often inclined to view school choice policies as a viable solution to these
problems (Moe 2001; Fusarelli 2003). To the extent that private school choice policy
options can be presented as a way to provide those particular sets of families with the
resources to remedy their situation, we may expect to find that more urbanized states
foster more opportunities for advocates to make that crucial connection between their
preferred policy goals and recognized public problems.

Public Problem & Resources Hypothesis 2: Private school choice bills are more
likely get signed by the governor and enacted into law when introduced into the
legislatures of more populous and urbanized states.
95

Urbanization is measured using figures provided by the Department of Education
Statistics’ Common Core of Data that counted the number of school districts classified as
located in a large city in 2002 for each state.

Public Problem & Resources Hypothesis: Racial Demographics
The racial composition of a state’s student population should be significant in a
couple of respects. The first is that racial minority students and families seem to bring a
different set of difficulties to the provision of education, as indicated by the consistently
lower rate of freshmen who graduate as compared to white students.
14
States with larger
racial minority communities are more likely to be dealing with the problems of the racial
gap in student performance, a public problem school choice proponents can capitalize on
by framing private school choice bills as a way to provide minority students more and
better educational opportunities.
15
The data on private school choice bill-introduction
activity provides a general suggestion that states with larger populations of minorities are
more likely to attract more legislative proposals. The second respect is that significant
evidence suggests that low-income and urban families can be a latent support base for
school choice reforms (Moe 2001; Fusarelli 2003; Gokcekus, Phillips, and Tower 2004).

14
As represented by National Center for Education Statistics, annual reporting of graduation and dropout
rates.

15
E.g. Howell, William G. and Paul E. Peterson (2002), The Education Gap, Brookings Institution,
Washington D.C.
96
Public Problem & Resources Hypothesis 3: Private school choice bills are less
likely to get signed by the governor and enacted into law in states with larger white
majorities regardless of the state’s overall population.

White population percentages are measured using self-identified racial
distribution data collected by the U.S. Census Bureau for 2002.

Socioeconomic Hypothesis: Educational Competitiveness
Existing policy standards and outputs can also affect the combination of
recognized education-related public problems and innovation resources (Sabatier and
Jenkins 1993 and 1999). Private school choice proposals are, by and large, presented as a
cost-effective way to improve school performance and empower parents. It makes sense
that this argument would find more acceptance among policymakers when there is some
indicators to suggest that the state’s current public school outputs are sub par, especially
in relation to other states. Indeed, policy innovation scholarship has identified the
competitive position of a state in relation to other states as a motivator for adopting
policy changes (Peterson 1995). As policymakers in states with below-average public
education outputs attempt to increase their system’s competitive position in relation to
other states, they can be reasonably expected to be more open to education reform in
general and more receptive to private school choice reform arguments in particular.
Conversely, in states where public schools are perceived by policymakers and the public
as successful and/or competitive, there should be less of a felt need to consider reform
97
proposals that alter the status quo, especially when those proposals tend to generate vocal
opposition from teachers unions and the parents and neighbors associated with high
performing public schools (Brunner and Sostelie 2003; Kenny 2005).

Public Problem & Resources Hypothesis 4: Private school choice bills introduced
into state legislatures in states with lower levels of public education performance and
output are more likely to get signed by the governor and enacted into law.

Education performance requires a measure that legitimately and consistently
compares statewide public school performance and outputs across states. Although
several indicators present themselves – high school graduation and drop-out rates,
average scores for math and/or reading on the National Assessment of Education
Progress exam carried out by the U.S. Department of Education, and average class size,
among others – these indicators are by nature partial assessments of a state’s education
system. In order to use a more complete measure, the “Education State Rankings”
published by CQ Press for 2005 – 2006 will be used. Its assessment and subsequent
scores and rankings combine 21 indicators of public school performance (these factors
will be provided in Appendix B), including graduation rates, exam scores, class size,
student to teacher ratios, expenditures toward instruction, average daily attendance, and
teacher salaries. These indicators are compared to the national average, resulting in a
positive score for states above the national average and a negative score for those states
below the average, with state scores ranging from -17.81 to 17.58 (see Morgan and
98
Morgan 2005). The 2005-2006 assessment will be used, as it is the earliest measure
within the observation time-frame in which factors of student achievement were
emphasized over factors of education spending.

Political Power Explanations

Scholars have long accounted for political will, pointing out that significant
change in who holds control of government has important implications for what policy
changes reach the policy agenda (Ginsberg 1976; Sinclair 1977; and Brady 1982 and
1989). Kingdon (1985) contributed to this argument by demonstrating in interviews with
political insiders that elected politicians are considered some of the most influential
policy actors and that a significant change in controlling politicians affects the range of
policy problems given agenda attention as well as the range of policy alternatives given
consideration. The political power of electoral constituencies may also come into play, as
the re-election incentive may often be so powerful that politicians are more concerned
with claiming credit for legislation favored by their constituents than with creating
"good" policies (Mayhew 1974). The state innovation literature usually includes
considerations of elections and partisanship in its studies, with one of the most prominent
considerations being the political orientations of both elites and the masses (Blomquist
1999). External political pressures can also move politicians to work against or oppose
certain legislation and scholars have pointed to the ability of politically influential
99
interests to prevent policy changes that challenge their current benefits and privileges (see
for example: Lowi 1964 and 1979).
Chapter Three’s examination of private school choice proposal distribution across
state legislative venues indicated that Republican partisan control of state legislatures and
the political conservatism of state voters weighted more heavily on those state
legislatures that ended up reporting a high and average number of proposals out of their
initial committee assignment, with low bill-attention states being significantly more
likely to avoid Republican control of government. The political hypothesis generated
below will also focus on partisan dynamics and voter conservatism.

Political Power Hypothesis: Partisan Control
By its very nature, private school choice strongly implicates values particularly
associated with center-right politics and conservative ideological values. The intent of
free-market education reform, as laid out by Friedman (1965) and Chubb and Moe
(1990), is to decrease public and bureaucratic control over education and expose public
schools to free-market competition. Naturally, this approach is more likely to be viewed
favorably by those who place more faith in individuals, markets, and competition, and
less faith in government authority and collective decision-making to yield effective
results: a trait more conventionally associated with contemporary American conservatism
and the Republican Party – i.e. its platform, its elected officials, and its supporters. The
idea that Republicans at both the mass and elite levels are significantly more likely to
view private school choice as a viable and desirable policy alternative has a strong anchor
100
in the existing scholarship on school choice political coalitions (Moe 2001; Brunner,
Sostelie, and Thayer 2001; Brunner and Sostelie 2003; Gokcekus, Phillips, and Tower
2004; and Kenny 2005).
Political parties hold strong importance to private school choice reform due to the
overwhelming tendency of teachers’ unions to support Democrats over Republicans. At
both the state and national level, teachers’ unions are one of, if not the, most important
sources of interest group opposition to education reform in general (Boyd, Plank, and
Sykes 1998) and school choice reform in particular (Mintrom 2000). They are generally
heavily invested in the established public school system and view school choice reforms
as challenging and undermining that system. Although the level of opposition teachers’
unions direct at public school choice proposals can vary (Mintrom 2000), when it comes
to ideas that allow public monies to support private education providers, their opposition
is generally unequivocal and supported by the national organization and other labor
organizations.
These ideological and interest-group factors serve to highlight the applicability of
the policy change scholarship’s focus on changes in political and partisan leadership for
understanding the legislative progress of private school choice bills. The American model
of divided political power creates a system of at least three major institutional
checkpoints in process of drafting laws, with policy proposals having to succeed in two
distinct legislative chambers and survive an executive veto. Majority party legislative
leaders have several tools at their disposal, such as control of committee chairmanships
and floor scheduling, that provide them with significant discretion as to which bills will
101
be placed on the agenda at crucial veto points in the legislative process. Governors tend
to be the most recognizable political leader in the state, providing them a potential bully
pulpit greater than that provided to legislative leadership. The combined clout of
executive veto and executive power affords governors powerful bargaining tools that can
make their opposition or endorsement of a policy idea more significant to its success.
Republican control over any, and especially all, of these lawmaking institutions should be
likely to increase the chances of a private school choice bill to achieve higher levels of
progress in a state’s legislative process.

Political Power Hypothesis 1: Private school choice bills introduced into state
legislatures in which the Republican Party controls both the legislature and the
governor’s office are more likely to get signed by the governor and enacted into law.

Partisan control is measured using two dichotomous variables. The first indicates
whether that particular state’s Republican Party has a majority in both of the state’s
legislative chambers, coded as 1, or in only one or neither chamber, coded as 0, during
the year the particular individual private school choice bill was introduced.
16
The second
is a dummy variable indicating whether or not the state’s sitting governor is a member of

16
It may be noted that this measure does not capture whether the Republican Party controls the particular
chamber a bill is introduced in, which could possibly make a difference in terms of whether or not
Republican control influences legislative outcome. However, it seems likely that private school choice
advocates attempt bill introductions in Republican-controlled chambers first. Also, most bills are able to
find sponsors in both chambers and are often introduced into both chambers at roughly the same time,
meaning that most bills will be introduced into a Republican-controlled chamber if such a chamber exists at
the time of introduction
102
the Republican Party during the year that a bill is introduced into the state’s legislature.
Cases are coded as 1 if this is the case and 0 if it is not.

Political Power Hypothesis: Inter-Party Competition
Moving beyond simple partisan control of a legislative institution, the level of
inter-party competition in a state political system may also significantly influence the
willingness of a state legislature to take on a private school choice proposal. In general,
large electoral margins in state legislatures have been viewed as indicative of lower inter-
party competition, which allows more elected representatives to feel more secure in their
positions and can lead them to be more willing to attempt more controversial policy
changes (Berry and Berry 1999). Assuming that private school choice reforms can
accurately be considered as controversial policy options, increased electoral security and
decreased party competition should help create a political environment more favorable to
legislative progress since legislators will have less reason to fear the issue being used
against them as they seek re-election. Larger margins in states with Republican
legislative control would also seem likely to provide a more secure position for
conservative political leaders to risk bringing private school choice bills up for a vote. By
increasing the number of Republican defectors needed to defeat a bill up for a floor vote,
the chances of political leaders to suffer an embarrassing defeat should theoretically
decrease, making floor votes and floor vote success more likely.

103
Political Power Hypothesis 2: Private school choice bills introduced into state
legislatures with lower levels of inter-party competition are more likely to get signed by
the governor and enacted into law.

Inter-party competition is measured using the inter-party balance of seats held by
the majority party in the legislature overall. This was calculated by counting the total
number of seats contained in both chambers of a state’s legislature and than measuring
the margin of difference between the number of seats held by Democrats and the number
of seats held by Republicans during the 11 year observation period. For example, the
Alabama state legislature during the legislative session of 2000-2001 had 140 seats
between its House and Senate; the Democrats held 92 of those seats and the Republicans
held 48, meaning that the Democrats held 65.7 percent of the seats available in the
legislature and held a majority balance of 31.4 percent.
17

Political Power Hypothesis: Professional Legislature
The institutional characteristics of legislatures has also been a consideration in
policy innovation studies, as the longer sessions and larger staffs of more
professionalized legislatures have been argued to have a higher tendency to adopt policy
reforms (Sigelman, Roeder, and Sigelman 1981). Such legislatures are considered to have
a greater capacity to deal with a wider range of public problems at any given time, as well

17
Although it has been suggested that using district level election vote totals are the more accurate way of
measuring inter-party competition, the use of the inter-party balance of seats held in the state legislature has
typically been the more common approach to measuring such competition in state politics studies
(Barrilleaux, Holbrook, and Langer 2002) and is more suitable for an analysis of the scope presented here.
104
as more resources to create coherent and effective policies. This can be an important
factor in whether private school choice reform bills get the attention of legislators or their
staffs on a crowded legislative agenda.

Political Power Hypothesis 3: Private school choice bills introduced into more
professional state legislatures are more likely to get signed by the governor and enacted
into law.

The professionalism of state legislatures is measured by utilizing the National
Conference of State Legislatures’ categorization of full-time legislatures as opposed to
part-time legislatures. They categorize legislatures roughly as full-time, part-time, or a
hybrid based on the amount of time a legislator needs to spend fulfilling their legislative
duties, the size of their legislative staffs, and the extent to which they are compensated
enough to devote their full attention to their legislative duties – in other words, whether
the legislature is organized to allow legislators to act as a full-time professional or
structured toward the concept of citizen-legislators; private professionals who also act as
part-time legislators. Full-time legislatures that have longer sessions, larger staffs, and
higher compensation are coded 2. Part-time legislatures that have short sessions, small
staffs, and low compensation rates are coded as 0. Hybrid legislatures have some
combination of these characteristics or toe a middle ground between the high and low on
each and are coded 1.

105
Political Power Hypothesis: Citizen Ideology
Constituency pressures have been found to impose meaningful constraints on
legislators (Kingdon 1989) and influence policy change (Verba and Nie 1972; McClosky
and Zaller 1984). American political parties are not homogenous organizations, and
elected officials from a particular party do not necessarily have the same constituency
pressures placed on them across states. It is not only important to identify what the
partisan majorities are in a particular state government during the session of a particular
proposal, but also to identify the extent to which the constituency environment of those
politicians is more closely associated with general conservative ideology. Higher levels
of conservatism among a state’s citizenry have been linked to variation in state welfare
payments (Hill and Leighly 1992) and state revenue allocations to school districts (Wood
and Theobald 2003). Both economic and religious conservatives have been found to
express support for private school choice reform proposals (Moe 2001; Brunner and
Sostelie 2003). Political policymakers seeking to get elected or govern a state with a
more moderate or liberal set of voters may not be able, or perhaps even willing, to
approach a particular issue in the same way as politicians working with a more
conservative set of voters.

Political Power Hypothesis 4: Private school choice bills introduced into states
with more politically conservative populations are more likely to get signed by the
governor and enacted into law.

106
Political conservatism is tested by using the state citizen ideology measures
originally developed by Berry et al (1998) and updated through 2006. Their ideology
measure focuses on the left-right ideology of district incumbents over the ideology of
their challengers and looks at district support for the incumbent than complies this data to
create an average score for the state as a whole.
18
This results in each state having an
overall score theoretically ranging from 0.00 as fully conservative and 1.00 as fully
liberal. For this analysis, the citizen ideology scores for a given state from 1997 to 2006
are averaged to come up with an overall state citizen ideology score for the period of
observation. The state scores are averaged because they tend to vary little during the
period of observation, and it is expected that consistent citizen ideology over time will to
have a greater impact on structuring the constraints on state legislators than a single
year’s score that seems to be an outlier relative to the state’s average score. In this way,
the measure seeks to limit the impact of outlier years and focus instead on a state’s
general citizen ideology relative to other states, a characteristic that is probably more
observable and identifiable to private school choice advocates than seeking to identify
year by year changes in citizen ideology.

18
For a full explanation of their measures,0 see William D. Berry, Evan J. Ringquist, Richard C. Fording,
Russell L. Hanson., 1998. “Measuring Citizen and Government Ideology in the American States, 1960-93”.
American Journal of Political Science, Vol. 42, No. 1 (Jan.), pp. 327-348.

107
Content and Framing Explanations

A significant portion of policy change scholarship has focused on the notion that
ideas matter and that policy decisions are not just about who holds political power at any
given time but are at least in part about the substance and presentation of the policy
change being considered. Kingdon argued that "the content of the ideas themselves, far
from being mere smokescreens or rationalizations, are integral parts of decision making
in and around government (1985)." Derthick and Quirk (1985) and Mucciaroni (1995)
demonstrate the ability of ideas to be persuasive in such a way as to enhance their ability
to “catch on” and overcome powerful and entrenched interests. Baumgartner and Leech
(1999) note that "policies are made in government through a process of argumentation"
and that "literature has begun to attempt to build explicit models of issue-definition,
framing, and argumentation into studies of lobbying." Baumgartner and Jones (1993) and
McFarland (2006) make a legitimate case for the importance of defining issues and
framing arguments in explaining policy change over time. These scholars suggest that
policy change is in part, about tailoring general policy ideas to fit the particular policy
problems that are currently attracting the attention of decision-makers and getting
decision-makers to pay attention to particular policy problems.
Although Sabatier and Jenkins (1999) do not directly incorporate issues of content
and framing into their original policy change hypothesis, the revised version of their
policy change hypothesis concedes that external shocks to a policy subsystem are a
“necessary but not sufficient” condition for policy change, and that the other part of the
108
equation is the need for a minority coalition to effectively capitalize on the change in
political environment. Citing Mawhinney’s (1993) study of Francophone advocacy for
French-speaking secondary schools in Quebec, they acknowledge the significance of
framing and argumentation in exploiting the opportunities created by shocks to a
subsystem, and that doing so can be done more or less skillfully.
Policy change scholarship has had a tendency to study policy content and framing
more broadly, discussing the ability of reformers to frame policy ideas and connect them
to recognized problems in a macro policy-change, context instead of a micro policy-
change context. By macro policy-change I mean they focus on the development,
argumentation, adoption, and implementation of general reform ideas over time instead
of focusing on the passage of particular legislative reform proposals that represent the
particular way those ideas have taken physical form.
19
Those studies that have analyzed
particular legislative reform attempts have generally been case studies designed to
develop a compelling historical narrative about how a particular policy change idea was
or was not able to build a reform coalition and gain traction on the policy-making
agenda.
20
These macro policy change studies have convincingly highlighted the
importance of reform advocates building coalitions, making arguments that highlight
previously ignored public problems, defining existing policy issues in contrast to by the
existing subsystem, and framing reform ideas as a solution to those previously ignored

19
Perhaps the most celebrated example of this type of scholarship is Baumgartner and Jones’s (1993) study
of reform advocacy efforts in several policy issue areas to develop their punctuated equilibrium model of
policy change in which reform ideas are presented and argued in multiple policy venues until they
eventually catch on and rapidly replace the issue definitions of existing policy arrangements.

20
E.g., Jacob Hacker’s (1997) analysis of why President Clinton’s push for healthcare reform through the
idea of managed competition ultimately failed to create policy change.
109
problems. However, school reformers have been highlighting the problems of
bureaucratic inefficiencies, lack of accountability, performance inequalities, and union
capture in public education for decades – at least since the Reagan administration
commissioned “Nation at Risk” report was published – and the need for some form of
public education reform has been gaining currency for just as long. Accordingly, it seems
that a pertinent question at this point may not be whether private school choice policy
alternatives can and have been framed to address the increasing problems and criticisms
associated with public education, but whether the particular ways they are presented to
policymakers can have distinct and significant relationship to policy change in a micro
policy-change context.
Using individual bills as the unit of analysis allows for a more specific analysis of
the potential relationship between bill content or policy design and the chances of bill
enactment. As shown in Chapter Two, private school choice legislation can be very
diverse in terms of the combination of subsidy mechanisms, values qualification, target
populations, and funding or participation restrictions incorporated into a given bill. These
content characteristics can result in proposals that are quite expansive and ambitious,
offering relatively attractive subsidies to a large portion of a state’s student population, or
they can result in weaker proposals that are narrow and supplemental in their policy
impact. Content characteristics can also help to frame a proposal as a more specific
policy solution to a particular public school-related problem. Given the politically
controversial and policy-consequential nature of free-market education reforms, these
differences in content are expected to influence a bill’s chances of enactment. The
110
following hypotheses are developed to capture the major ways bills can be designed with
different potential policy impacts and the ways they can explicitly implicate particular
framing strategies as outlined in Chapter Two.

Content Hypothesis: Subsidy Mechanism
Subsidy mechanisms can be seen as creating more or less obvious links between
public resources, government power, and private - particularly religious - education
providers. They can also be predicted to facilitate more transfers from public to private
schools based on whether costs are subsidized upfront or through reimbursement. Even
within the private school choice movement itself, issues of facilitating greater choice
versus school responsiveness and bureaucratic influence are not easily settled. Publicly-
funded scholarships are the subsidy framework apt to be most controversial, since they
create an obvious link between public resources and private schools, are conceivably the
easiest mechanism for parents to affordably use, and retain clear bureaucratic
administration of subsidy disbursement.

Content Hypothesis 1: Private school choice bills that propose a voucher program
are less likely to get signed by the governor and enacted into law than private school
choice bills that propose expense or donation tax credits.

If the bill’s language establishes a scholarship or certificate to be paid for using
any existing or to-be-established, taxpayer-funded revenue source and that can be
111
redeemed at a privately-affiliated, non-charter school, than it is coded 1; all others are
coded 0.

Content Hypothesis: Value Qualification
The value of the subsidy offered by a private school choice proposal relative to a
state’s existing public school funding per pupil can vary widely between different
proposals. Bills that contain qualifications that limit subsidy values to percentages or
dollar amounts that are relatively low, compared to the state’s average per pupil funding
for public schools, have the potential to significantly reduce the threat of large numbers
of families suddenly switching out of their local school. It can also signal to policymakers
that the proposed program’s intended constituency will be families who are dissatisfied
enough with their public school options that they are willing to make up the difference
between the subsidy and the actual private school costs. Since subsidy values tend to load
differently on different subsidy mechanisms…

Content Hypothesis 2: Private school choice bills that explicitly subsidize private
school costs at higher rates in relation to the state’s average per pupil funding are less
likely to get signed by the governor and enacted into law than private school choice bills
that subsidize costs at higher rates.

Value qualification is tested through a set of dummy variables indicating whether
the bill contained a value qualification of 50 percent of state PPA or more (yes = 1, no =
112
0). If a bill specifies a specific dollar amount instead of a percentage value of state per
pupil funding, the dollar amount is compared to the dollar amount of state per pupil
funding for the year it was introduced and converted into a percentage.

Content Hypothesis: Target Populations
Chapter Two revealed five distinct types of provisions targeted to student
subpopulations that consistently appeared in bills introduced across multiple state
legislatures: students with disabilities, low-income students, urban school students, poor-
performing school students, and overcrowded school students. It is possible that the
inclusion of any of these common target population provisions increases the chances of
legislative adoption. This is because all of these provisions that direct subsidies to a
specific subset of a state’s student population serve to both mitigate the potential scope
and impact of a proposed program, and link its benefits to students who have arguably
been underserved or disadvantaged under existing public school arrangements.
Considering the scholarly importance of connecting preferred policy solutions to
recognized public problems (Kingdon 1985), bill language that appears to explicitly
direct subsidies and benefits toward problem areas may improve framing efforts and
provide these bills with an advantage.

Content Hypothesis 3: Private school choice bills that are explicitly targeted to a
disadvantaged student population are more likely to get signed by the governor and
enacted into law than bills that are not targeted to any student populations.
113

Overall target population provisions are tested by simply counting the number of
target population provisions a bill contains. Since the fewest number of target population
any bill contained was 0 and the largest number was 3, bills are coded 0, 1, 2, or 3.

It is also possible that some target population provisions are more significant than
others. This could be because they are perceived as being more likely to limit the policy
impact of the proposed program, or because they are linking the program to a more
recognizable problem area or a more sympathetic student subgroup. For instance,
students with disabilities probably represent a smaller portion of the student population
than urban school students as well as probably representing one of the most sympathetic
student populations. The issues surrounding low-income minorities in public education
and problematic urban schools both have a vocal constituency willing to support school
choice reforms (Moe 2001; Fusarelli 2003; Gokcekus, Phillips, and Tower 2004).

Content Hypothesis 3a: Private school choice bills that are explicitly targeted and
limited to either – students with disabilities, low-income students, urban school students,
poor performing school students, or overcrowded school students – are more likely to get
signed by the governor and enacted into law than private school choice bills that are not
targeted to any of these student populations.

Individual categories of target population provision are tested using a dummy
variable for each different type of population targets, so bills are coded as follows:
114
a) If it contains language that limits eligibility for participation in the proposed
program to only students who are determined to have a qualifying disability, it is
coded as 1, all others as 0
b) If it contains language that limits eligibility for participation in the proposed
program to only students whose families are determined to be low-income, it is
coded as 1, all others as 0
c) If it contains language that limits eligibility for participation in the proposed
program to only students who are attending or assigned to attend schools that are
part of a large urban school districts, it is coded as 1, all others as 0
d) If it contains language that limits eligibility for participation in the proposed
program to only students who are attending or assigned to attend schools
determined to be underperforming or failing coded as 1, all others as 0
e) If it contains language that limits eligibility for participation in the proposed
program to only students who are attending or assigned to attend schools and/or
school districts determined to be overcrowded, it is coded as 1, all others as 0

Content Hypothesis: Funding and Participation Caps
Provisions that limit program funding and/or participation serve to more explicitly
mitigate the budgetary and policy impacts of a proposed private school choice policy.
Bills that contain these restrictions - as opposed to bills with no limits or soft, flexible
limits - generally create limits significant enough to ensure that a program will, at least
initially, constitute a relatively minor part of overall education policy. Legislative
115
proposals that are designed and framed to be small and experimental should be able to
mitigate concerns about unintended or unanticipated policy consequences relative to
more ambitious reform proposals. Bills with hard limits may also have an easier time
being framed as a supplemental solution to particular public education policy problems
and allow proponents to downplay arguments that the bill is intended to destabilize, or
will result in undermining, existing public schools.

Content Hypothesis 4: Private school choice bills that contain specific limits on
overall subsidy costs or overall participation are more likely to get signed by the
governor and enacted into law than bills that contain more flexible limits or no limits.

Content Hypothesis Four is tested by coding a bill as 1 if it contains language that
specifically limits the amount of money available to fund a proposed program, or the
number of families and/or students who can participate in the program, to a particular
level – i.e., hard limitations. Bills that contain language that create more flexible – i.e.,
soft – limits, or that contain no such language, are coded as 0.

Other Possible Explanations

There are some other factors that may also be relevant to school choice reform but
do not readily fall into one of the three explanation families outlined. Two of these
factors – expansion bills and federal policy change – can be considered possible policy
116
resources that school choice proponents can use to increase the chances of private school
choice success in a state’s legislative process. Two other factors, state constitutional
language and existing charter school law, are existing policies that may provide school
choice opponents with useful resources to frame private school choice reforms as
unnecessary or legally improper. To properly examine the influence of proposal content,
state socioeconomics, and state government politics on the enactment of private school
choice bills, the possible impact of these other factors needs to be controlled.

Control Hypothesis: Expansion Bills
In states that have already enacted some form of private school choice legislation,
subsequent legislation can either expand on the existing framework or propose a new and
distinct reform law. Usually, when a private school choice law already exists, a new and
distinct reform law proposes a different type of subsidy mechanism – e.g., a donation
credit law already exists and a voucher law is being proposed – or targets a different
student population subgroup. It makes sense that expanding on laws that have already
been enacted would be easier than getting lawmakers to approve a law with different
subsidy mechanisms or framed and targeted to a different student group.

Control Hypothesis 1: Private school choice bills that are expansions of
previously - passed private school choice laws are more likely to get signed by the
governor and enacted into law than bills that are a new policy innovation.

117
To test this hypothesis, bills that increase the benefits, eligibility, or funding of
previously-enacted laws are coded as 1; bills that do not contain language expanding
previous law are coded as 0.

Control Hypothesis: Federal Policy Changes
Sabatier and Jenkins (1993 and 1999) hypothesize that policy core attributes can
be changed - i.e. major policy change can occur - by the outputs of “a hierarchically
superior jurisdiction”. Although federal level power has not formally imposed any type of
private school choice incentives or schemes on the states, 2002 heralded a couple of
significant federal policy changes that may influence state level decision-making on the
issue. The first is the No Child Left Behind Act, signed January 8th 2002, which requires
that the progress of all public school students in math and reading be tested every year in
grades 3 through 8, and at least once in high-school in order to receive federal funds. It
also requires that states make available "report cards" on public school performance on
these tests and establish standards of achievement. Schools that do not achieve these
standards and do not show improvement for two consecutive years are required to
provide supplemental instruction and allow students an option to transfer to better-
performing public schools. This creates an established mechanism for singling out public
schools as inadequate and may create tools and rationales that can be built into private
school choice arguments and proposals.
The second federal policy change is the Supreme Court's Zelman v. Simon-Harris
decision, which reversed an appellate court decision and was handed down on June 27th
118
2002. It ruled that it is constitutional for Ohio’s school voucher program in Cleveland to
use public money to subsidize parochial school tuition because parents, not state officials,
were deciding at which institution their public assistance money would be spent. The
ruling signaled to both state governments and school choice proponents that the federal
court system would not easily strike down enacted private school choice programs on
establishment clause grounds, mitigating a significant source of policy uncertainty given
the Cleveland voucher cases’ demonstration of the ability of school choice opponents to
challenge private school choice policies on establishment clause grounds.

Control Hypothesis 2: Private school choice bills introduced into state legislatures
after the passage of the No Child Left Behind Act the U.S. Supreme Court’s Zelman v.
Simmon Harris ruling are more likely to get signed by the governor and enacted into law.

Both the No Child Left Behind Act and the Supreme Court’s decision occurred
within about six months of each other (January 8 and June 27,

respectively) and both are
hypothesized to favor the enactment of private school choice laws. Thus, both federal
policy changes will be measure as a single dummy variable, with all bills introduced
before the 2002 legislative session coded as 0 and all bills introduced during or after that
session are coded as 1.

119
Control Hypothesis: Restrictive State Constitutional Language
State political leaders and school choice proponents have to deal with their state
court system as well, which has the potential to strike down an enacted private school
choice program and create policy uncertainty just like the federal court system. The
existence of constitutional language that can be interpreted as prohibiting or restricting
government from providing public money to parochial schools could be considered as
ready-made tools that school choice proponents can use to persuade a state court to strike
down a private school choice program, creating enough uncertainty about future state
court actions that they influence the ability of proposals to succeed. Two types of
constitutional provisions have historically been adopted by states in order to restrict
taxpayer monies from going to religious institutions: Blaine amendments and Compelled
Support clauses. Although these constitutional characteristics are from the 19
th
and 18
th

centuries, respectively, their relevance to contemporary school choice advocacy grew in
importance after the U.S. Supreme Court removed establishment clause violations as a
viable basis for court challenges.

Control Hypothesis 3: Private school choice bills introduced into state legislatures
in states whose constitution contains language originally intended to prevent public
money from being used to support parochial education are less likely to get signed by the
governor and enacted into law.

120
Anti-school choice language is measured using two dummy variables, each
denoting the presence of absence of either a Blaine Amendment or compelled support
clause; those with amendment or clause presence are coded 1 and with absence coded 0.

Control Hypothesis: State Charter School Law
States that already possessed charter school laws experienced significantly more
bill introductions and reported more bills out of committee than states that possessed no
charter school law or that passed a law after 1997. This is significant because ,given the
identifiable divide between public and private school choice policy ideas, with the former
being able to garner broader political acceptance than the latter, it is reasonable to suspect
that the existence or passage of charter school laws may be used by private school choice
opponents to redirect school choice reform energy toward policy options that maintain
the link between public funding and public control of education in order to keep private
market dynamics from gaining a policy foothold. Existing charter school laws may also
be used by opponents to argue that school choice reform has already been tried in the
state and that more time is needed to study and understand the extent to which it is
effective. In a paper examining the impact of teachers’ union power on education reform,
Terry Moe (2005) notes that the unions often treat charter school laws as their
compromise school choice reform, using it as a fall-back position in order to yield some
ground in school choice debates without funding schools completely outside the public
sector. Since charter schools are considered independent but not private direct political
121
control is maintained over the new choice schools, allowing for the possibility of
alteration and even rollback down the road.

Control Hypothesis 4: Private school choice bills introduced into state legislatures
in states that already possess a charter school law are less likely to get signed by the
governor and enacted into law.

State charter school laws are measured using a dichotomous variable indicating
the presence or absence of a charter school law before the private school choice law was
introduced, with the presence of a charter school law coded as 1 and the absence of a
charter school law coded as 0.

Chapter Four Conclusion

This chapter has utilized the relevant political science literature on policy change
and state policy adoption in conjunction with the descriptive information provided in
Chapter Two and Three to develop three distinct families of explanation and 16
hypotheses, which together account for the public problems and reform resources
presented by states; the reservoir of potential political will for reform both inside and
outside of elected office; and the policy impact and policy framing presented by
individual reform proposals. Previous studies of policy change have tended to either
study only content and framing factors or only socioeconomic and political factors for
policy change, instead of trying to account for all three sets of explanations at the same
122
time. Taken in combination, these hypotheses incorporate many of the lessons from
previous scholarship on American policy change while also directly including measures
of individual policy design and potential policy impact. In this way the developed
hypotheses have the potential to a) provide a more comprehensive approach to explaining
why and how individual private school choice bills succeed in state legislative processes,
and b) test the extent to which measures of policy design variation are able to add to our
overall understanding of private school choice politics and policy change. In the next
chapter we will see to what extent they are able to fulfill that potential.
123
CHAPTER FIVE:
THE DETERMINANTS OF SUCCESS

Of the 568 bills introduced into state legislatures, nineteen bills in ten different
states were able to successfully survive the legislative gauntlet and become enacted into
law (see Map 5.1, below). Of those nineteen bills, four were substantive expansions of
previously enacted programs, while the other fifteen bills created a private school choice
subsidy that did not exist in the state before. This is a very small percentage of all
collected bill introductions during the 11-year observation period and starkly
demonstrates the difficulty facing free-market education policy change efforts. Thus, this
chapter focuses on providing a better understanding of the factors that increase the
enactment chances of private school choice bills introduced into state legislatures. Given
the amalgam private school choice bills introduced into state legislatures, what types of
bills in what types of state legislatures had a better chance of overcoming the policy-
change resistance inherent to American legislative processes in general, and to a long-
established public education policy subsystem in particular?
This question is addressed through a logit analysis of the relationship between
each of the three families of explanation with their associated hypotheses – as developed
in the previous chapter – and the decision by state lawmakers to pass or reject a given
private school choice bill. To set up the logit analysis, the process of collecting the data to
be analyzed is explained and the dependent variable specified. The results of the logit
estimation are then explained, elaborated, and interpreted.
124
Map 5.1: Private School Choice Proposal Outcomes by State

No bills signed into law
At least one bill signed into law

Data Collection and Measures

The data to be analyzed consists of individual private school choice bills
introduced in any of the fifty American states from 1997 to 2007. This time frame begins
right after the concerted push for private school choice bills at the state level truly got
underway in 1995. It also occurs after limited voucher programs had been adopted for
125
Milwaukee, Wisconsin (1989), and Cleveland, Ohio (1995). In this way, all cases fall
under the context of private school choice as an already established issue on the
education reform agenda with existing, observable voucher programs. Because both the
passage of No Child Left Behind, and the Zelman v. Simmons-Harris school voucher
decision land in the middle of the selected time frame (January and June of 2002,
respectively), the selected cases also allow for an important variation between proposals
introduced before and after these potentially relevant policy events were in place.
The collected dataset of cases was acquired by searching the Lexis-Nexis State
Capital database for individual bills introduced into each state’s legislature between
January 1, 1997 and December 31, 2007. Bills comprising the dataset contain language
that specifically proposes: a publicly-funded tuition scholarship program, a tax credit or
deduction for education expenses, or a tax credit for donations to private scholarship
organizations. Of course, the legislative process is messy, and legislators may be more
creative in their bill labels and descriptions, as well as their use of parliamentary rules,
than the search parameters in a Lexis search; undoubtedly some independent private
school choice bill introductions have fallen through the cracks. To compensate, the data
collection process was supplemented by utilizing databases of voucher and tax
credit/deduction legislation maintained by organizations involved in tracking school
choice laws and policies: the Heritage Foundation, the Education Commission of the
States, and State Net Capital Journal. These supplementary data sources recorded not just
legislation that was enacted but also bills that made legislative progress or garnered
media attention. This made them more useful to an effort to capture all significant pieces
126
of voucher and/or tuition tax break legislation proposed at the state level between 1997
and 2007. The Heritage database, in particular, made a considerable effort to provide
more complete histories of significant school choice legislation for each state, though
even it did not capture as many bills as a more labor intensive Lexis search.
Each bill was coded according to their introduction characteristics (i.e. date of
introduction, state introduced in, etc.), its last recorded legislative action (i.e., assigned to
committee, reported out of committee, etc.), and its content (i.e., type of reform proposed,
participation restrictions, etc.). Each case was earmarked with three primary points of
variation: year, state, and proposal content. If two private school choice proposals
substantially match along all three points, they were not considered as distinct cases. If
two proposals differ on any one of these points, they were considered distinct cases, even
if they match perfectly on the other two dimensions. Thus, a proposal was treated as a
distinct case from other proposals as long as it met one of the following conditions:

1) It was substantively different (i.e., contained distinct funding requirements,
participation restrictions, implementation periods, etc.)
2) It was proposed in a different state legislature
3) It was proposed in a different year

Under this collection scheme, the same bill can be introduced into the same
legislature in a subsequent legislative session and still be considered a distinct case. Five-
hundred-sixty-eight distinct cases of private school choice legislation were collected,
127
including 23 cases in which a proposal did not create a new program but expanded the
size and/or scope of a previously enacted program, making the total number of distinct
cases 545. These cases represent voucher activity as it occurs in different states, different
years, under different political conditions, and under different existing educational and
demographic environments.

Dependent Variable
As has been alluded to in previous chapters, the dependent variable for the
analysis will be individual private school choice bill success in achieving enactment into
law, meaning that the unit of analysis is individual bills. Typically, state policy output
studies use states as their unit of analysis because they are approaching the subject of
policy change from the perspective of pinning down why some states innovate or adopt
new policies. To accomplish this, they have incorporated interval level measures of
policy outputs – such as public expenditures or number of welfare recipients (Hill and
Leighley 1992, Barilleaux and Miller 1988) – or they have used a dichotomous measure
indicating state adoption or non-adoption of a type of policy change over a period of
time, including the adoption of lotteries (Berry and Berry 1990), hate crime laws (Haider-
Markel 1998), and charter school laws (Mintrom 1999), to name just a few of the many
issues covered by such analysis. These measures do a fine job of capturing comparative
differences in adopted policy outputs between states, but they do not capture comparative
differences between policy change legislation attempted and policy change legislation
enacted. This study is different in that it approaches policy change and state policy
128
outputs from a position of trying to understand if and why some policy change legislation
has a better chance of achieving enactment. This means accounting for both the
characteristics of what was proposed and the characteristics of the states in which bills
were proposed. A focus on individual bill adoption, although requiring significant data
collection effort, facilitates the goals of the analysis better than a focus on state policy
adoption. Accordingly, the dependent variable is a dichotomous measure with a value of
1 when a bill has been signed into law by a state’s governor and a value of 0 if the bill
died in the legislative process.

Logit Results

Due to the categorical nature of the dependent variable – enactment or non-
enactment of an individual bill – logit estimation was used to test the 16 hypotheses
developed in the previous chapter. Table 5.1 displays the results of the analysis and
predicted probabilities.

Table 5.1 Private School Choice Bill Success: Logit Results

Hypothesis Variable Expected
Change
(+/-)
Coefficient
(Std. Error)
Change in x
(from, to)
Marginal
Effect per
unit
change in
x

Socio-
Economic

# Urban Schools

+

.001 (.001)

(.49, 33.87)

.04%
Gross State Product per cap.

+ .050 (.086) (26.1, 64.5) .17%
% white population

– .103** (.047) (24.9, 95.3) .86%
Public school performance

– –.157** (.075) (-17.81, 17.58) –.77%
129
Table 5.1 Private School Choice Bill Success: Logit Results (Continued)

Political GOP control of legislature

+ 1.694** (.772) (0, 1) 1.16%
GOP control of Governor

+ 1.460** (.772) (0, 1) .66%
inter-party balance of seats

+ – 2.879 (3.583) (6, 70) –.65%
professional legislature

+ .914 (.687) (0, 2) 1.11%
Content/
Framing
Subsidy mechanism = voucher

– –.214 (.642) (0, 1) –.10%
High Subsidy Value

– –.322 (.760) (0,1) –.16%
Hard limits - funding / partic.

+ 1.944*** (.682) (0, 1) 2.06%
Students with disabilities

+ 2.153** (.912) (0, 1) 2.98%
Low income students

+ –.380 (.680) (0, 1) –.17%
Urban school students

+ – 2.625 (1.542) (0, 1) –.57%
Low perf. / overcrowded sch.

+ .799 (.851) (0, 1) .53%
citizen left/right ideology

– –.019 (.053) (22.7, 81.5) –.13%
Control Expansion bill

+ 2.187** (.917) (0, 1) 3.33%
NCLB / Zelman Decision

+ –.716 (.709) (0, 1) –.34%
Charter School Law

– .857 (1.429) (0, 1) .32%
Blaine Amendment

– –.013 (1.020) (0, 1) –.01%
Compelled Support Clause

– –.833 (.806) (0, 1) –.43%
Constant

– 16.034
(6.579)

N

562
Log Likelihood

– 54.864
Prob > chi2

.0001
Pseudo R2

.365
Correctly Classified

96.80%
Note: *p < 0.10, **p < 0.05, ***p < 0.01

In line with the literature, each family of explanation is relevant in some way.
Critical for this project, given its deliberate choice to use individual policy proposals as
130
the unit of analysis and individual bill enactments as the dependent variable, is the logit
model’s estimate of a relationship between the specific policy content included in
legislative proposals and bill enactment. Bills that specifically limit a proposed policy’s
budget impact and participation levels, and bills that are targeted to students with a
disability, have a positive relationship with bill enactment and are statistically significant
at the 0.05 level. Although subsidy mechanism would seem to be a fundamental point of
design difference between proposals, the subsidy mechanism measure was not
statistically significant, though bills that contain voucher subsidy mechanisms had the
hypothesized relationship of decreasing the odds of bill enactment. High-value
qualifications were also in the hypothesized negative direction and not statistically
significant. It was somewhat surprising to see that no other provisions directing subsidy
benefits to particular student populations were estimated to have much of an impact on
the chances of bill enactment besides those targeted to disability students. Considering
the rhetorical emphasis that many school choice advocates and scholars have placed on
using subsidies as a way to address issues of social equity and fairness – an emphasis
captured by the fact that low-income targets were included in almost one third of all
introduced bills – it was surprising to find that low-income provisions are not only not
estimated to be significant but, indeed, had a negative coefficient.
As hypothesized, and in conformity with conventional associations of partisan
leadership to legislative behavior and of Republican partisanship with free-market policy
solutions, Republican control of a state’s legislative and executive institutions has a
significant and positive effect on the chances of a private school choice bill achieving
131
enactment. Republican control of the legislature was in the hypothesized direction and
significant at the 0.05 level. While a Republican governor was also estimated in the
hypothesized direction, it was only significant at the less desirable 0.10 level. On the
other hand, neither electoral margins nor the professional nature of the legislature are
estimated to influence bill enactment chances. It may well be the case that these
particular factors come into play earlier in the legislative process, such as during
committee debate and floor consideration. It certainly seems logical that electoral
margins would matter more during floor votes than they would for overall bill success,
which is a multi-institutional process. The level of conservative sentiment present in a
state’s population underperformed in the logit model, as it was not estimated to be
statistically significant and was actually estimated with a coefficient in the non-
hypothesized direction. It seems that although free-market policy reform is associated
with conservative politics, it is not so easily associated with conservative populations.
The only explanation from the public problem and public resource family to be
significant and in conformity with its specified hypothesis was the measure of a state’s
existing public-education ranking. It is significant at the 0.05 level and in the
hypothesized negative direction, indicating that states with lower-ranked public schools
and less-competitive public school systems are more likely to have an observed private
school choice bill enactment. The only other socioeconomic explanation to have a
statistically significant relationship with bill enactment was state racial composition;
however, despite its statistical significance, the logit estimation indicates that states with
higher percentages of white citizens actually had a positive influence on the odds of
132
enactment instead of the hypothesized negative relationship. This result is especially
puzzling given that, in Chapter Three, we found states with high bill introductions and
high bill attention scores also tended to have smaller white majorities in their population,
while states with low bill-introduction and bill-attention scores tended to have larger
white majorities. The idea that states with larger white majorities were less likely to
witness a bill successfully being reported out of committee – but more likely to enact a
bill into law – seems quite counterintuitive, and at the very least means the null
hypothesis for the relationship between race and bill enactment cannot be rejected.
Looking at the control variables, bills that are designed to expand previously
enacted private school choice laws are estimated to be significant at the 0.05 level and in
the hypothesized positive direction. This confirms the importance of controlling for bills
designed to expand previously enacted law and, unsurprisingly, suggests that these types
of bills have an easier time getting signed into law than bills proposing new policy
innovations. The two measures of federal-level policy decisions related to school choice
did not have a discernable influence on bill passage, since introducing a bill after 2002’s
double policy change – No Child Left Behind and USSC’s favorable voucher decision –
was not significant to enactment and actually had a negative coefficient in the model.
However, neither of these policy changes really provided a direct incentive for states to
adopt private school choice as much as they simply made it a bit easier for lawmakers to
pursue school choice policies if they wished to do so. At the state policy level, both
Blaine and Compelled Support amendments in state constitutions were also not estimated
to be significant. Considering the older and more arcane nature of these provisions, their
133
contemporary weight in a state’s jurisprudence may simply be too uncertain and in doubt
for policymakers to concern themselves – unless, of course, the state courts force them to.
To further analyze the coefficients, the changes in predicted probability of bill
enactment that occur in conjunction with changes in each independent variable’s value
are also calculated and displayed. The change from 0 to 1 is used for all dichotomous
variables, while the change from one-half standard deviation below mean to one-half
standard deviation above mean is displayed for the non-dichotomous variables. The
default probability of bill enactment for the full model is 0.48 percent, or less than a one
percent chance of seeing an introduced bill achieve enactment. This is a very small,
though not surprising, probability of observing any particular bill getting signed into law.
Relative to this small chance of success, the increases in the probability of success
estimated for value changes in the independent variables are significant. For instance,
changing from a non-Republican governor to a Republican governor yields the smallest
increase in bill success for a statistically significant variable, yet even at 0.66 percent this
change increases the probability of bill enactment by over 100 percent of the default
probability.
The second-largest increase in bill enactment probability is seen with expansion
bills. The increase in probability estimated between new innovation bills and expansion
bills is almost three percent (3.33%), or over almost six times the default probability.
Expanding on previously enacted private school choice laws and programs is estimated to
have a substantially increased chance of success than attempting initial legislative reform
proposals. The variables that were estimated to have the first-and third-largest increase in
134
bill enactment probability were the two content variables estimated to be statistically
significant. Specifically targeting a bill to students with disabilities ends up increasing the
probability by over six times the default probability (2.98%), while including specific
cost and eligibility caps increases it by more than four times the default probability
(2.06%). Republican control of both legislative chambers is significantly more relevant to
bill enactment than Republican control of the executive, as the former increases
enactment chances to a little over 1 percent (1.16%) while the latter only increases those
chances by a quarter of a percent over the default probability (.66%). The public problem
measures are actually estimated to have a larger effect than a GOP governor – though,
interestingly enough, the race measurement was estimated to increase enactment
likelihood more than the public school measurement (0.010 to 0.008), despite a
coefficient result in the opposite of our hypothesized direction. Overall, the chances of
observing a private school choice bill enactment are fairly small, but the content variables
outperform the political and socioeconomic variables by a significant margin when it
comes to increasing the probability of witnessing a bill enactment.

Causation and Political Battle over Private School Choice

Correlation is not causation, and although a significant and hypothesized
relationship was estimated in the logit model for at least one factor for each of the three
families of explanation, it is worthwhile to explore whether there is a basis for inferring
that the observed relationships can and/or should be considered causal relationships. In
135
this regard, we turn to a few examples of struggles over private school choice legislation
to anecdotally illustrate how the statistically significant variables can and have led to bill
progress or failure, beginning with the more established causal relationship between
policy adoption and socioeconomic and political variables.

Public School Ratings, Partisan Politics, and Private School Choice Bill Success
Both socioeconomic and political explanations already have a well-established
theoretical basis as causal factors in state policy change studies. This scholarship has
typically drawn from systems theory, approaching policy change as occurring when
socioeconomic factors influence the development of political system conditions, which in
turn produce policy outputs. Recent studies conceptually disentangled socioeconomic and
political factors, considering the two factors as more distinct elements in the policy
change dynamic.
21

The concept that the socioeconomic and political factors of public school rank and
Republican control of legislative and executive offices can play an important causal role
in private school choice policy change finds support in the reform struggles that occurred
in Louisiana and Colorado. Media accounts in both states depict state legislators
previously opposed to or unsure of private school choice reform pointing to chronically
problematic schools or school districts as the reason they considered giving current
choice reform efforts a chance. In Louisiana, several members of the state’s Legislative
Black Caucus (representing New Orleans area voters) expressed openness to Republican

21
See Blomquist (1999) for a good review of the use of systems theory in state policy adoption studies.
136
Governor Mike Foster’s Catholic Church-supported push for a voucher subsidy program
targeted to failing New Orleans schools, pointing to the need for drastic action to address
the dismal educational options provided to the students of their constituents. In an
interview with a local reporter covering the voucher attempt for the Times-Picayune, the
president of the Louisiana Federation of Teachers admits that the general frustration of
state lawmakers with the consistently dismal track records of some public schools has
made it more difficult to keep the bill off of the agenda.
22
In Colorado, we find a similar
situation. A Republican Governor, Bill Owens, publicly advocates for a voucher program
as a way to address the issue of failing public schools and, in doing so, attracts the public
support of some Democratic and Hispanic lawmakers who point to chronically
underperforming Denver public schools as the reason for breaking with their traditional
public sector allies.
23

Although the private school choice battles in Louisiana and Colorado in 2003
contained similarities, two key differences illustrate the importance of Republican
partisanship for bill enactment. The first key difference is that in Colorado, the state had
just elected new Republican majorities in each chamber, while in Louisiana, Democrats
still held large majorities in each chamber. The second key difference is that in Colorado,
a voucher bill eventually passed both legislative chambers and was signed into law; not
one of six separate private school choice bills was even passed out of committee in
Louisiana. Media accounts of Colorado’s successful voucher program push partly credit

22
Maggi, Laura, 2003. “Political landscape tipping toward school vouchers; Proponents still face
significant hurdles,” Times-Picayune, March 23.

23
Richardson, Valarie, 2003. “Colorado nears voucher approval; Similar bills pass Senate, House to aid
low-income youths,” The Washington Times, February 27.
137
the success to the Republican legislative takeover, since the state’s Democratic leadership
had demonstrated staunch opposition to voucher proposals.
24
The significance of
Republican governors is aptly demonstrated by the fact that, out of the eight private
school choice bills in the collected dataset that were vetoed by state governors, seven
were Democrats; and the only Republican governor to veto a private school choice bill,
Utah’s Olene Walker, subsequently failed to retain her party’s support for re-election.

Private School Choice Policy Design and Bill Success
The explanatory factors associated with individual policy design and framing are
tenuous in their causal implications. State policy adoption studies generally do not
account for policy design, and the existing literature on the importance of issue-framing
has considered the subject from the perspective of ideas that are able to gain political
traction and agenda attention, as opposed to individual policy proposals able to gain
enactment. Indeed, the system theory approach suggests that the policy design of reform
proposals is properly considered a result of the interplay of socioeconomic and political
factors, and in many ways the content of individual reform proposals is considered a
product of idiosyncratic or case-specific negotiation within a policy coalition and
between reform supporters and potential reform supporters to find a workable
compromise.

24
Morgan, Ryan, 2003. “Voucher bill headed for Senate OK Court challenge predicted,” The Denver Pos.t,
March 27.

138
Despite a lack of theoretical support for causal dynamics in existing literature,
there are good examples of legislative struggles in which proposals presenting smaller
programs with restricted policy impacts were indeed an important causal factor in private
school choice bill enactment. In these examples, on-the-fence legislators became
convinced that private school choice reforms were worth a try if they were limited and
experimental. The first example focuses on the battle over Pennsylvania’s initial private
school choice law. On three separate occasions, Republican Governor Tom Ridge put
significant effort into pushing a voucher program through the state legislature while
Republican majorities controlled both chambers, and on each attempt the bill died in
committee. These voucher proposals did not have any specific caps attached to either
funding or participation levels but were all targeted to moderate-to low-income families
and limited to below 20 percent of average state per pupil funding. Finally in 2001, on the
recommendation of the school choice advocacy groups he was working with, Ridge
backed an indirect tax credit for business donations to non-profit scholarship-granting
organizations that were worth up to $100,000 per participating business – but were also
capped at $30 million annually in funding.
25
It was this attempt that was ultimately
approved by both chambers and sent to his desk for enactment.
Although the voucher program for Washington, D.C., signed into law by
President Bush in 2004 was not a state-level bill enactment, and as such has not been
included in the analysis until now, it provides a useful example of how a private school

25
Woodall, Martha, 2002. “A new way to afford private schools: Pennsylvannia businesses that fund
scholarships may seek tax credits. Lower-income students stand to benefit,” The Philadelphia Inquirer,
January 22.
139
choice program that is framed as a more limited and experimental program can attract
legislative votes that might not otherwise be available. The D.C. School Choice Incentive
Act of 2003 offered publicly-funded tuition scholarships worth up to $7,500 to families
earning less than 185 percent of the poverty level. The program’s funding was capped at
$13 million annually. A key point in the Congressional battle for the voucher program
came when California’s Democratic Senator Diane Feinstein, originally an opponent of
school choice, switched her position and voiced her intent to support the bill. Her new
position broke a stalemate in the Senate Appropriations Committee caused by Republican
Senator Arlen Spector’s opposition to the voucher program and allowed the bill to move
forward. Her declared reasons for the change of heart were that D.C. Mayor Anthony
Williams was on board with the legislation and that the bill was explicitly experimental,
declaring that “local leaders should have the opportunity to experiment with programs
that they believe are right for their area”.
26

In some cases, it is not the undecided or leaning-opposed legislators who limit
their support for private school choice proposals to policies that are explicitly restricted in
their policy impact, but previous key political figures in enacting past voucher legislation
who vocally hold such a position. This was the case in Florida when both the Republican
Governor Jeb Bush and the Republican Senate President John McKay publicly stated
their non-support for an ambitious voucher proposal that would have expanded the state’s
existing Opportunity Scholarship program’s eligibility to all public school students. What
makes their declared non-support even more interesting is that Jeb Bush was the primary

26
Hsu, Spencer S., 2004, “How Vouchers Came to D.C.”, Education Next vol. 4 no. 4.
140
policy entrepreneur behind Florida’s 1999 Opportunity Scholarship voucher program for
students in failing schools, an issue on which he both campaigned on and heavily
prioritized upon taking office in 1999. Senate President McKay was the driving policy
entrepreneur behind Florida’s voucher program for disability students, also enacted under
Jeb Bush in 1999, which is why the law was named the McKay Scholarship for Children
with Disabilities. These two very public supporters of previous voucher laws voiced their
disapproval of the ambitious 2004 voucher bill because, according to Governor Bush, the
ambitious choice reform proposal was not appropriate “at this time”. Senator Mckay also
expressed concerns about the bills impact, stating that he did not know if it is manageable
by the Department of Education.
27
It would appear that even strong school choice
reformers value the framing and design of private school choice reforms to be more
limited and experimental in their impact on existing public school arrangements.

Interpreting the Results

So what types of private school choice proposals in what types of state
legislatures are more likely get enacted into law? Based on the logit results and causation
examples, it would be those proposals that are:
a) explicitly restrained in their policy reform impact, either by including specific
maximums on their funding or participation or by targeting subsidy benefits to students
with recognized learning disabilities, and

27
Ulferts, Alisa, 2002. “Voucher Bill Heads to Full House,” St. Petersburg Times, February 15.
141
b) are introduced into Republican-controlled legislatures in states with a
Republican Governor and a comparatively less-competitive public school system.
Another way to look at these results is that private school choice policy change
was more likely to occur when reform advocates were willing to get behind explicitly
small and incremental reform programs, present them to a less hostile set of decision-
makers, and make a credible argument that education reform is needed and that the
reform proposed is supplementing existing policy arrangements and/or is a limited and
experimental program with clearly-defined cost and benefit.
An important take-away from the analysis, in terms of content and framing
explanations, is that the policy design of private school choice reform proposals has a
clear and distinct relationship with bill enactment success because of its ability to frame a
proposed private school choice policy as a small and experimental program with more
explicitly-defined budget and policy impacts that is intended to supplement the existing
public school system, as opposed to challenge it. Framing a reform proposal this way has
made the difference between support and opposition for both Republican-controlled
legislatures and individual Democratic lawmakers. Even Republican lawmakers with well
established credentials as school choice advocates have seen fit to restrict their support of
private school choice reform to these types of proposals.
As we saw in Chapter Two, the majority of collected private school choice
legislation was designed so that it implicated a more moderate policy impact, with about
80 percent of reform proposals including some form of impact restrictions. However,
despite the diverse ways in which bills were designed to present less-ambitious reform
142
policies, it was only those bills that contained specific funding or participation
maximums, or that were targeted to only those students with a state-recognized learning
disability, that were able to demonstrate a statistically relevant relationship in the logit
model. Out of all the different types of provisions that can and have been included in
private school choice bills, it is arguably these two particular types of provisions that
create the most clearly defined and restrictive policy impacts, making them the most
effective provisions for framing bills as limited and experimental.
Most obviously, hard-cap restrictions on a proposed program’s funding levels or
the allowable number of granted subsidies create very clear impact boundaries, by
definition. No other type of policy restriction so explicitly communicates that the
proposed private school subsidy will only cost “this much” and/or will only cost a
particular set of public schools “this many students,” at most. Soft-cap provisions and
student population targets do place boundaries on the impact of proposed reforms, but
these are categorical – and therefore flexible – limitations as opposed to boundaries based
on hard maximums, therefore they cannot limit reform impact in the same way. Including
a provision that limits subsidy benefits – only to families below the poverty level, or to
schools in a particular city, or to students in first through third grade – may clearly define
the boundaries of who is or is not eligible to receive benefits, but it does not provide a
clear indication of how many families will actually take advantage of the program.
Subsidy value qualifications are even more unclear in the impact boundaries they create.
This is because, to the extent that they are able to restrict the size and cost of a program,
143
it is done by reducing its utility and desirability so that fewer people take advantage of
the subsidy – and predicting what “fewer” will look like is quite a difficult task.
On the other hand, if defined policy impacts are so important and student target
provisions create boundaries that are too flexible, why do disability student targets fare so
well in the logit analysis? This particular student target provision arguably fares well
because students with disabilities represent a conceptually and legally distinct minority in
a state’s student population. Disability students are distinct from the other types of often-
targeted disadvantaged student populations because they are often more sympathetic, and,
under the Individuals with Disabilities Education Act of 1974, they already have a history
of being placed in private schools when public schools are determined not to have
appropriate accommodations.
Students with disabilities are a more sympathetic category of disadvantaged
students because, unlike low-income or urban students, children with disabilities are
more evenly distributed across socioeconomic boundaries, meaning that a broader
spectrum of people can relate to families who deal with the challenges of educating a
student with disabilities. They are also more evenly distributed across schools and school
districts. Low-income and urban students, as well as those in poor-performing schools,
are usually concentrated in certain schools and districts, allowing opponents to argue that
any enacted school choice reform would disproportionately impact those particular
schools and policymakers should instead boost their support for the public schools at
issue.
144
The legal requirement established by IDEA that children with disabilities have
legal right to a free and appropriate education, even if that education must be provided at
a private school, along with the right it granted parents to challenge the adequacy of their
child’s learning accommodations, means that public policies directing state funds to
private schools have a more firm establishment in the law. Many states may already be
using taxpayer funding to pay for some of their disability students to attend private
institutions because their parents believe those schools are better suited to their child’s
needs and mounted a successful challenge to their public school placement. To the extent
that this is already occurring, it is not a significant legal or policy leap to allow parents of
students covered under the IDEA an easier opportunity to choose private institutions.
Turning our attention to public problem explanations, a key aspect of initiating
policy change, according to the literature, is finding and capitalizing on opportunities to
connect a preferred policy solution to a public policy problem that has been or can be
readily recognized by key policymakers. The significance of a state’s public school
system score certainly fits in line with this aspect of policy change. Including hard caps
on funding or participation levels or disability student targets can effectively frame a
private school choice policy solution as limited and supplemental, but those provisions
cannot as effectively make the argument that the existing public school system needs
supplemental reform or that education reform should be a priority of state policymakers.
States with low-ranking public school scores are more likely to have problematic schools
brought to the attention of state policymakers, meaning they are also more likely to be
willing to experiment with education reform. This was the case for the private school
145
choice battles in Louisiana and Colorado during 2003. The public school systems of both
states received negative scores, and in both states, legislators pointed to problematic
schools and the generally-acknowledged need for education reform to justify their
willingness to give a limited private school choice program a chance.
Turning to the political environment private school choice bills are presented in,
the logit analysis and discussion of causation indicate that private school choice is
primarily a partisan issue with much more saliency at the elite level than the mass
population level. In large part, this makes sense, private school choice is a major change
from the government controlled system of free public education that most people have
been familiar with their entire lives. It is also a major policy change that people do not
have much knowledge about, a fact Terry Moe (2001) acknowledges when describing the
issues involved in creating a survey instrument to measure voucher opinions. Most
people have not really heard of private school choice in any form and so do not have a
firm view or understanding of the issue. Indeed, when private school choice reform
attempts have drawn the attention of the general public, they have not fared well. No state
referendum on vouchers – either proposing a voucher program or requesting repeal of an
enacted voucher program – has resulted in majority voter support for private school
choice reform (Kenny 2005). When, under the strong advocacy of Governor John
Huntsman, the Utah state legislature pushed through a very ambitious voucher subsidy in
2007 – it would have offered scholarships worth between $500 and $3,000 depending on
income, to every public school student in the state by 2020 – public school supporters
were able to put the new law to a public referendum vote and get the new voucher
146
program rejected by 62 percent of the voters. Florida Governor Jeb Bush’s Opportunity
Scholarship program, which was not hard-capped – though it was limited to students in
poor performing schools – also generated a majority negative opinion from voters in
post-enactment polling.
28
The lack of a public referendum law in Florida may be the only
thing that saved it from a similar fate.
In both Utah and Florida, people expressed concerns that the new private school
choice programs would have an adverse impact on existing public schools, a concern that
opponents were understandably quite eager to exploit.
29
It would seem that the dearth of
developed opinions on private school choice, combined with the prevalence and
importance of public schools to so many people’s everyday lives, means that the general
populations of even relatively conservative states like Utah and Florida are prone to view
private school choice reform as a threat. Thus, the opposition of Jeb Bush and Senator
McKay to the ambitious statewide voucher program moving through the Florida House
can be viewed as a truly pragmatic position to protect the two fledgling school choice
programs they had worked so hard to previously enact from the negative public
consequences of pushing too hard too quickly.

28
Bridges, Tyler, 2002. “Florda Voters Cool to Governor’s A+ Plan, Poll Says.” The Miami Herald, June
14.

29
Marshall, Janet, 2007. “State expects a rise in voucher offerings.” Sarasota Herald-Tribune, October 24;
“A Choice Showdown.

147
Chapter Five Conclusion

There is a cautious, almost under-the-radar quality to passing private school
choice reform, at least when it comes to initial policy innovation. In many ways, private
school choice reform is politically defined more by its sources of opposition than its
sources of support. Getting introduced into legislatures with Democratic leadership
almost ensures a reform proposal will not become law, but getting introduced into a
Republican-controlled legislature only has a marginal effect on increasing enactment
chances. Even when private school choice was put before Republican legislative
majorities by a Republican Governor it more often than not had to be designed as a small,
limited, and perhaps experimental reform that presented no immediate threat to public
schools. This is, of course, quite understandable, private school choice threatens to make
fundamental changes to a policy arrangement that has remained a static and major part of
most Americans’ lives for generations and has one of the most powerful interests in state
level politics invested in ensuring it continues unadulterated.
Overall, the logit analysis reveals an almost overwhelming resistance to private
school choice reform, the default chance for a private school choice bill to successfully
proceed from introduction to enactment is estimated as a tiny half-percent, which
certainly helps explain why only nineteen bills were enacted out of 568 introduced. This
anemic pass rate for introduced bills can not be blamed on the majority of them being too
ambitious or too often attempted in unfriendly partisan majorities in states with above
average school systems because none of the variables estimated to be significant had a
148
marginal effect on bill enactment chances of more than a 1 or 2 percent. Clearly, passing
private school choice through the legislative process is a difficult uphill battle and there
are no easy levers that proponents can manipulate for anything close to reliable results.
They can gain a small edge by designing reform policies that are unthreatening,
connecting them to real and recognizable deficiencies in the state school system, and
presenting them before Republican leaders who will at least be more likely to give them a
chance to be heard, but these are no substitute for the hard and particular work of framing
arguments, building coalitions, and negotiating compromises on a case by case basis.
149
CHAPTER SIX:
CONCLUSION

Looking back this project has taken a unique approach to studying the
contemporary political struggle over private school choice reform. Previous scholarship
has tended to view the issue of school choice as a free-market reform idea struggling to
gain currency and “catch on” on the contemporary policy and political agenda. Studying
school choice in this vein has led to a focus on examining how the school choice idea as
developed over time, who its champions are, the ways it has been argued and
communicated, and which social and political groups are more inclined to support it.
Scholars have not concerned themselves much with how the school choice idea has
actually manifested in a diverse multitude of legislative proposals with distinct policy
designs implicating a wide range of reform impacts, which have in turn been introduced
before various sets of state lawmakers with very different results. Instead of essentially
ignoring or by-passing this variation, this project get it’s hands dirty with the nitty-gritty
of a decade’s worth of private school choice proposals, 568 in all, revealing what was
proposed as well as what actually passed. It has viewed the school choice issue as a set of
individual policy proposals as opposed to generalizing very different policy reforms into
an abstract whole, assuming that the variation in policy design is important for
understanding private school choice reform and for understanding why and how it passes.
Chapter Two was an effort to get a handle on the policy diversity within the
private school choice family. It demonstrated that private school choice legislation can
come in a dizzying array of shapes and sizes, with severely different implications for
150
their ability to affect or alter existing state public school policies. Analyzing the design
and content of individual legislative proposals revealed four major policy dimensions:
subsidy mechanism, subsidy value qualifications, target student populations, and funding
or participation caps. Each of these four dimensions could be designed to produce
stronger or weaker reform impacts in a multitude of combinations. Indeed, the actual
private school choice bills introduced into state legislatures over the observed decade
seemed to take full advantage of this design variability, since – even when the
dimensions were significantly generalized – no particular combination of them was able
to capture more than a fifth of the collected bill population on its own, and the three most
common combinations were only able to capture 48 percent of that population.
Chapter Three mapped the geographic diversity of the private school choice issue.
Instead of only considering which states had adopted or debated some form of private
school choice innovation, an analysis of individual bills allowed consideration of the
frequency of reform legislation and legislative attention across state policy venues. The
discovered dichotomy between high-frequency states and low-frequency states was
significant. High bill-introduction activity states and low-activity states gape between as
little as zero to as many as 32 bill introductions, with low bill-attention states also at zero
and high-attention states seeing four or more bills go through. However, bill introduction
and attention activity did not always correlate; in other states, like Vermont, many bills
are introduced but paid no legislative attention, while in some, like South Dakota, a few
bills are introduced and half of them receive attention. The chapter demonstrated that
state characteristics matter for describing the different types of states, but they don’t
151
matter powerfully. The correlation isn’t a direct one, as if, say, high-activity states are
almost always also high-population. California isn’t a high-activity state, and it has the
highest population in the country. Characteristics, than, point to tendencies rather than
define traits; they shed light, but as through a glass darkly.
Both state characteristics and bill content provide a base of explanations for bill
enactment. The fourth and fifth chapters made a first cut at trying to explain individual
private school choice bill enactment by incorporating specific policy design variables,
along with the more traditional socioeconomic and political variables, into an explanation
of private school choice policy adoption at the state level. The results of the logit analysis
indicate that each family of variables can play a role in the success or failure of
introduced reform proposals. Most of the socioeconomic and political variables mattered
to success in ways we would expect. Given the conservative and free-market associations
of the school choice issue and the close ties between teachers’ unions and the Democratic
Party, it was important that reform bills were significantly more likely to achieve
enactment when state legislatures were under Republican political leadership or when the
governor’s office was occupied by a Republican. It is also not surprising that the relative
rank of a state’s public schools was also significant to the success chances of bills framed
as necessary reforms of the public school system. Indeed, if these variables did not
matter, there would be concern for the validity of the analysis.
More significantly, the logit results also indicated that policy design can and has
mattered for the adoption of private school choice reforms. Not only did policy design
matter, but it demonstrated a larger marginal effect on bill enactment than the more
152
traditional socioeconomic and political explanations. Proposals that targeted disability
students or placed hard caps on program funding and participation were able to entice
support from lawmakers open to exploring the potential benefits of private school choice
but also resistant to changing long-established policy arrangements. Viewed cynically,
these limiting provisions may simply allow certain elected lawmakers to feign support for
new and innovative education reforms intended to create equity between urban and
suburban schools without actually having to support a proposal that threatens meaningful
policy change. Either way, the analysis demonstrates the ability of policy design to have
an observable influence on bill enactment that is not attached to particular political or
socioeconomic conditions.
The story that emerges from the analysis of individual private school choice
proposals is one of complexity. Private school choice can take many shapes in legislative
proposals, and the dynamic, multifaceted nature of combining policy dimensions to
create different reform impacts defies easy capture. States experienced private school
choice activity in various ways. Many experienced a few bills and ignored them all,
others experienced lots of bills and still chose to ignore them all, while still other states
experienced lots of bills and paid attention to many of them.
Complexity was also spotlighted in the logit estimations, which found that bills
have a half-percent chance of getting passed overall, and that even the most powerful
variable – bills that expand on previously enacted choice laws – was only able to increase
that enactment chance to a little over 3 percent. This underscores the multitude of factors
that can be involved in decisions to advance any bill through the multi-step process of
153
legislative policymaking. It also highlights an important limitation on attempts to explain
legislative success at the altitude necessary to systematically capture broad variation in
both policy design and venue characteristics of individual bills. Analyzing the legislative
success of multiple individual bills across legislative venues may reveal some key general
or non-case-specific explanations for bill enactment, but these factors can only account
for the margins, and the bulk of explanation remains in the idiosyncratic dynamics
between the decision-makers, the process and the content.
Looking forward, an analysis of individual bill success reinforces conventional
wisdom about policy change, and reveals future avenues for studying school choice
activity and policy change.
Private school choice policy change is incremental, path dependent, and complex.
The story presented was one of traditional, incremental policy change. The overwhelming
majority of reform proposals presented were specifically designed to be moderate and
supplemental in their impact on existing public school systems, and the inclusion of the
most clearly-defined and restrictive content provisions was estimated to have the largest
impact on success chances. Big, formal revisions signaling that an important idea’s time
had come were not a prominent feature of the analysis. In this way it connects with other
studies of policy change that have come to similar conclusions, arguing that significant
changes in public policy occur over a decade or more (Sabetier and Jenkins 1993 and
1999), and that legislative dynamics are a complex, shifting, and often personality driven
process that is difficult to capture quantitatively (Kamienieki 2003???). Accordingly, the
struggle for private school choice should be seen as a marathon, not a sprint. Free-market
154
reform advocates may, in their desire for reform, push for larger, less-restrained programs
or shun bills they believe to be too meager. But analysis indicates that when they are
willing to get behind explicitly small, supplemental programs, the chances increase of
gaining a foothold on existing educational policy arrangements that can be subsequently
expanded.
Although the analysis of bill success reinforces conventional wisdom, the
description of legislative activity in state legislative venues confounds private school
choice’s conventional association with conservative politics. The frequency of bill
introduction and bill attention recorded across states revealed some puzzling groupings.
According to the table of legislative activity by state (see Appendix), North Carolina,
Nebraska, Oregon, Kentucky, and Michigan each experienced only five or six bill
introductions, which the legislature largely ignored. That these state containing red and
blue states, more populous urbanized states and less populous more rural states, and
containing states in very different regions of the country. We see a similar situation in
other state groupings. Legislative activity plays out in such a way that states that seem
dissimilar at first blush – Kansas with Massachusetts, South Carolina with Hawaii – are
unexpectedly similar in their amounts of bill introduction and attention.
Thus, a puzzle emerges to be explored. Socioeconomic and political
characteristics of states relate only weakly to their bill activity, unexpectedly grouping
red states with blue states and sparsely-populated Midwestern states with urbanized New
England states, begging the questions: Why do seemingly dissimilar states have similar
legislative activity when it comes to private school choice?; Why are private school
155
choice dynamics only marginally connected with conservative politics given the strong
conventional association of free-market reform and religious accommodation with the
political right?
Future scholarship seeking to understand private school choice dynamics may be
better tailored by selecting cases of state school choice struggles based on their bill
activity as opposed to selecting cases for more in-depth analysis based on their assumed
relationship with conventional socioeconomic and political characteristics. In this way,
selecting particular state cases of private school choice battles to study can be more
deliberate and purposely tailored to explain the puzzling patterns of legislative activity
revealed here. A comparative case analysis of two or more seemingly opposite states –
such as Kansas and Massachusetts – both of which experienced an above average number
of bill introductions and had legislatures choosing to ignore every bill, can more
effectively get at the policy and political dynamics not captured by either this systematic
analysis or previous case studies using case selection methods based on opportunity or
conventional assumptions. Another potentially fruitful exercise would be to select cases
of states that would seem to have favorable characteristics yet have very little legislative
activity to show for it. Conversely, selecting cases of states that would not conventionally
be associated with above average private school choice reform activity yet experienced a
large number of bill introductions – such as Vermont or Connecticut – or even provided
bills a measure of attention – such as Hawaii and Louisiana – should more effectively
explore the way private school choice as a policy issue is or can be distinct from other
issues of conservative policy change such as restrictions on abortion providers.
156
Private school choice is an issue that is more complex then it appears at first
glance. The different ingredients that make up a legislative battle or legislative success
are legion and difficult to capture. This project has identified some of those ingredients as
influencing legislative activity and success on a broader, across state level but it has also
revealed conventional associations of private school choice with conservative, public
education, and urban politics to be partial and problematic. In order to more effectively
explore and understand the unconventional dynamics that can be apart of school choice
politics it is important that case analysis select states and cases that will lend themselves
to addressing the partial and puzzling nature of the issue revealed. This project’s analysis
of individual private school choice legislation was able to highlight details of legislative
activity by state in a way that should better facilitate more deliberate, purposive, and
interesting case selection and analysis on the politics of private school choice.

157
BIBLIOGRAPHY

Barrilleaux, Charles J., and Mark E. Miller. 1988. “The political economy of state
Medicaid policy.” The American Political Science Review 82(4), 1089-1107.

Barrilleaux, Charles, Thomas Holbrook, and Laura Langer. 2002. “Electoral
Competition, Legislative Balance, and the American Welfare State.” American
Journal of Political Science. 46(2), 415-427.

Berry, Francis Stokes, and William D. Berry. 1990. “State Lottery Adoption as Policy
Innovations: An Event History Analysis.” American Political Science Review 84(2),
395-415.

--- 1999. “Innovation and diffusion models in policy research.” In Theories of the
policy process, edited by P. A. Sabatier. Boulder: Westview Press.

Berry, William D., Evan J. Ringquist, Richard C. Fording, and Russell L. Hanson.
1998. “Measuring citizen and government ideology in the American states.”
American Journal of Political Science. 42, 337-346

Blomquist, William. 1999. “The policy process and large-N comparative studies.” In
Theories of the policy process, edited by P. A. Sabatier. Boulder: Westview Press.

Boyd, William Lowe, David N. Plank, and Gary Sykes. 1998. “Teachers' unions in
hard times.” Paper presented at Conference on Teachers Unions and Educational
Reform, Kennedy School of Government, Harvard University, Cambridge, MA.

Brady, David. 1982. “Congressional Party Realignment and Transformations of
Public Policy in Three Realignment Eras.” American Journal of Political Science 26,
333-360.

Brady, David. 1989. Critical Elections and Congressional Policymaking. Stanford
University Press.

Bridges, Tyler, 2002. “Florda Voters Cool to Governor’s A+ Plan, Poll Says.” The
Miami Herald, June 14.

Brunner, E., Sonstelie, J., & Thayer, M. 2001. “Capitalization and the voucher: An
analysis of precinct returns from California's proposition 174.” Journal of Urban
Economics, 50(3), 517-536.

158
Brunner, E., & Sonstelie, J. 2003. “Homeowners, property values, and the political
economy of the school voucher.” Journal of Urban Economics, 54(2), 239-257.

Baumgartner, Frank, and Beth L. Leech. 1999. Basic Interests. Princeton University
Press.

Baumgartner, Frank, and Bryan D. Jones. 1993. Agendas and Instability in American
Politics. University of Chicago Press.

Chubb, John E., Terry M. Moe. 1990. Politics, Markets, and America's Schools,
Washington, D.C.: The Brookings Institution.

Coulson, Andrew. 2004. Forging Consensus: Can the School Choice Community
Come Together on an Explicit Goal and a Plan for Achieving It? Mackinac Center for
Public Policy.

Derthick, Martha, and Paul J. Quirk. 1985. The Politics of Deregulation. Washington
D.C.: Brookings Institution Press.

Dye, Thomas R. 1966. Politics, Economics, and the Public. Chicago: Rand McNally.

Elazar, Daniel. 1984. American federalism: A view from the states. 3rd ed. New York:
Harper and Row.

Friedman, M. 1955. “The role of government in education.” In Economics and the
public interest, ed. R. A. Solo, 123-44. New Brunswick: Rutgers University Press.

--- 1962. “Capitalism and Freedom.” Chicago: University of Chicago Press.

Fusarelli, Lance D. 2003. The Political Dynamics of School Choice. New York, NY:
Palgrave Macmillian.

Ginsberg, Benjamin. 1976. “Elections and Public Policy.” American Political Science
Review 70, 41-49.

Gokcekus, E., Phillips, J. J., & Tower, E. 2004. “School choice: Money, race, and
congressional voting on vouchers.” Public Choice, 119(3-4), 241-254.

Gray, Virginia. 1973. “Innovation in the states: A diffusion study.” The American
Political Science Review 67(4), 1174-85.

Hacker, Jacob S. 1997. The Road to Nowhere: The Genesis of President Clinton’s
Plan for Health Security. Princeton, N.J.: Princeton University Press.

159
Haider-Markel, Donald P. 1998. “The politics of social regulatory policy: State and
federal hate crime policy and implementation effort.” Policy Research Quarterly.
51(1), 69-88.

Henig, Jeffrey. 1994. Rethinking School Choice. Princeton N.J.: Princeton University
Press.

Hill, Kim Q., and Jan E. Leighley. 1992. “The policy consequences of class bias in
state electorates.” American Journal of Political Science. 36, 351-65.

Hofferbert, Richard I. 1966. “Ecological Development and Policy Change in the
American States.” Midwest Journal of Political Science 10:464

--- 1974. The Study of Public Policy. Indianapolis: Bobbs-Merrill.

Howell, William G. and Paul E. Peterson. 2002. The Education Gap. Washington
D.C.: Brookings Institution.

Hsu, Spencer S., 2004, “How Vouchers Came to D.C.”, Education Next vol. 4 no. 4.

Kamieniecki, Sheldon. 2006. Corporate America and Environmental Policy: how
often does business get its way? Stanford University Press.

Kenny, Lawrence. 2005. “The Public Choice of Educational Choice.” Public
Choice.124(1-2), 205-222.

Kingdon, John W. 1985. Agendas, Alternatives, and Public Policies. 2
nd
ed. Boston:
Little, Brown.

--- 1989. Congressmen's voting decisions. Ann Arbor: University of Michigan Press.

Levin, H. M. 1980. Educational vouchers and social policy. In School finance policies
and practices: The 1980s: A decade of conflict, ed. I. W. Guthrie, 235-63. Cambridge,
MA: Ballinger Publishing Company.

--- 1987. “Education as a public and private good.” Journal of Policy Analysis and
Management 6, 628-41.

--- 1990. “The theory of choice applied to education.” In Choice and control in
American education: Vol. 1. The theory of choice and con-trol in education, eds. W.
H. Clune and I. EF Witte, 247-84. New York: Falmer Press.

Lindblom, Charles E. 1959. “The Science of Muddling Through.” Public
Administration Review 19, 79-88.
160

--- 1977. Politics and Markets. New York: Basic Books.

Lowi, Theodore J. 1964. “American Business, Public Policy, Case Studies and
Political Theory.” World Politics. 16, 677-93

--- 1979. The End of Liberalism. 2
nd
ed. New York: Norton.

Maggi, Laura, 2003. “Political landscape tipping toward school vouchers; Proponents
still face significant hurdles,” Times-Picayune, March 23.

Marshall, Janet, 2007. “State expects a rise in voucher offerings.” Sarasota Herald-
Tribune, October 24; “A Choice Showdown

Mawhinney, Hanne B. 1993. “ An Advocacy Coalition Approach to Change in
Canadian Education.” In Policy Change and Learning: An Advocacy Coalition
Approach. edited by Sabatier, Paul A., and Hank C. Jenkins-Smith. Boulder. Colo.:
Westview Press.

Mayhew, D. 1974. Congress: The electoral connection. New Haven: Yale University
Press.

McClosky, Herbert, and John Zaller. 1984. The American ethos: Public attitudes
toward capitalism and democracy. Cambridge, MA: Harvard University Press.

McFarland, Andrew. 2004. Neopluralism. University Press of Kansas.

Miller, Edward A. 2004. “Advancing Comparative State Policy Research: Toward
Conceptual Integration and Methodological Expansion.” State & Local Government
Review. 36(1), 35-58

Mintrom, Michael. 1997. “Policy Entrepreneurs and the Diffusion of Innovation.”
American Journal of Political Science. 41(3), 738-770

--- 2000. Policy Entrepreneurs and School Choice. Washington, DC:
Georgetown University Press.

Moe, Terry M. 2001. Schools, Vouchers, and the American Public. Washington, DC:
Brookings Institution Press.

--- 2005, “Union Power and the Education of Children,” paper presented at the
Teacher Collective Bargaining Conference at the Urban Institute, Washington, D.C.,
May 16-17.

161
Mooney, Christopher Z., and Mei-Sien Lee. 1995. “Legislative Morality in the
American States: The Case of Pre-Roe Abortion Regulation Reform.” American
Journal of Political Science 39(3), 599-627

Morgan, Kathleen O’Leary, and Scott Morgan. 2005. Education State Rankings,
2005-2006. Washington, D.C.: C.Q.Press.

Morgan, Ryan, 2003. “Voucher bill headed for Senate OK Court challenge
predicted,” The Denver Pos.t, March 27.

Morken, Hubert, Jo Renee Formicola. 1999. The Politics of School Choice. New
York: Rowman Littlefield.

Mucciaroni, Gary. 1995. Reversals of Fortune: Public Policy and Private Interests.
Washington D.C.: Brookings Institution Press.

Peterson, Paul E. 1995. The price of federalism. Washington, DC: Brookings
Institution.

Richardson, Valarie, 2003. “Colorado nears voucher approval; Similar bills pass
Senate, House to aid low-income youths,” The Washington Times, February 27.

Sabatier, Paul A., and Hank C. Jenkins-Smith, eds. 1993. Policy Change and
Learning: An Advocacy Coalition Approach. Boulder. Colo.: Westview Press.

Sabatier, Paul A., and Hank C. Jenkins-Smith, eds. 1999. “The Advocacy Coalition
Fremework: An Assessment.” In Paul A. Sabatier, ed., Theories of the Policy
Process. Boulder. Colo.: Westview Press.

Sigelman, Lee, and Roland E. Smith. 1980. “Consumer Legislation in the American
States: An Attempt at Explanation. Social Science Quarterly 61(1), 58-69.

Sigelman, Lee, Phillip W. Roeder, and Carol Sigelman. 1981. “Social service
innovation in the American states: Deinstitutionalization of the mentally retarded.”
Social Science Quarterly 62(3), 503-15.

Sinclair, Barbara Deckard. 1977. “Party Realignment and the Transformation of the
Political Agenda.” American Political Science Review 71, 940-953

Smith, Kevin B. and Kenneth J. Meier. 1995. The Case Against School Choice:
Politics, Markets, and Fools. Armonk, New York: Sharpe

Ulferts, Alisa, 2002. “Voucher Bill Heads to Full House,” St. Petersburg Times,
February 15.
162

Verba, Sydney, and Norman H. Nie. 1972. Participation in America: Political
democracy and social equality. New York: Harper & Row.

Walker, Jack L. 1969. “The Diffusion of Innovations Among the American States.”
American Political Science Review 63(3), 880-899.

Wildavsky, Aaron. 1984. The Politics of the Budgetary Process. 4
th
ed. Boston: Little,
Brown.

Wood, B. Dan, and Nick A. Theobald. 2003. “Political responsiveness and equity in
education finance.” Journal of Politics 65(3), 718-38.

Woodall, Martha, 2002. “A new way to afford private schools: Pennsylvannia
businesses that fund scholarships may seek tax credits. Lower-income students stand
to benefit.” The Philadelphia Inquirer, January 22.

163
APPENDIX A:
LEGISLATIVE ACTIVITY OF SPECIFIC STATES FROM 1997 TO 2007

Table A.1 Individual Bill Activity by State

State # of Bills
Introduced
# of Bills
Reported out
of Committee

# of Bills on
Floor
Scheduling
# of Bills
Passing at least
one Chamber
# of Bills
Signed into
Law
Louisiana 32 4 4 1
Arizona 28 15 14 10 3
Mississippi 26
Missouri 25 7 2
N. Mexico 24 4
N. York 21
Colorado 20 10 7 5 1
Florida 20 15 12 9 3
Indiana 20 5 3 3
Connecticut 18
Texas 18 4 1 1
Vermont 18
N. Jersey 15
S. Carolina 15 2 1 1
Hawaii 14 1 1 1
Kansas 13
Massachusetts 13
Minnesota 13 1 1 1 1
N. Hampshire 13 5 4 4
Utah 13 11 10 6 2
Oklahoma 12
Maine 11
Illinois 10 4 2 2 1
Nevada 10 4 1 1
Virginia 10 1
California 9 1 1 1
Rohde Island 9
West Virginia 9
Georgia 8 1 1 1 1
Ohio 8 3 3 3 3
Arkansas 7
Delaware 7
Iowa 7 3 3 3 2
Maryland 7
Pennsylvania 7 2 2 2 2
Washington 7 1

164
Table A.1 Individual Bill Activity by State (Continued)

N. Carolina 6
Nebraska 6
Oregon 6 1
Kentucky 5
Michigan 5
Tennessee 5 1
Wisconsin 5 2 2 2
Montana 4 1
Alabama 2
Idaho 2 1 1 1
N. Dakota 2
S. Dakota 2 2 2 1
Alaska 1
Wyoming 0
Total 569 117 80 63 19

165
APPENDIX B:
Indicators of Public School Performance
30

Positive Factors Considered (1-13)

1. Public Elementary and Secondary School Revenue per $1,000 Personal Income
2. Percent of Public Elementary and Secondary School Current Expenditures used
for Instruction
3. Percent of Population Graduated from High School
4. Public High School Graduation Rate Percent of Public School Fourth Graders
Proficient or Better in Reading
5. Percent of Public School Eighth Graders Proficient or Better in Reading
6. Percent of Public School Fourth Graders Proficient or Better in Writing
7. Percent of Public School Eighth Graders Proficient or Better in Writing
8. Percent of Public School Fourth Graders Proficient or Better in Mathematics
9. Percent of Public School Eighth Graders Proficient or Better in Mathematics
10. Average Teacher Salary as a Percent of Average Annual Pay of All Workers
11. Average Teacher Salary as a Percent of Average Annual Pay of All Workers
12. Average Daily Attendance as a Percent of Fall Enrollment in Public Elementary
and Secondary Schools in 2004
13. Percent of School-Age Population in Public Schools

30
As published in Morgan and Morgan (2005)
166

Negative Factors Considered (14-21)

14. High School Drop Out Rate
15. Special Education Pupil-Teacher Ratio
16. Percent of Public Elementary and Secondary School Staff Who are School
District Administrators
17. Average Class Size in Public Elementary Schools
18. Average Class Size in Public Secondary Schools
19. Median Pupil-Teacher Ratio in Public Primary Schools
20. Median Pupil-Teacher Ratio in Public Middle Schools
21. Median Pupil-Teacher Ratio in Public High Schools

Abstract (if available)

Abstract The idea of using publicly funded vouchers to subsidize parents’ ability to pick and choose education providers has been part of the contemporary policy debate for decades. Beginning in the 1990s, this private school choice idea began to be translated into specific legislative proposals showing up on the agenda of state legislatures with increasing frequency, with a few states going so far as to enact some form of private school choice law. Yet, scholarship on the issue has remained focused on debating the substance, advocates, and constituencies of the general private school choice idea and little attention has been directed toward understanding the diverse landscape of actual legislation driving attempts to introduce private school choice reform into state education policies.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Income eligible school choice: the effect of parent knowledge on advocacy

PDF

Improving college participation success in Glendale Unified School District: An application of the gap analysis model

PDF

Technical efficiency change in an era of adequacy

PDF

The impact of Connecticut legislators on teacher diversity

PDF

Successful resource allocation in times of fiscal constraint: case studies of school-level resource use in southern California elementary schools

PDF

Personal agency and educational decision-making: insights from middle- and emerging-middle-income school parents in Costa Rica

PDF

Atole con el dedo: learning from low-income parents' lived experiences with educational choice policies

PDF

Essays on the politics of the American welfare state

PDF

Local votes and outside money: campaign contribution geographic origins and their impact on Los Angeles City Council election outcomes

PDF

Oregon education policy implementation: a case study of the achievement compacts information dissemination of the Oregon Education Investment Board

PDF

Multiracial politics or the politics of being multiracial?: Racial theory, civic engagement, and socio-political participation in a contemporary society

PDF

Levels of interest: the effects of teachers' unions on Congress, statehouses, and schools

PDF

The role of judicial decisions on access and equity: a case study of Daniel v. State of California (1999) and the AP Challenge Grant, Senate Bill 1689 (2000)

PDF

Can election reforms improve democracy?

PDF

The impact of principal leadership and model schools on state‐wide school reform: case studies of four Oregon elementary model schools

PDF

The influence of counselors and high school organization on the selection of participants for a dual credit program

PDF

Corporal punishment of students by teachers in elementary and middle schools in Taiwan: the relationship with school level, gender, school location, academic performance, and emotional reactions

PDF

The factors supporting or inhibiting Teachers of Color to accept and stay in an international school in Southeast Asia

PDF

How do consumers use brands for identity signaling: Impact of brand prominence on consumers' choice and social interaction

PDF

A call for transparency and accountability: continuous improvement in state education agency policy implementation

Asset Metadata

Creator Jones, Matthew Daniel (author)

Core Title Private school choice at the state level: diversity and success

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Politics

Degree Conferral Date 2010-12

Publication Date 11/29/2010

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag education reform,legislative success,OAI-PMH Harvest,policy change,school choice,state policy innovation,vouchers

Place Name USA (countries)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Barnes, John E. (committee chair), Picus, Lawrence (committee member), Wong, Janelle S. (committee member)

Creator Email matthedj@usc.edu,mdjones75@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m3564

Unique identifier UC1438999

Identifier etd-Jones-4213 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-420283 (legacy record id),usctheses-m3564 (legacy record id)

Legacy Identifier etd-Jones-4213.pdf

Dmrecord 420283

Document Type Dissertation

Rights Jones, Matthew Daniel

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu