The influence of sector on human resource management

Material Information

The influence of sector on human resource management a study of public, charter, and private schools
Added title page title:
Study of public, charter, and private schools
Williamson, Aimee Lynn ( author )
Publication Date:
Physical Description:
1 electronic file (xiii, 249 leaves) : ;


Subjects / Keywords:
School personnel management -- United States ( lcsh )
Public schools -- United States ( lcsh )
Charter schools -- United States ( lcsh )
Charter schools ( fast )
Public schools ( fast )
School personnel management ( fast )
United States ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


The public education system of the United States is perceived by many as failing to provide an adequate education for the nation's children. Although the current system has supporters and the actual state of American public education is a topic for debate, the public and policymakers from both sides of the political spectrum are demanding improvement in public education. A number of scholars have argued for the institution of market-based school reforms as a means to transform the current system. Charter schools have emerged as a form of school choice reforms that propose to improve the educational system through choice, competition, and autonomy from the bureaucracy of traditional public school districts. The implications of these charter schools are not yet fully understood. This mixed-method study analyzes human resource management practices across the public, private, and charter school sectors, uncovering differences, particularly in the areas of formalization and autonomy. In support of other research on public-private management differences, findings suggest that public schools are more formalized and have less autonomy than do public and charter schools.
Thesis (Ph.D.)--University of Colorado Denver, 2008.
Includes bibliographical references (leaves 242-249).
Statement of Responsibility:
by Aimee Lynn Williamson.

Record Information

Source Institution:
University of Colorado Denver Collections
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
988950379 ( OCLC )

Full Text
Aimee Lynn Williamson B.A., University of Notre Dame, 1993 M.P.A., University of Colorado at Colorado Springs, 1995
A thesis submitted to the University of Colorado Denver in partial fulfillment of the requirements for the degree of Doctor of Philosophy Public Administration

by Aimee Lynn Williamson All rights reserved.

This thesis for the Doctor of Philosophy degree by
Aimee Lynn Williamson has been approved by
Paul E. Teske
Alex Medler

Williamson, Aimee Lynn (Ph.D. Public Administration)
The Influence of Sector on Human Resource Management: A Study of Public, Charter, and Private Schools
Thesis directed by Professor Paul Teske
The public education system of the United States is perceived by many as failing to provide an adequate education for the nations children. Although the current system has supporters and the actual state of American public education is a topic for debate, the public and policymakers from both sides of the political spectrum are demanding improvement in public education. A number of scholars have argued for the institution of market-based school reforms as a means to transform the current system. Charter schools have emerged as a form of school choice reforms that propose to improve the educational system through choice, competition, and autonomy from the bureaucracy of traditional public school districts. The implications of these charter schools are not yet fully understood. This mixed-method study analyzes human resource management practices across the public, private, and charter school sectors, uncovering differences, particularly in the areas of formalization and autonomy. In support of other research on public-private management differences, findings suggest that public schools are more formalized and have less autonomy than do public and charter schools.
This abstract accurately represents the content of the candidates thesis. I recommend its publication.
Paul E. Teske

I dedicate this dissertation to my family. For Rob Williamson, a true public servant, whom Im proud to call my husband, and for our children, Colton and Helena, who
are our proudest accomplishments.

I wish to express my sincere appreciation to all the people who have provided assistance and support that made this dissertation possible. Between the faculty, administration, and students at the School of Public Affairs, as well as family and friends, there are many to whom Im grateful, but in the interest of brevity, Ill only note a few below.
My sincere thanks go to my committee chair, Paul Teske, whose support has extended far beyond this dissertation. He has provided me with opportunities that I never expected during graduate school and for that, I will always be grateful. I would also like to thank the rest of my committee, Jody Fitzpatrick and Linda deLeon of the School of Public Affairs and Alex Medler of the Colorado Childrens Campaign; all of whom have also had a larger impact on me than their assistance with this dissertation. In addition to Jody Fitzpatricks clear and constructive feedback, the research experience I gained working for her provided an excellent foundation for this project. Linda deLeons expertise on public management, excellent teaching skills, and dedication to students is sincerely appreciated. I also thank Alex Medler for his vast knowledge on charter schools and his continuous support, both on this dissertation and in the workplace.
I thank my department chair, Michael Lavin, as well as other members of the faculty and administration at Suffolk University, for their willingness to hire me ABD and their patience and support while I finished this dissertation.
I would also like to thank Antoinette Sandoval of the School of Public Affairs, who knows the answer to every question and seems to solve every problem.
Most importantly, I thank my family, including my husband Rob, my children Colton and Helena, and my parents Suzanne and Alec Vezina. All of them have made sacrifices for my education and their support has never waivered.

1. INTRODUCTION....................................................1
The State of Public Education...............................1
School Choice...............................................4
Market-Based Reform and Charter Schools................6
General Differences Between Public and Private Schools.....10
Problem Statement..........................................11
Research Purpose and Research Questions....................14
2. REVIEW OF THE LITERATURE.......................................16
School Choice..............................................16
School Choice Origins.................................17
A Proposal for Charter Schools........................19
The Public v. Private Management Debate....................27
Schools of Thought.................................. 27
Hybrid Organizations..................................30
Empirical Studies.....................................31

Empirical Research on Human Resource Management in Schools.35
Public Choice Theory........................................44
Literature Review Conclusion................................47
3. METHODOLOGY.....................................................50
Research Purpose........................................... 50
Study Design............................................... 51
Sampling for States....................................52
Sampling for Subjects..................................62
Survey Protocol........................................64
Interview Protocol.....................................65
Chi-Square Analyses....................................66
Independent t tests....................................67
Multiple Regression....................................68
Cronbachs Alpha.......................................69
Problems and Limitations....................................70
4. RESULTS.........................................................74
Descriptive Statistics for the Survey Sample................74

Descriptive Statistics for the Qualitative Sample............80
Quantitative Results.........................................82
Additional Findings of Interest........................120
Qualitative Results........................................ 123
Contribution to Theory: Public-Private Differences in Formalization, Autonomy, and Time...........................154
Charter Principals as More Likely to Decentralize
Decision Making........................................171
Contribution to Theory: Public v. Private Management Debate.176
Contribution to Theory: Charters as a Means to Inject
Market Forces...............................................180

Contribution to Theory: Fairness v. Effectiveness Tension..182
Areas for Future Research..................................184
Role of Additional Groups on Decision Making..........184
Analyzing Time........................................186
Core v. Dimensional Approach to Public-Private Management Debate.....................................186
Effect of No Child Left Behind on Teacher Certification Requirements..........................................189
Effect of Age of School...............................190
Desire for Formalization and Autonomy.................193
Challenges in Defining Formalization and Autonomy.....195
Union and Professional Group Constraints..............196
Hiring Challenges and Application Pools...............198
Limitations and Challenges.................................200
Weak Statistical Power................................202
C. HUMAN SUBJECTS APPROVAL................................240

3.1 Sampling overview......................................................54
4.1 Sample descriptive statistics by state.................................76
4.2 Sample descriptive statistics by sector................................77
4.3 Descriptive statistics for the qualitative sample......................81
4.4 Regression analysis summary for variables predicting summated
formalization scale 1................................................ 86
4.5 Desire to adapt evaluation policies across school sectors..............88
4.6 Liberty to diverge from personnel policies across school sectors.......89
4.7 Written policies for evaluation across school sectors..................90
4.8 Rules and procedures across school sectors.............................92
4.9 Tenure across school sectors...........................................93
4.10 Salary schedule across school sector...................................94
4.11 Summary of chi-square analyses for Hi..................................96
4.12 Regression analysis summary for variables predicting
summated autonomy scale...............................................100
4.13 Evaluation autonomy across school sectors.............................101

4.14 Union and professional group constraints across school sectors..........102
4.15 Strong traditions across school sectors...................................104
4.16 District staff hiring influence across school sectors ................... 107
4.17 District staff dismissal influence across school sectors..................107
4.18 Evaluation influence of teachers across school sectors....................109
4.19 Dismissal influence of teachers across school sectors.................... 109
4.20 Dismissal influence of parent association across school sectors..........110
4.21 Hiring influence of curriculum specialists across school sectors..........Ill
4.22 Summary of discussed chi-square analyses for H2...........................115
4.23 Regression analysis summary for variables predicting weeks to
dismiss an ineffective teacher........................................... 117
4.24 Regression analysis for variables predicting weeks to fill a teachers
4.25 Dismissal influence of others across school sectors.......................121
4.26 Ineffective teacher result across school sectors..........................122
4.27 Summary of quantitative and qualitative findings related to Hi........134
4.28 Summary of quantitative and qualitative findings related to H2........145
4.29 Summary of quantitative and qualitative findings related to H3........150
A. 1 Summary of regression results.........................................206
A.3 Summary of chi-square analyses for Hi: Personnel policies are more
formalized in more public schools.......................................207

A.4 Summary of chi-square analyses for H2: Principals in more public schools have less autonomy than principals in more private schools in terms of personnel decisions................................... 208
A.5 Summary of chi-square analyses for H3: From the time where a
need to hire or fire personnel is recognized, such personnel decisions take longer in more public schools........................212
A.6 Summary oft tests....................................................213

INTRODUCTION The State of Public Education
The public education system in America has been a central topic on the countrys political agenda for the past two decades. The public, the media, and politicians from both parties agree that the public education system is in great need of improvement. The means to greater achievement, however, is the source of extensive argument across America (Schneider, Teske, & Marschall, 2000). There are simultaneous calls for both decentralization, including reforms such as charter schools, vouchers, and site-based management; and greater centralization, reflected by standardized testing, national legislation (e.g., No Child Left Behind), and proposals for national standards.
Both sides of the debate cite evidence, such as testing results and graduation statistics, that the American public school system is failing to educate the nations children effectively. The problem is even more pronounced for minority students and those living in poverty (Ladd, 2002). As Moe (2001b, p. xv), argued: The evidence is plain that many urban school districts are in crisis, often failing to graduate even half of their students, and turning out graduates who in many cases can barely read, write, or do basic arithmetic.

Following the publication in 1983 of the Department of Educations A Nation at Risk, states engaged in rapid reforms in the attempt to improve the public education system:
Their piecemeal efforts coalesced into a forceful national movement for academic excellence, and the reforms that followed were far more extensive in breadth arid magnitude than anything the nation had see since the early years of Progressive institution-building (Chubb &
Moe, 1990, p 10).
Calls for educational reform, variously defined, have continued over the past twenty-five years and scholars argue that the American education system remains in a state of crisis (Lips, 2008).
Education reform has been a major issue for most of the last century (Ravitch, 1995), but gained greater prominence around the turn of the century: Education is currently at the forefront of the nations political agenda: everyone, regardless of political persuasion, wants to see an improvement in the performance of U.S. schools (Figlio, 2001, p. 7). In recent years, other issues, such as the economy, the war in Iraq, and energy costs, have eclipsed education as top concerns on the national policy agenda; but education remains an important topic for politicians, scholars, and the general public. In the current presidential race, both major presidential candidates emphasized the importance of education in their convention speeches, with Senator Barack Obama commenting on our moral obligation to provide every child a world-class education (Barack Obama, Illinois, 2008) and Senator John McCain asserting a recurring theme that education is the civil rights issue of this century (Full text:

Remarks by John McCain, 2008). In the academic literature, opposing sides continue to debate the need for widespread and comprehensive reform in terms of school management and education policy, but not the importance of education. Despite the ascension of other policy issues, such as the economy, education remains an important issue for many Americans. A CNN/Opinion Research Corporation Poll from the summer of 2008 found that 44% of registered voters reported education as extremely important in their vote for president, and an additional 39% reported education as very important, representing a total of four out of five registered voters {Problems and Priorities, 2008).
Despite various reform efforts over the last twenty years and relatively large expenditures for education, the American educational system has struggled to increase the achievement of students. An examination of standardized test scores over time indicates that in 1996, reading scores were no higher than in 1984 and math scores were no higher than in 1988. A 2008 report, however, points to gains from 2002-2007 in reading and math scores, as well as a narrowing of achievement gaps for African American and low-income students (Kober, Chudowsky, & Chudowsky, 2008). The United States still ranks low among OECD nations in graduation rates, however, despite having the third-highest expenditures among economically advanced countries (Walberg, 2001, p. 67). One out of four of our freshmen students did not graduate from high school with their peers in 2006. The situation is even more critical within certain subgroups of the populationthat same year, two

out of five Hispanic and black students failed to graduate with their peers (Stillwell & Hoffman, 2008). Critics of reliance on within system change argue that previous reforms have been ineffective and should be replaced with more significant change mechanisms: The modem history of American education reform is a history of dashed hopesand continuing demands, as a result, for real reforms that will bring significant improvements (Moe, 2001b, p. xvii).
Effecting education reform has continued to be a priority for recent presidential administrations, including those of George H.W. Bush, Bill Clinton, and George W. Bush. The most recent wave of federally directed education reform, under the presidency of George W. Bush, includes the passage of the No Child Left Behind Act, which focuses on accountability in public schools, particularly through the establishment of widespread standardized testing requirements and provisions to institute market-like reforms in failing schools. There has been talk of reauthorization of this legislation over the past couple of years, but there is uncertainty as to whether this will occur in 2008.
School Choice
School choice, broadly construed, refers to a range of reform efforts which decentralize educational decision-making from higher levels of bureaucracy to parents (Schneider et al., 2000). School choice comes in a variety of diverse forms. Although school choice is often presented as a new and innovative reform in education, various forms of school choice are relatively widespread, uncontroversial,

and long-standing elements of the public school system (Schneider et al., 2000). Examples include open enrollment; magnet schools; and the ability of parents, at least those who are financially able, to choose a particular school or district by moving within its boundaries (Peterson, 2001).
In present day policy terminology, however, school choice typically refers to contemporary market-based reforms such as voucher programs and charter schools. Voucher programs, which are one of the most controversial contemporary reforms, allow students to use the money allocated for their public education to pay for private (either secular or religious) education (Schneider et al., 2000). Levin defined vouchers as a system of public educational finance in which parents are given a tuition certificate by the government that can be used to pay tuition at any approved school, public or private (Levin, 2002, 159). Voucher programs can take various forms, depending upon the structure of the program. For example, policy decisions include whether to include religious schools, the amount of the voucher, eligibility requirements (e.g. low-income families, failing schools), and accountability standards.
Charter schools, which are the focus of this study, are publicly funded schools, operating under a charter from an authorizing entity or board. Levin defined charter schools as public schools that receive a specified sum of funding for each student from their local school district or state and are released from compliance with many local and state regulations providing that they adhere to their declared mission

or charter (Levin, 2002, 159). There are a wide variety of charter schools, but the common characteristic is greater autonomy from the public school bureaucracy. Unlike private schools, charter schools are accountable to a charter granting entity, but have far greater discretion in the operation of the school than traditional public schools.
Market-based Reform and Charter Schools
Contemporary market-based education reforms, such as vouchers and charter
schools, originated in the writings of Milton Friedman, who proposed the concept of
educational vouchers as an application of economic theory to reform the public
education system. Although Friedmans proposal didnt gain widespread attention
until he reiterated it in Capitalism and Freedom (1962), he first described the concept
in 1955, in a book chapter entitled The Role of Government in Education:
Government, preferably local governmental units, would give each child, through his parents, a specified sum to be used solely in paying for his general education; the parents would be free to spend this sum at a school of their own choice, provided it met certain minimum standards laid down by the appropriate governmental unit. Such schools would be conducted under a variety of auspices: by private enterprises operated for profit, non-profit institutions established by private endowment, religious bodies, and some even by governmental units (1955).
Friedmans argument was based on the perception that the public education system was monopolistic and, thus, inescapably inefficient. His proposed voucher system sought to instill competition and choice in public education as a means to

correct the inefficiencies and lack of innovation resulting from a publicly-run education system:
They would bring a healthy increase in the variety of educational institutions available and in competition among them. Private initiative and enterprise would quicken the pace of progress in this area as it has in so many others. Government would serve its proper function of improving the operation of the invisible hand without substituting the dead hand of bureaucracy (1955, p. 144).
Although the charter school movement moves away from Friedmans original recommendation that the government only provide subsidies, rather than act as a supplier of education, the concepts of competition and choice that were central to his argument are also at the foundation of the charter school movement. In fact, critic Jeffrey Henig (1994) contended that Chubb and Moes charter school proposal, to be discussed at length in chapter 2, was primarily an elaboration of the Friedman model, (p. 89) with strategically refurbished terminology (p. 87) as a means to avoid the stigma of vouchers. At any rate, Friedman has continued to articulate his proposal and support the voucher movement throughout the years (Friedman, 1995, 1997, 1999; Friedman & Friedman, 1984).
The application of market-based reforms to the public sector has been a critical issue in public affairs for most of the past decade. As noted above, the education sector has not been exempt from such calls for market-based reform. Recent movements in education reform include calls to institute market mechanisms

into public education. Charter schools, which originated in practice in Minnesota in 1992 (Schroeder, 2004), are an example of market-based reforms in education.
Charter schools have become widespread as a means through which to instill choice and competition in the public education system. As defined by Peterson and Campbell (2001, p. 6), charter schools are those schools granted a charter by a state agency giving them the right to receive state funds in exchange for commitments contained in the charter. In other words, charter schools are publicly-funded schools that may be managed by non-public entities such as private companies or non-profit organizations. These schools compete for students to ensure their survival. If they are unable to market their services to draw in students, they will eventually cease operation because funding is directly tied to enrollment.
For proponents motivated by instilling market mechanisms into public education, the conceptual origin of the charter school movement is attributed to John E. Chubb and Terry M. Moes influential and controversial book Politics, Markets, and America's Schools (1990).1 This book presented the findings from research examining the characteristics of effective versus ineffective schools. Its fundamental argument was that the primary problem with public education in the U.S. is institutionalthe bureaucratic organization of public schools impedes the ability of schools to organize effectively.
1 For alternative perspectives on the origin of charter schools, see Rofes and Stulberg (2004).

Chubb and Moe (1990) identified three of the most significant findings from their study: (1) school organization influences school effectiveness, (2) autonomy contributes to school effectiveness, and (3) the current public education system lacks the characteristics and autonomy to encourage the development of effective schools. Theoretically, the authors argued that the fundamental differences between public schools and private schools from this institutional perspective are that public schools are too heavily bureaucratic and that politics fosters this degree of bureaucracy, whereas private schools are controlled by the market rather than politics (p 26). Chubb and Moes analysis of findings included the assertion that students in effectively organized schools (which are more likely to be private schools) achieve more than a full year more of academics than students in ineffectively organized schools, with all other factors (including socio-economic status) held constant.
In the eighteen years following the publication of Politics, Markets, and Americas Schools, over 4,000 charter schools have emerged in 40 states and the District of Columbia, educating over 1.24 million students (Allen & Consoletti, 2008). Charter schools have become a part of the public education landscape and have generated political support from both parties. Presidents and presidential candidates from both parties over the past eight years have demonstrated support for charter schools, including former President Bill Clinton and President George W. Bush (Manno, 2000; Peterson & Campbell, 2001). The significance of the charter school movement is recognized by scholars:

In a short time, these independent public schools of choice, freed from rules yet accountable for results, have spread like wildfire across much of the land, providing schooling alternatives for hundreds of thousands of families and challenging some basic assumptions about public education (Finn, Manno, & Vanourek, 2001, p. 19).
Despite relatively strong support for the charter school movement, the
implications for such reforms are not yet fully understood. Questions remain about
the effectiveness of charter schools in generating the administrative and curricular
innovations, and subsequent improvements in achievement, that charters claim to
generate (Lubienski, 2003).
The debate surrounding charter schools, as well as other market-based education reforms, has roots in critical issues in public affairs, including the public/private management distinction debate and public choice theory. This proposal explains each of these theoretical concepts and discusses how they relate to school choice.
General Differences between Public and Private Schools By definition, the primary distinction between private and public schools concerns whether they are owned, funded, and/or operated by governments or private companies. Alt and Peter (2002, p. 1) explained the defining characteristics of private schools: Private schools are owned and governed by entities that are independent of any governmenttypically, religious bodies or independent boards of trustees. Private schools also receive funding primarily from nonpublic sources... Public schools, on the other hand, are governed by state and local education agencies

(districts) and publicly elected or appointed school boards and such schools receive nearly all their funding from local, state, and federal governments (Alt & Peter,
2002, p. 1). The authors also explained the role of choice in private schools as an important characteristicparents choose private schools for their children and private schools have some choice in which students attend their school.
Recent reforms, such as charter schools, blend the above distinctions between public and private schools, resulting in a kind of hybrid organization, as charter schools have been recognized in the literature (Manno, 2000; Vergari, 2007). The boundaries for defining a public school become blurred as non-profit or even for-profit schools receive public funding in the form of charters or vouchers. As Lubienski (2003, p. 404) explained, charter schools are public-private hybrids, which occupy an area between the different institutional roles (and consequent structural characteristics) typically played by public and private schools. Charter schools are funded by government, but may be managed by private firms who contract with the authorizing body. The implication of this hybrid form of public-private educational provision is one of the central issues of this dissertation.
Problem Statement
The problem is two-fold: 1) The implications and effectiveness of the past fifteen years market-based education reforms, specifically charter schools, are not clearly understood, and 2) There is an ongoing debate as to the differences, if any, between public and private management. The issue becomes even more timely and

important with the implementation of the No Child Left Behind Act (2002), which requires a highly qualified teacher in every classroom, defined as a teacher having at least a bachelors degree, state certification, and able to prove they know the subject(s) they teach. This legislation extends to charter schools, as well as traditional public schools, and likely imposes constraints on their autonomy. Influences such as these further muddy the distinction between public and private in the world of educationan issue that the next chapter addresses at more length.
The charter school movement, which attempts to inject private forces into public education, implies that (1) there are management differences between public schools and private schools, and (2) private schools (and perhaps public schools in competition with each other) are more effective than public schools in terms of the educational achievement of students. Reform movements, such as charter schools, are sweeping the public education system, without a clear understanding of the implications for management. This research examines the implications of market-based reforms of education for school management.
Some critics of market reforms, generally speaking, argue that public and private organizations are too different for the appropriate transfer of best practices from one sector to the other. Boyne (2002, p. 118), summarized the argument of such critics that management techniques cannot be exported successfully from one sector to another because of differences in organizational environments, goals, structures, and managerial values. He suggested, however, that there is relatively limited

empirical support for this argument, and that there are few solid empirical grounds for rejecting the application of successful private practices to public organizations
(P- H8).
This debate surrounding public-private management distinctions had emerged by the 1970s (Rainey, Backoff, & Levine, 1976), but little consensus has developed over the past thirty years. Scholars continue to argue whether management is generic, whether there are fundamental differences between public and private management that preclude generic organizational theory, and whether those differences are based on ownership or other factors. This research contributes to this ongoing body of theory, an area with recognized significance for a number of issues, including the following:
privatization of public services; allocation of functions and tasks among sectors; the nature of the sectors themselves; the dimensions that define the sectors, including their complex overlapping and blurring with the third and nonprofit sectors; administrative reforms and organizational change; and the theoretical and practical analysis of major administrative topics, such as organizational goals, structure, and individual motivation and work attitudes (Rainey & Bozeman,
2000, p. 2).
These are all considerations in designing and implementing educational reform policies.
Despite the recognized significance of this debate, there has been relatively little empirical research addressing the issue, both generally and specifically in terms of educational policy:

In short, the academic community has not taken seriously the need to evaluate public management reforms. Most work in this field has analyzed the economic, political, and administrative antecedents of reform. By contrast, the consequences of reform for the performance of public organizations have been neglected (Boyne, Farrell, Law, Powell, & Walker, 2003, p. 2).
In terms of educational reform, one area which remains relatively unexplored
is the difference in human resource management between traditional public,
charter, and private schools. One of the stated goals of the charter school movement
is to grant more autonomy to principals, including granting more flexibility in
personnel issues. Public schools are generally thought to be relatively bound in
personnel decisions by the formalization of procedures, including an emphasis on
seniority and the influence of unions. In contrast, private schools are thought to have
more autonomy in these decisions, resulting in a greater ability to make personnel
decisions in line with the mission and needs of the school. These characterizations
deserve further examination, to assess the differences between public and private
schools, the impact of charter schools, and the correlation between such practices and
student achievement.
Research Purpose and Research Questions The primary purposes of this research are to examine differences between public and private management with regard to schools and to examine the implications of charter schools on school management. The research question is as follows: Do differences in human resource management exist between public, private, and hybrid (charter) schools?

This study has implications for both public management and education policy. In terms of public management, the study contributes to the existing body of literature by identifying differences between public and private organizations in terms of autonomy and formalization. This study also addresses the public-private distinction debate by providing support for the core approach and, to some extent, to the dimensional approach to the public-private management debate. In terms of education policy, this study provides empirical evidence of the implications of charters on school organization, suggesting that charters do demonstrate the autonomy they are generally designed to achieve, specifically with regard to human resource management practices.

School choice, public-private distinctions, and public choice theory are all relevant conceptual frameworks with regard to this study. As this study focuses on the anti-bureaucratic rationale for charter schools, the school choice proposal of charter schools, articulated by Chubb and Moe (1990), and the research that forms their justification for such a proposal, are central to this study. In a sense, their research played the role of pretest for the present study, as explained in more detail below. Public management literature, specifically the public-private organization debate and public choice theory, contributes to the study as well, serving as conceptual frameworks.
School Choice
School choice involves a broad range of reforms, designed to provide parents with new options for their childrens schooling and to decentralize decision making. The history, including the various motivations for charter advocacy, as well as the theoretical framework underlying school choice are important bodies of literature for this study.

School Choice Origins
The difficulty in determining the starting point of school choice history is noted in the literature, as components of the movement go back to the 1700s, with the works of Thomas Paine and the Founding Fathers (Coons & Sugarman, 1978; Stulberg, 2004). Despite these early discussions, contemporary school choice discussions date back to the 1950s, where school choice programs have roots in the work of economist Milton Friedman. Friedman argued that the government could pay for education through vouchers without operating the schools, thereby giving more choice to parents, instituting competition, and resulting in a more efficient means of providing for education (1955; 1962).
In addition to roots in free-market, conservative ideology, there have been parallel movements advocating charter schools with alternative motivations. Conservatives were not the only ones who took notice of Friedmans proposal liberal academic Christopher Jencks advocated vouchers for low-income students as a means to advance equity in the 1970s (Marron, 2002; Stulberg, 2004). Also in the 1970s, as a precursor to todays charter schools, Ray Budde proposed a concept providing public school teachers with autonomy to run an existing public school for a period of years without significant intrusion from administration, an idea that gained prominence when supported by A1 Shanker of the American Federation of Teachers in the late 1980s (Medler, 2004). Shankers conception of a charter school, however, was quite different from what is typically in practice today. He envisioned schools

where teachers were empowered to create their own programs to foster innovation with union approval (Malin & Kerchner, 2007). Very few charter schools today are unionized, but many advocates share Shankers focus on teachers autonomy as a driver for the charter movement and progressive policy leaders have played a significant role in the emergence and subsequent expansion of the charter school movement. In fact, the first state charter law, passed in Minnesota in 1991, was the result of work by a coalition of non-partisan, progressive, and Democratic policy leaders and analysts, such as Ted Kolderie, Joe Nathan, and State Senator Ember Reichgott-Junge (Medler, 2004).
School choice remains a complex movement, comprised of proponents with varying and sometimes overlapping motivations for instilling choice in the American education system. Todays charter advocates continue to reflect a diverse and unexpected mix of motives and theories, including supporters from both sides of the political spectrum (Vergari, 2007). As evidenced above, school choice proponents have different motivations for supporting school choice, resulting in diverse coalitions of supporters. School choice supporters include individuals with a social justice focus, seeking to improve conditions for underserved populations such as minority and low-income children; those opposed to the bureaucracy of public education; members of the establishment seeking autonomy to pursue a particular curriculum or more freedom in the classroom or looking for an alternative to

accountability through standardized testing; and individuals seeking to further religious education (Medler, 2004; Moe, 2001a).
This study focuses on the rationale for charter schools arguing against the bureaucratic structure of American public education and examines some of the claims of this subset of the charter movement, which includes the work of John Chubb and Terry Moe.
A Proposal for Charter Schools
In their groundbreaking work, Politics, Markets, and Americas Schools, Chubb and Moe (1990) examined four areas of schools: personnel, goals, leadership, and practice. Their research involved two data sets, the High School and Beyond survey and the Administrator and Teacher Survey, both in the 1980s. The first provided detailed information on students, including student achievement and student background. The second provided information on school organization. The authors ran various regression analyses on the data to establish examining relationships between student performance and school organization, and between school organization, bureaucratic constraints, and school sectors. Based on these analyses, Chubb and Moe (1990, p. 99) argued, first that effective schools had goals that were clearer and more academically ambitious, principals who were stronger educational leaders, and teachers who were more professional and harmonious, coursework that was more academically rigorous, and classrooms that were more orderly and less bureaucratic. Second, they argued that these characteristics were

more likely to be found in private schools. In sum, Chubb and Moe suggested that the institutional setting of a school influenced significant dimensions of a schools organization:
Democratic control tends to promote bureaucracy, markets tend to promote autonomy, and the basic dimensions of school organization personnel, goals, leadership, and practicetend to differ in ways that reflect (and support) each sectors disposition toward bureaucracy or autonomy (Chubb & Moe, 1990, 61).
This study focuses on the personnel aspects of Chubb and Moes study, which the authors emphasized as a particularly significant area in affecting organizational effectiveness: Personnel constraint may well be the critical determinant of whether public school organization is affected adversely by bureaucracy (1990, p. 277). In terms of their analysis of personnel policies, the authors found that public principals were more constrained in terms of personnel decisions as a result of statutes and requirements as well as the unionization of teachers. Private school principals had more discretion in personnel practices such as the hiring and firing of teachers. In other words, personnel decisions were more formalized and less autonomous in public schools than in private schools. A related argument in their work was the concern that public school principals were typically denied the discretion for effective leadership; generally, private principals were not.
As explained above, the authors found more levels of bureaucratic constraint, including in the area of personnel, in public schools than in private schools. In addition, they found significant differences in organizational

effectiveness based on levels of bureaucratic constraint: A change in administrative and personnel constraint from their top to their bottom quartiles is associated with an increase in organizational effectiveness of 68.4 percentiles (p. 277). Their argument suggested that the bureaucratic constraints of personnel policy in public schools resulted in less organizational effectiveness than the more autonomous personnel policies that are more characteristic of private schools.
In examining differences in personnel constraints between public and
private schools, Chubb and Moe acknowledged the influence of other factors,
such as the possibility that lower achieving schools result in more personnel
constraint, rather than the other way around. They found, however, that even
taking other factors into account, higher levels of personnel constraint resulted
in lower levels of effectiveness:
There is no ambiguity, moreover, about the more basic influence of institutions. When all else is equalwhen schools are serving the same kinds of students and dealing with the same kinds of families, when schools are situated in the same locations, including urban locations where large educational institutions are conducive to bureaucratizationschools in the private sector are likely to experience far less administrative and personnel constraint than schools in the public sector (Chubb & Moe, 1990, p. 180).
Charter schools have since emerged as an alternative to the bureaucracy of public schools. Their proponents have suggested a number of benefits of instituting this market-based reform into education, including the creation of schools that are

more autonomous, flexible, innovative, and effective than public schools, as well as being more accountable due to their dependence on the choice of parents/students and the need to uphold performance contracts with funding agencies (Bulkley & Fisler, 2002).
Chubb and Moes analysis of the differences between public and private schools and their subsequent call for the development of charter schools as a means to cure the ills of public schools has generated a contentious debate surrounding the appropriateness of their proposed reforms. Two of the forceful critics of Chubb and Moes proposal are Kevin B. Smith and Kenneth J. Meier, who have argued strongly against school choice (Smith, 1994; Smith & Meier, 1994, 1995a, 1995b). In direct opposition to the claims of Chubb and Moe, Smith and Meier proclaimed that school choice was an ineffective and inappropriate means of school reform: We conclude that the theory providing the rationale for school choice simply does not fit with reality. Instead of being a panacea, school choice has the potential to affect adversely the overall performance of Americas education system (1995a, p. xiii). Critics of school choice, generally, and charter schools, specifically, have leveled a number of arguments against such programs. This thesis focuses on the ones most relevant to the current research.
The argument against school choice most directly related to this proposal is the criticism leveled against Chubb and Moe for their negative portrayal of bureaucracy. Such critics point to the importance and effectiveness of government

and bureaucracy for solving complex problems in society (Henig, 1994; Smith & Meier, 1995a). For example, Smith and Meier acknowledged that bureaucracy often has a legitimate and useful function (p. 129). Bureaucracy is effective at managing the diverse problems and issues that emerge in public school environments. As a result, Smith and Meier argued for policies resulting in greater uniformity and centralization, such as increased standardized testing, a national curriculum, graduation tests, and teacher certification, as opposed to market reform.
There has been some empirical investigation of this debate over the relationship between bureaucracy and school effectiveness in recent years. For example, John Bohte (2001) examined data from school districts in Texas and found that higher percentages of administrative personnel (both in terms of school administrators and central administrators) resulted in lower student scores on Texas Assessment of Academic Skills (TAAS) and SAT exams, a finding that lends some support to Chubb and Moes argument that bureaucracy negatively impacts organizational effectiveness.
On the other hand, Kevin B. Smith (1994) examined the impact of school organization and competition on student performance in Florida school districts, with results that generally did not support the school choice argument that democratic control and bureaucracy are conversely related to academic performance. Democratic control, measured as whether the superintendent was elected (or appointed), did not have a statistically significant negative relationship with performance. Also, although

initial analyses did indicate a negative relationship between bureaucracy and performance, secondary analyses suggested that the existence of large bureaucracies in school systems was need-driven rather than a problem resulting from democratic control or the organizational strength of teachers (p. 489). Furthermore, the author suggested that his findings provided some support for the existence of the creaming effect, suggesting that charter schools draw already successful students from the public school system. There is ample room for criticism of this study, however, particularly the choice of an elected or appointed superintendent as the measure of democracy. Even Smith (with Larimer, 2004) himself later criticized the common practice of operationalizing multidimensional concepts with one dimension. Nonetheless, Smith represents part of the cadre of scholars arguing in support of the bureaucratic system of public schools.
Smith and Larimers (2004) study provided additional insight into these issues. The authors examined bureaucracy and school performance, finding a negative relationship between bureaucracy and school performance in terms of test scores, but a positive, in the sense of beneficial, relationship between bureaucracy and attendance rates, as well as dropout rates. Bureaucracy had a positive correlation with attendance rates and a negative correlation with dropout rates.
A second related criticism against school choice is the question as to whether school choice reforms, specifically charter schools, lead to the autonomy envisioned by its founders. A study comparing charter schools founded by or with for-profit

educational management organizations (EMOs) with all other charter schools, found that the market oriented charter schoolsthose founded or co-founded by a for-profit EMOexhibited significantly less school-level decision making in the areas of curriculum, testing and standards, student discipline, facilities, and general administration (Brown, Henig, Lacireno-Paquet, & Holyoke, 2004, p. 1044). In other words, do charter schools lead to increased autonomy for principals or is the bureaucracy of the public school system replaced by that of an educational management organization? On the other side of the argument, however, such a finding could reflect an EMOs desire and ability to replicate a proven school design.
Another argument against school choice is the suggestion that the claimed failures of the public school systems are greatly exaggeratedthat American schools are keeping up with other comparable countries, that SAT scores are increasing, and that drop out rates may be misleadingly high due to increases in enrolled immigrants who are not adequately prepared for the demands of the education system (Smith & Meier, 1995a).
Other critics of school choice note the downside of parental involvement in the school choice process. Such criticisms include the argument that some parents are incapable of making sound choices for their students (Themstrom, 1990) or the argument that parents base their decisions (or make demands of their schools) that are based in issues besides educational quality, such as a desire for religious education or racial segregation (Smith & Meier, 1995b; Wrinkle, Stewart, & Polinard, 1999).

One of the other major criticisms of school choice that is beyond the scope of this thesis, but important to note, is the potential for school choice programs, including charter schools, to adversely affect equity (Henig, 1994). The equity concern is also commonly called a creaming effect (Smith, 1994). Some critics claim that charter schools will pull the students who are already successful out of traditional public schools, leaving such schools with a less conducive environment for education. This argument rests partly on the premise that those parents who are already involved in their childrens education are the ones most likely to seek out innovative charter schools and transfer their children to such schools. In addition, it is in the best interest of charter schools to accept students who are easier to educate since their survival is dependent upon their success. As Smith and Meier (1995a, p. 131) explained: The market seeks efficiency. Equity is not efficient. Research has produced mixed findings on this hypothesis, with Smith and Meier (1995b, p. 475) finding strong support for the cream-skimming hypothesis, but a study replicating Smith and Meiers work did not find evidence that private schools drain the public i schools of their best students (Wrinkle et al., 1999, p. 1252).
In light of these criticisms and the mixed findings of prior studies, there is clearly a need, however, for more empirical research, particularly methodologically sound research, to examine the major opposing sides of the school choice debate, as represented by the opposing arguments over school choice. Some of the existing work in this area is open to criticism for its methodology or selection of variables. In

the Smith (1994) study, for example, the author based findings on district-level data rather than school-level data, which would seem to mitigate any influence of particularly well-run organizations on school effectiveness. Furthermore, the study used variables such as size of bureaucracy, measured as the number of school officials per student (p. 480) as the variable to test Chubb and Moes (1990) argument that bureaucracies diminish autonomy, which has a negative impact on effectiveness, rather than measuring levels of autonomy itself. As Bohte (2001, p.
93) points out, despite the fierceness of the arguments surrounding these issues, our knowledge of the impact of bureaucracy on school performance remains limited.
The Public v. Private Management Debate There has been a long standing debate in organizational theory as to whether generic organizational theories apply to both public and private sector organizations, or whether differences between the sectors preclude the applicability of theories from one sector to the other. Despite decades of research on this debate, there still exists uncertainty. Even where there is agreement that there is a difference between public and private management, the nature and extent of that difference is a source of further argument.
Schools of Thought
Scott and Falcone (1998, pp. 1-3) characterized three schools of thought in the public-private management debate. First, they identified the generic approach, which argues that public and private management is primarily the same. Second, the

core approach argues that there exist fundamental differences between public and private management. Third, the school they term the dimensional approach, represents those arguments acknowledging more of a continuum between the two and suggests that publicness involves more than an organizations formal, legal status.
The generic approach is also referred to as the universalistic perspective and suggests that management practices are not sector specific, and thus, best practices can be developed across sectors (Harel & Tzaffir, 2001). Although this approach is uncommon among contemporary public affairs scholars, there have been supporters (Baldwin, 1987; Gold, 1982; Murray, 1975).
The core approach to public-private management differences asserts a line of demarcation between public and private organizations based on legal ownership. If an organization is owned by government, it is characterized as public. If it is owned by a private individual/company, it is characterized as private (Antonsen &
Jorgensen, 1997). There are significant differences between the public and private sectors that provide support for the core approach. Three of these differences include the lack of profit motive in the public sector, the legal/constitutional environment of the public sector, and the diffusion of authority in the public sector as opposed to a single, formal line or chain of command (Harel & Tzaffir, 2001, pp. 321-322). Aligned with this approach is the argument that the fundamental differences between

the sectors precludes the effective transmission of management practices across public and private organizations (Allison, 1980).
The dimensional approach, first articulated by Myron Fottler in Is Management Really Generic? (1981) and Barry Bozeman in All Organizations are Public (1987), suggests that the difference between public and private is a matter of degree; publicness is both a behavioural category, not a legal one, and multidimensional (Antonsen & Jorgensen, 1997, p. 338). One of the major assets of this approach is its ability to incorporate mixed organizational forms, such as government enterprises and privately owned government contractors, which do not fit appropriately within the core approach framework. Scholars have recognized that .. .it has become increasingly difficult, and perhaps irrelevant, to classify organizations as purely public or purely private using a formal criterion. The boundary between the two is too blurred and the public sector acquires private sector characteristics at a rapid rate (Antonsen & Jorgensen, 1997, p. 338). One of the limitations of this approach, however, is the lack of consensus as to which factors to include in determining degrees of publicness. Bozeman (1987) focused on the impact of external political authority; Bozeman and Bretschneider (1994) focused on funding and ownership; and Antonsen and Jorgensen (1997, p. 337) defined publicness as organizational attachment to public sector values: for example, due process, accountability, and welfare provision.

Studies have come to different conclusions regarding the appropriateness of the three approaches. Bozeman and Bretschneider (1994) and Scott and Falcone (1998) provided some support for both the core and dimensional approaches. In both studies, some measures supported the argument that the public-private distinction is based on ownership (core approach); other measures supported the argument that other factors are also important. Furthermore, Scott and Falcones findings indicated that the core approach was more useful in explaining personnel functions and external outputs, but the dimensional approach was better at explaining large-scale procurement and internal research funding functions (p. 9).
Hybrid Organizations
As noted above, complicating the debate on public vs. private management is the existence of market-like reforms, which often have the effect of further blurring the lines between the sectors (Antonsen & Jorgensen, 1997). Education reforms are no exception; for example, the charter school movement results in schools that are funded by government but operated by private or non-profit organizations. As such, the dimensional approach is particularly useful to the present study.
Bozeman and Bretschneider (1994, p. 197) defined publicness as a characteristic of an organization which reflects the extent the organization is influenced by political authority. Applying this logic to schools, publicness would decrease based on the autonomy of the school. In other words, the dimensional approach to education allows for the development of a continuum of publicness, as

applicable to schools. Following Bozeman and Bretschneiders (1994) logic, traditional public schools would exhibit the highest degree of publicness, private schools the lowest, and charter schools would fall somewhere in the middle. Based on this logic, the charter schools would fall closer to public schools or private schools dependent upon the autonomy granted to a charter school by state statutes. In practice, however, there is much variation within school sectors. For example, EMO managed schools and Catholic schools may face large bureaucracies, particularly as compared to small, grass-roots charters or independent private schools. Vergari (2007), outlined four important areas influencing a schools place along a public-private continuum: 1) groups that found or operate the school, 2) school funding, 3) issues of governance and regulation, and 4) purposes, including a schools mission. This study acknowledges the important role of all four, but operationalizes sector only with regard to the first two, as outlined in the methodology section.
Empirical Studies
Over the years, scholars have examined differences between the sectors empirically. Studies have examined numerous aspects of management across sectors, including goals (Rainey, Pandey, & Bozeman, 1995), motivation (Baldwin, 1987), and leadership (Hooijberg & Choi, 2001). Studies examining the differences between the sectors in the areas of formalization, autonomy, and time required for decision making are the most applicable to the present study. As such, they will be the focus of this proposal.

Formalization has been defined as the extensiveness of rules and formal procedures and their enforcement (Rainey & Bozeman, 2000, p. 7). Some empirical studies have suggested that public organizations have more rules and procedures, both generally (Antonsen & Jorgensen, 1997; Chubb & Moe, 1990), and specifically in terms of personnel systems (Bozeman & Bretschneider, 1994). In other words, public organization personnel processes are more formalized. Other studies, however, have resulted in conflicting findings. For example, Kurland and Egan (1999, p. 447) reported that their findings indicated that public employees did not perceive higher levels of job formalization. Perception, however, may play an important role in such findings.
Within studies, formalization is closely tied to bureaucratization and red tape. In explaining the relationship between red tape and bureaucratization, Boyne (2002, p. 101) portrayed red tape as a pathological side-effect of bureaucracy.. .The existence of red tape implies an unnecessary and counter-productive obsession with rules rather than results, and with processes instead of outcomes. Although some scholars have been careful to distinguish between formalization and red tape (Bozeman, 2000), other scholars combine formalization variables within a measure of bureaucratization and/or red tape, as the two are closely related. As such, formalization is not always analyzed independently. For example, Bozeman and

Bretschneider (1994) looked at the time taken for hiring and firing personnel across
sectors as a component of their measure of bureaucratization/red tape.
Formalization, as well as red tape, has been identified in public human
resource management specifically (Bozeman, 2000). Scholars have noted differences
between the public and private sector in terms of human resource management:
Organizations in the public sector use more elaborate schemes of employee selection.. .and provide a voice mechanism... .However, there is less reliance on performance when it comes to determining the compensation of employees and managers..., and there is also less investment in training and development..., as well as fewer opportunities for employee participation in organizational decisionmaking... (Harel & Tzafrir, 2001, p. 337).
The literature acknowledges that public organizations are generally larger, older, and more unionized than private sector organizations (Harel & Tzafrir, 2001). These characteristics may lend themselves to greater formalization in human resource management practices. For example, one of the hypothesized reasons behind the suggested greater formalization in the public sector is the extent of unionization of that sector: The higher level of unionization in the public sector also leads to a higher level of formalization in the process of recruitment and selection in order to comply with the collective bargaining agreement (Harel & Tzafrir, 2001, p. 325). Along this argument, if public schools are heavily unionized and private schools are not, the public schools will more likely be subject to greater levels of formalization.

The literature suggests that public sector managers have less autonomy in their decision-making (Baldwin, 1987; Boyne, 2002). This appears to be particularly true in terms of human resource management capacity and is related to the literature on formalization and red tape. If public managers are subjected to large numbers of inflexible rules concerning hiring, firing, and promotion, it follows logically that their autonomy in decision-making is lower. Empirical research has supported these claims that public organizations are less autonomous than private organizations. Coursey and Rainey (1990) found that managers in public agencies have less autonomy over personnel issues than their counterparts in the private sector (See also Antonsen & Jorgensen, 1997). Research in the area of schools has supported the findings of public sector organizations, generally, with regard to autonomy. For example, research comparing public and charter schools has found that charter schools generally have more autonomy than public schools (Bulkley & Fisler, 2002; Podgursky & Ballou, 2001).
The literature also suggests that personnel decisions, such as hiring and firing personnel, take longer in public organizations than in private organizations (Rainey & Bozeman, 2000; Scott & Falcone, 1998). In terms of the present study, this would suggest that it takes longer to hire and fire teachers in public schools than in private schools. Presumably, if Chubb and Moe are correct, charter schools would fall

somewhere in the middle or perhaps closer to private schools than traditional public schools.
There is, however, some evidence that some of these issues concerning formalization, autonomy, and time are changing in the public sector, generally: For many years, public organizations have been less flexible than private organizations in choosing human resource and developing them. However, awareness of the strategic role of human resource practices has finally penetrated to the public organizations (Harel & Tzafrir, 2001, p. 345). The present study examines some of these classic issues in the area of human resource management in public and private sectors, as they apply to schools. In addition, it examines whether charter schools have had the desired effect of decentralizing decision-making processes, in terms of human resource management, to the schools, as well as within the schools.
Empirical Research on Human Resource Management in Schools
There have been a limited number of studies examining human resource practices in charter schools (Triant, 2001) and across school sectors (Ballou & Podgursky, 1998; Podgursky & Ballou, 2001). Thus far, research has generally supported the contention that charter school principals have more autonomy than their public sector counterparts. Bill Triants (2001) report, while quite limited in scope, illuminated some of the issues of formalization, autonomy, and time as they relate to charter schools. When asked what they believe is the best part of running a charter

school, the eight Massachusetts charter school principals2 he interviewed brought up issues such as the sense of freedom and the lack of teacher unions (p. 4).
Michael Podgursky and Dale Ballou are perhaps the most notable researchers examining human resource management across traditional public, private, and charter schools. In Teacher Recruitment and Retention in Public and Private Schools, (1998) they examined differences in personnel practices between public and private schools. In Personnel Policy in Charter Schools, (2001), they broadened the scope of personnel policies examined and focused on charter schools (including comparisons with traditional public and private schools, where possible). Their findings echoed some of the broader literature on personnel policy across sectors, specifically in the areas of formalization and autonomy. Overall, they concluded that, at least within the states examined, when given the opportunity, charter schools pursue innovative personnel policies that differ in key respects from those of traditional public schools and more closely resemble the practices of private school (2001, p. 24). As such, their research deserves substantial attention and serves as the focus of prior research in this area.
Prior studies have examined personnel policies across school sectors with regard to formalization and autonomy. Several of the notable areas have included hiring, firing, and compensation. In terms of hiring policy, Ballou and Podgursky (1998) examined personnel practices at public and private schools and found that
2 Titles also included Headmaster and Director.

private schools seemed to do a better job of retaining and developing effective teachers than public schools. Private school administrators usually have the ability to hire teachers who lack state certification, which is considered an important factor in this regard. Their 2001 report found that charter schools behaved more closely to private schools in hiring flexibility with respect to state certification: charter schools were much more likely than traditional public schools to employ teachers who lack regular state certificationeither those holding no license or those employed on an emergency license. Open-ended responses to their survey supported this finding, with a number of respondents commenting on the schools ability to hire non-certified teachers as among the most important ways that recruitment in their schools differ from that in traditional public schools. Open ended responses also included the perception that the small size and independence of the school permitted them to avoid the red tape of dealing with a personnel office, and that the recruitment process involved parents, teachers, and board members (p. 10-11). Qualitative research has supported Podgursky and Ballous findings, indicating that autonomy in teacher hiring, including the lack of state certification requirements were valuable characteristics of charter schools (Triant, 2001). There was at least some concern in Massachusetts, however, that new requirements, such as state certification for teachers, would reduce principals flexibility (Triant, 2001).

Davies and Quirke (2007) also presented findings on differences in teacher certification requirements, comparing public and private schools. Although their study didnt focus on human resource practices, they examined teacher certification differences, as a measure of formalization, between public schools, well-established, elite private schools, and a group of new sector private schools, which they argued were more subject to market pressures. With Toronto as the site of their study and without the inclusion of charter schools, the findings cannot be extrapolated to charters in the U.S, but are certainly relevant to the present study. Their logic is also of interest and involves competing theories as to whether schools respond to deregulation with a weakening of formalization (as suggested by market theory) or that such a response is mitigated by institutionalized perceptions of a school (as suggested by institutional theory). Their findings included extensive staff accreditation for public schools, less accreditation for elite schools, and the lowest level for new sector schools. All of the public schools in their study had an accredited principal and hired only accredited teachers. Nearly all of the elite schools (94%) had an accredited principal, but only one-third (33%) hired only accredited teachers. For new sector schools, the percentages were even lower, with less than half (49%) having an accredited principal and 14% hiring only accredited teachers.

Dismissal policies are also central aspects of human resources examined with regard to differences among school sectors. Ballou and Podgursky (1998; 2001) have investigated administrators flexibility in firing poor teachers. They describe charter schools as fairly aggressive in ridding themselves of ineffective teachers (2001, p. 15). Length of teacher contracts and the role of collective bargaining are important issues with regard to teacher dismissal. For example, their 2001 report indicated that very few charter school teachers worked under contracts of more than one-year (4%), which is typical of teachers in traditional public schools. The majority of charter school teachers (63%) worked under one-year contracts and one-third (33%) of them were employees at will. The authors attributed this at least in part to the lesser role of collective bargaining in charter schools as compared with traditional public schools, where faculties are significantly more likely to be represented by unions. For example, 90% of the charter schools with collective bargaining granted teachers tenure, but only 8% of the charter schools without collective bargaining granted tenure. Charter schools with collective bargaining were less likely to dismiss teachers than charter schools without collective bargaining. Relatively few charters, however, were subject to collective bargaining agreements. The Center for Education Reforms 2008 Annual Survey for Charter Schools, reported that only 15% of charter schools participated in a union or collective bargaining agreement. Triants (2001) qualitative study contributed some depth to this finding, with principals commenting on the lack of teacher unions as a benefit to being the principal of a charter school. For example,

one principal noted that he could call a teacher in and tell him that he is not doing a good job without being concerned that he is going to call the union rep (pp. 8-9).
Contracts and tenure have been a source of disapproval in traditional public schools for serving as a constraint on the autonomy of public school administrators: Teacher contracts in traditional public schools have been criticized for unduly constraining managerial prerogatives. Tenure protects ineffective teachers from dismissal (Podgursky & Ballou, 2001, p. 13). In Triants (2001, p. 6) study, more than half of the principals interviewed commented on their ability to remove bad teachers with reasonable ease. Several used at-will contracts that could be easily terminated at the end of a year.
Podgursky and Ballou (2001) also examined compensation policy and the use of salary schedules at traditional public, charter, and private schools. The majority of schools in each category used salary schedules, but there were notable differences. For example, nearly 100% of traditional public schools used such schedules, whereas 71% of charter schools and 63% of private schools used salary schedules. Furthermore, only 23% of the charter schools shared a salary schedule with the local school district. Whereas traditional public school salaries primarily depended on experience and education, charter schools considered factors such as superior performance and expertise in hard-to-staff subject areas (p. 16). In sum, the authors found that charter school salary offers appeared to resemble those in private schools, which had greater flexibility in structuring compensation packages and were

also much more likely to use incentives to recruit teachers in shortage subjects than
traditional public schools (p. 16). The authors noted, however, that they didnt have
comparable data on traditional public schools in terms of factors affecting salary
offers and recognized that some public school systems do have policies that permit
higher offers to teachers in hard-to-staff areas. Again, Triants (2001, p. 6)
qualitative study indicated similar findings, with one principal explaining the
importance of compensation flexibility:
In the regular district school, you are allowed to hire a certain number of teachers and the salary of these teachers is determined by the central office from an amount thats mostly determined by the teachers union. If you want more teachers at regular school you cant just have them. I can decide how many teachers I can have and, based on an objective criterion, how much I can pay each one.
In addition to the differences in offer salaries, Podgursky and Ballou (1998;
2001) found significant differences in compensation policies for existing teachers. In
the 2001 report, charter schools resembled private schools in the use of performance-
based pay. Half of the private schools and almost half of the charter schools used
performance-based pay and 31% of the private schools and 25% of the charter
schools used school performance bonuses. Traditional public schools were less likely
to use such performance-based incentives and even when it was used, it was in a more
limited way than charter schools:
Most charter schools that took individual performance into account rewarded it in a variety of ways, including one-time bonuses, advancing the teacher an extra step on the salary schedule, and additions to base pay in some other form. In traditional public

schools, by contrast, the dominant method for rewarding such incentives is a one-time cash bonus.. .Because such awards are not built into the teachers base pay, their long-term impact on compensation is much smaller (p. 17).
Furthermore, charter schools with merit pay generally applied it to more of their teachers than traditional public schools with merit pay, who generally used it mostly to reward the exceptional employee (p. 17). Open-ended responses to survey items asking charter school administrators how their compensation policies differed from traditional public schools reflected similar perspectives to the quantitative data. One interesting point was that some of the administrators responded that their performance-based pay was tied to areas such as standardized test score improvements, parent surveys, student attendance, and/or peer observation.
Ballou and Podgursky (1998) also considered effectiveness by having each principal rate their new teachers and experienced teachers. By having the principals make two rankings, they were able to mitigate the potential for sectoral bias. Their findings indicated that public school principals seemed to rate new teachers the same or slightly higher than principals in private schools. However, private school principals rated experienced teachers significantly higher than public school principals. From these analyses, the authors argued that Whatever the absolute quality of the initial workforce, then, private schools appear to do a better job of retaining superior instructors and developing the talents of their staffs (p. 406).

Other studies have looked at standardized test scores as a measure of charter
school effectiveness, with mixed results in terms of which sector is more effective in
educating students. Some studies have indicated that charter school students
demonstrate greater achievement (Greene, Forster, & Winters, 2003; Hoxby, 2004;
Solmon & Goldschmidt, 2004). For example, in a study of Arizona schools, Solomon
and Goldschmidt (2004, p. 21) found that students at charter schools began with
lower test scores, but that they demonstrated faster achievement growth than
students in traditional public schools, particularly in lower grade levels:
Thus, charter school students start with lower scores. They have patterns of faster achievement growth in elementary school, growth patterns roughly equivalent to students in traditional middle schools, and slower growth in high school. Even so, charter school students who completed the twelfth grade surpassed traditional public school students.
Other studies have found that charter schools exhibit lower levels of student achievement than traditional public schools (Crew & Anderson, 2003; Nelson, Rosenberg, & Meter, 2004). For example, in a study of Florida schools over the 1999-2000 school year, Crew and Anderson (2003, p. 198) found that students in regular school outperformed those in charter schools in every grade level in every subject. The authors did note, however, that the majority of charter schools in Florida did not serve traditional student populations. Furthermore, other studies have found mixed results or no difference in performance between charter and traditional public schools (Buddin & Zimmer, 2003, 2005). At any rate, measuring the

performance of charter schools is clearly a contentious issue. The difficulty in doing so is acknowledged in the literature (Greene et al., 2003).
Michael Podgursky and Dale Ballou provided significant research in the area of personnel policies in charter schools; however, there is clearly room for additional research in this area. Podgursky and Ballous 2001 report gave an account of research collected during the 1997-1998 school year. As a decade has passed since their research, the charter school landscape has changed dramatically during that time. Charter schools have become much more prevalent during that time and some have been in operation for close to 15 years. In addition, state policies may have changed over the years, which have the potential to dramatically impact the nature of charter school personnel policies. This study both supplements and updates the work theyve already done in the area.
Public Choice Theory
A brief discussion of public choice theory is central to the present study. Public choice theory serves as a conceptual foundation for school choice, including Chubb and Moes (1990) argument for the necessity of charter schools. School choice reforms are rooted in the idea that traditional public schools are too bureaucratic, which constrains innovation. The implementation of market-based reforms, such as charter schools, institutes competition and choice into public education. Subsequently, such competition and choice lead to the innovation that critics of traditional public schools claim they presently lack (Lubienski, 2003).

These claims of school choice proponents are based in public choice theory, which argues that bureaucratic public organizations lack the efficiency and effectiveness of private sector organizations. George Boyne and his collaborators (2003, p. 6) explained:
Three specific characteristics of public bureaucracies are believed to lead to a lack of efficiency and effectiveness: the monopolistic structure of public service markets; the absence of valid indicators of organizational performance; and the large size of many government agencies.
Public choice theory, epitomized by William A. Niskanens Bureaucracy and Representative Government (1971), also asserts that the lack of adequate performance indicators in the public sector exacerbates the problem. Whereas the market naturally encourages ineffective organizations to improve or dissolve, the absence of these indicators in the public sector allows ineffective organizations to survive indefinitely.
Although focusing on educational reforms in England and Wales, Boyne and others (2003) discussed the link between public choice theory and educational reforms directed at instituting market forces in education. A number of these reforms were similar to those undertaken or considered in the U.S. over the past decade, including school choice, publication of information on school performance, and vouchers. Public choice theory suggests that more competition, better performance information and smaller and more autonomous organizational units will lead to improved performance (Boyne et al., 2003, p. 119).

Although some public choice proponents advocate nothing less than the complete privatization of all but the most pure economic public goods, others have generated solutions for bureaucratic maladies. Examples include the institution of competition into public programs, increases in public information on effectiveness, and smaller organizations (Boyne et ah, 2003). In states where charter schools are required to publicly announce standardized test scores, they generally exemplify all three solutions. First, charter schools must compete for their student body and resulting funding. Second, they may be subject to the same standardized testing requirements as public schools, which gives the public additional information regarding their effectiveness and informs choice. Third, charter schools are generally smaller than other public schools (Bulkley & Fisler, 2002) and are relatively independent from the district bureaucracy.
Public choice theory has significant critics who argue against its various propositions. As pertaining to this study, public choice has been criticized for its tendency to overlook the coercive potential of public authority (Zafirovski, 2001, p. 681). In other words, top-down educational reforms, such as greater teacher certification requirements and standardized curricula may be an alternative means toward effectiveness in schools. Critics also point out that there is a difference between politics and the market (Zafirovski, 2001). As such, market-reforms may not be appropriate for the political environment. The main issue at hand, however, is not

the accuracy of public choice theory, but how some of the underlying principals have emerged as school choice reforms and the implications of such reforms.
Literature Review Conclusion
As addressed throughout this review, the literature on school choice, including Chubb and Moes (1990) proposal, and the literature on differences between public and private management are central to this study. Public choice theory provides further theoretical background for the market approach to school choice.
The school choice literature provides various rationales for the development of charter schools, often focusing on the need for the decentralization of authority within the public school system, and Chubb and Moes (1990) analysis is an important example of the market approach to education. Despite the extensive wealth of literature on charter schools today, very few empirical studies have looked at human resource practices in charter schools. In Politics, Markets, and America's Schools, the authors examined management differences between public and private schools, as well as between effective and ineffective schools. Their study, however, took place prior to the emergence of charter schools. This study contributes to one aspect of their comprehensive analysishuman resource management, in light of the existence of this new, hybrid school sector. Public choice theory has informed the market-based argument for charter schools and thus supplements the school choice literature for analyzing whether the charter school movement is an effective and appropriate prescription for human resource management failures within the

traditional public school system. In examining human resource differences across school sectors, this study provides support for Chubb and Moes argument that charter schools are a means to provide more autonomy to principals and builds on the small body of literature on human resource practices across sectors.
The public-private management debate provides a theoretical framework for looking at charters, which serve as appropriate subjects for analyzing the various approaches in the public-private management debate. This study examines the relationship of charter schools to the generic, the core, and the dimensional approach to public and private management. One of the benefits of the dimensional approach, as applicable to the present study, is that the dimensional approach handles all organizations, not just pure types (Antonsen & Jorgensen, 1997, p. 338), which makes it particularly useful for examining charter schools, which are hybrids between the public and private sectors. This study focuses on funding and ownership as the two dominant factors in determining degrees of publicness, as did Bozeman and Bretschneider (1994). Despite a large body of literature reporting differences between the public and private sectors, there is no consensus on which approach is most accurate. Although the study does not provide conclusive evidence for one approach, it contributes to the body of literature.

In sum, the literature on school choice and the public-private management debate provides historical background, a theoretical framework, and an existing base of findings. Furthermore, it served as the basis for the methodology decisions outlined in the following chapter.

METHODOLOGY Research Purpose
As noted earlier, the primary purpose of this research was to examine differences between public and private human resource management with regard to schools and to examine the implications of charter schools on school personnel practices. The existence of Chubb and Moes research into public and private schools in the 1980s, prior to the emergence of the charter school movement, coupled with their subsequent proposal for charter schools and states acceptance of this proposal provided an opportunity for a unique natural experiment. This project re-examined the personnel dimension of the Chubb and Moe study, but rather than limit the study to private and public schools, this study included a category of charter schools to examine the validity of Chubb and Moes propositions concerning the effects of choice and autonomy on public schools.
The research question was as follows: Do differences in human resource management exist between public, private, and hybrid (charter) schools? This research question tested public-private management theories, as well as Chubb and Moes (1990) theoretical foundation for their charter school proposal. As a result, hypotheses for this question are outlined below.

The following three hypotheses were based on the research question and
captured some of the suggested differences in human resource management between
public, private, and charter schools as identified in the literature:
Hi: Personnel policies are more formalized in more public schools.
H2: Principals in more public schools have less autonomy than principals in more private schools in terms of personnel decisions.
H3: From the time where a need to hire or fire personnel is recognized, such personnel decisions take longer in more public schools.
Study Design
Data collection methods consisted of the following: 1) a survey to principals of elementary schools in Arizona, Colorado, Georgia, Massachusetts, Michigan, and the District of Columbia3 and 2) interviews with the principal/head of school at 16 schools in Michigan, Massachusetts, and Colorado, including a mix of public, charter, and private schools. These states and the District of Columbia4 served as an appropriate sampling frame for this study, as they provided the necessary variability to analyze some of the issues central to the charter school debate, but focused on the stronger states in terms of state charter laws, as assessed by the Center for Education Reform (Charter school laws across the states: Ranking and scorecard, 8th edition,
3 Kansas was originally selected as a seventh state to be combined with Georgia for comparisons between weak and strong charter states, but extremely low numbers of charter schools in that state prevented its inclusion in the study.
4 For the purposes of simplicity, the District of Columbia will be referred to as a state throughout this proposal.

2004). The organizations ranking system scores each state law based on 10 criteria, based on the extent the state law supports charter school development and autonomy. Focusing on these stronger charter states was important, both because these states had higher numbers of charters from which to sample and because charters in such states were more appropriate for testing Chubb and Moes claimsstates with stronger charter laws better reflect the components of Chubb and Moes proposal that free charters from many of the requirements imposed on traditional public schools.
Sampling for States
Arizona, Colorado, Massachusetts, Michigan, and the District of Columbia were the primary focus of analysis, as they have strong charter laws that are among the closest to that envisioned by Chubb and Moe (1990). All of these states were ranked in the top 10 of the 2004 Center for Education Reforms rankings of strength of charter school laws. Georgia, on the other hand, received weak designations from the Center of Education Reform in terms of their charter laws. This state was originally included to allow for secondary analyses based on strength of laws. Unfortunately, the low numbers of charters in Georgia prevented any meaningful analysis based on strength of state charter laws.
States were selected based on several considerations: 1) the need for several states with particularly high rankings in terms of strength of state charter laws, particularly in relevant sub-categories such as automatic waiver from state and district laws, legal and operational autonomy, and exempt from collective

bargaining agreement/district work rules, as measured by the Center for Education Reform 2) variability in charter laws across states (e.g., including a weaker state, a mix of right-to-work states and non- right-to-work states, including states with a significant for-profit EMO presence) 3) sufficient numbers of charter schools in each state, 4) states that have had charter laws in place for at least 5 years (all of these states had charter laws in place since at least 1996). The states selected fulfilled these requirements. These states were selected with thoughtful consideration, as charter school law provides a fruitful background for analysis. A brief overview5 of each state follows Table 3.1, which provides an overview of state sampling.
5 All state characteristics were best available data as of 2005, unless otherwise noted.

Table 3.1. Sampling overview
CER Ranking of Strength of Charter Law 1 3 5 6 9 26
Year of Charter Law 1994 1996 1993 1993 1995 1993
# of Charter Schools (Winter 2004) 491 43 210 50 93 36
Combined Score on 5 selected CER categories (total possible score = 25) 24 24 21 20 20 10
Right to Work State? Yes No No No No Yes
Charters may be granted directly to for-profit organizations? Yes No No No Yes No
Charter school may be managed by a for-profit organization? Yes Yes Yes Yes Yes Yes
Arizona was ranked highest in the nation in terms of strength of charter law by the Center for Education Reform in 2004. It received the highest score possible (5) in several important subcategorieseligible charter applicants, legal/operational autonomy, and fiscal autonomy. The states charter law was first passed in 1994

and the state had 491 charter schools in operation in 20046. The number of charter schools dipped slightly over the past few years to 479 in 2008. Arizona was a right-to-work state.7 Private organizations were eligible to apply for a charter and converted private schools were permitted. State certification was not required for charter school teachers in Arizona at the time of the ranking.
Colorado passed its charter law in 1993; the state had 93 charter schools in operation in 2004 and 140 as of 2008. Although Colorado charter law was ranked slightly lower than the other strong states (receiving a B instead of an A), Colorado was considered moderately strong (Lubienski, 2003, p. 409) and was an appropriate state in which to conduct this research. Additionally, Colorado charter schools were permitted to request waivers from some state requirements and waivers regarding personnel issues were considered among the most important for charter school employees (Bulkley & Fisler, 2002). For example, state certification for teachers was required, but could be waived. For-profit organizations were permitted to apply for a charter, but converted private schools were not permitted. Colorado was not a right-to-work state, but charter schools were not bound by school district
6 The information on year of charter law passage, rankings, number of charter schools in 2004, and additional information for each state was taken from the Center for Education Reforms Charter School Laws Across the States: Ranking and Scorecard, 8th Edition (2004) unless otherwise noted. The rankings were based on charter laws in effect as of December 2003. The 2008 number of charter schools came from the Center for Education Reforms Charter School Enrollment and Closures, by State, retrieved August 25, 2008, from
7 The information on Right to Work status for each state was taken from the National Right to Work Legal Foundation website at

collective bargaining agreements. Charter applicants with schools designed to serve low-achieving students were given priority (State charter law profiles, 2005). District of Columbia
The District of Columbia passed its charter law in 1996 and had 43 charter schools in operation in 2004. That number increased to 78 by 2008. Although the 2004 numbers were low compared with the other strong states, DC received the highest score possible in some of CERs subcategories, including eligible charter applicants, automatic waiver from state and district laws, and exempt from collective bargaining/district work rules. No other state received a 5 in both the automatic waiver and exempt from collective bargaining subcategories in that year, two areas which are particularly applicable to personnel policy issues. Additionally, it was ranked third in the nation in terms of strength of charter law. The District of Columbia is not a right-to-work state. Converted private schools were permitted, but for-profit organizations could not be direct recipients of charters.
Massachusetts passed its charter law in 1993 and had 50 charter schools in operation in 2004 and 62 in 2008. It was ranked 6th in 2004 in terms of strength of charter law. Massachusetts categorized its charter schools as Horace Mann (converted public schools) and Commonwealth (start-up charters). Certain aspects of the charter law (such as collective bargaining rules and waivers from state/district rules) were dependent upon which type of charter the school was, with

Commonwealth charters having more freedom than Horace Mann charters. Massachusetts is not a right-to-work state and state certification was required for teachers (or they had to pass the Massachusetts Educator Test). Charter applicants for poor performing districts were given priority.
Michigan has had its charter law in place since 1993 and had 210 charter schools in operation in 2004. By 2008, the number of charter schools increased to 245. The state was ranked 5th in terms of strength of charter law in 2004. For-profit organizations could not receive charters directly, but they were permitted to manage schools. In addition, converted private schools were permitted. Charter school teachers were required to hold state certification, with the exception that faculty from charter sponsors (universities/community colleges) could teach in one of their sponsored schools. Although charter schools were not given automatic waivers from state and district laws, they could apply for such waivers. Michigan was particularly interesting for its large presence of for profit EMOs.
As mentioned above, Georgia was selected to enable comparisons between strong and weak charter law states. Georgia received a C from the Center for Education Reform 2004 rankings, ranking 26th in terms of strength of charter law.
The state had 36 charter schools in operation in 2004, but this has since increased to 65. CER gave the state particularly low rankings in the following relevant sub-

categories: 0 in automatic waiver from state and district laws, 1 in legal/operational autonomy, and 1.5 in exempt from collective bargaining/district work rules. Local school boards were the only chartering authority for the state of Georgia. Converted private schools were not permitted and for-profit organizations could not receive charters directly, but were allowed to manage charter schools.
Much of the specifics, such as collective bargaining issues and state certification requirements were specified in each charter. Georgia is a right-to-work state.
As indicated by the state summaries, the sampled states clearly had variation in terms of the strength of the state charter laws, chartering authorities, eligible charter applicants, and right-to-work status. In addition, these states represented a large number of the charter schools operating in the United States and provided insights into the complex considerations of charter policy.
A survey was the primary means of data collection for this study. Survey methodology has been used elsewhere in examining similarities and differences between public and private organizations (Antonsen & Jorgensen, 1997), and specifically for differences in human resource management between public and private sectors (Harel & Tzafrir, 2001). The survey enabled the collection of data from a large number of schools and the qualitative interviews supplemented this survey with in-depth information on sixteen of the schools. I pilot tested all protocols (i.e. questionnaire and interview protocols) at two local elementary schools.

Survey participants primarily included the principal or head of school at each sample institution. For the sake of simplicity, the term principal is used to cover all such titles throughout this dissertation. Individuals in this position were appropriate participants, as they have knowledge of school/district policy and procedures, are involved with school personnel issues, and have a broad-based perspective on the performance of their teachers and students. Furthermore, principals tend to be the primary link between the school and any higher bureaucratic authority (e.g., local school board, charter board). Therefore, they are in a unique position to judge the external constraints on their authority and have been the subjects for other research studies examining personnel issues at the school level (Ballou & Podgursky, 1998; Triant, 2001).
Independent Variable
In this study, publicness was the independent variable, as represented by the levels of public, charter, and private schools. Publicness as an independent variable has been an approach taken by other studies (Antonsen & Jorgensen, 1997). Levels of the independent variable were operationalized as follows:
Public Schools: Schools that are funded, owned, and operated by the public sector. 8
8 The study requested that the principal or headmaster at the institution fill out the survey. This individual may have various titles, including head of school, headmaster, or director.

Private Schools: Schools that are funded, owned, and operated by nonpublic entities.
Charter Schools: Those schools operating under the charter school laws of their respective states. For these schools, the primary funding, ownership, and/or operation functions are generally divided between the private and public sectors.
This operationalization of variables was further supported by that of Scott and Falcone (1998) in their determination of public, private, and mixed sectors of research and development laboratories. In their study, they defined government laboratories as those solely owned and operated by an agency of the government, private laboratories as those solely owned and operated by a private-sector industrial organization or conglomerate and mixed sector laboratories as those owned and operated by more than one parent organization from public and private sectors or that have portions of their facilities and programs owned by an industrial parent and portions owned by a governmental organization (pp. 4-5).
Dependent Variables
There were four primary dependent variables: formalization of personnel procedures, principal autonomy, school level autonomy, and time for personnel decisions. Formalization of personnel procedures was operationalized as the extent to which personnel policies were written and inflexible. Indicators included the extent personnel rules were developed and written and the extent to which exemptions could be made. Principal autonomy was primarily defined as the relative ability of the principal to make personnel decisions independently. Indicators included the

influence of various groups involved in personnel decisions and the extent of freedom or autonomy principals reported. This study also investigated a second aspect of autonomyschool level autonomy. One scale combined multiple items indicating the influence of various groups at the school level, such as teachers and curriculum specialists, providing an alternative operationalization of autonomy based on the extent of school level influences. The time for personnel decisions was defined as the average time it takes for core personnel decisions (i.e., hiring and firing) to be made. Control Variables
Control variables included school size, location (i.e., urban, suburban, or rural area) and free and reduced lunch eligibility (as a proxy for socio-economic status of the student body)., School size may have a strong impact on the bureaucratization of personnel policies in a school. Other studies on public and private sector differences have also controlled for size (Bozeman & Bretschneider, 1994; Harel & Tzafrir,
2001) and the importance of doing so has been recognized (Boyne, 2002). This study operationalized school size as the number of students enrolled in the school. School location may also influence school organization, so this variable was included and operationalized as urban, suburban, and rural.
Free and reduced lunch eligibility, as a measure of socio-economic status, was also included as a control variable. Socio-economic status (SES) of the student body has been found to have a strong effect on student achievement. It is quite possible that high performing schools have fewer organizational challenges that lead to tighter

controls in the form of increased formalization and reduced autonomy for principals. As such, including this variable helps to mitigate any effects of student populations on the dependent variables. This variable was operationalized as the percentage of students eligible for free or reduced lunch at the school. Although free and reduced lunch eligibility is not a perfectly accurate measure of poverty in a schoolcriticisms include a lack of eligibility verification and decline in participation as students reach higher levels of schooleducation researchers use this measure almost exclusively, due to the lack of a viable alternative currently available. The National Center for Education Statistics is experimenting with new poverty measures (Viadero, 2006), but until a better measure is available, free and reduced lunch eligibility is the most appropriate indicator for use in this study.
Sampling for Subjects
The sampling frame was all elementary schools (K-5/6) in the following states: Arizona, Colorado, Georgia, Massachusetts, Michigan, and Washington, D.C. The sampling method was stratified random sampling. More specifically, surveys were initially sent to all charter schools in states with fifty or fewer elementary charter schools, with a corresponding number of surveys sent to a random sample of the same number of public and private schools in those states. For states with greater than fifty charter schools, surveys were initially sent to a random sample of 50 charters, 50 public schools, and 50 private schools in the state. Due to an extremely low response rate initially, which is discussed later in this chapter, I increased this

sample to include all charters in Colorado and Michigan, with an equal number of public and private schools in those states. The random samples were drawn with the use of a random numbers table and lists of schools in each state, which were available from each states department of education website9 (public schools and charter schools) and (private schools), based on the 2005-06 school year.
Sixteen elementary schools from the Colorado, Michigan, and Massachusetts sample participated in interviews. These schools were selected to achieve a mix of characteristics including variety among the following: socio-economic status of the student body, location of the schools (urban, suburban, rural), different types of charters (e.g., grass-roots, managed by for-profit EMO, core knowledge curriculum, etc.), and different types of private schools (e.g., religious affiliation, dual language focus, etc.). Particular emphasis was paid to include a couple of schools distinguished for high performance with high percentages of free and reduced lunch and minority student populations. There is no one charter model, so it was important to select schools with a mix of characteristics for the study. Seeking out schools with some of these characteristics, such as the existence of a management organization or religious oversight body, was particularly important, as such oversight bodies may impose constraints on principals that are not typical of other schools in the sectors.
9 Arizona: Colorado: home.htm. Massachusetts: Washington, D.C.:

As an incentive to improve response rates, the requests for participation included information on a lottery for all participants. The winner of the lottery, a public school in Arizona, received a $300 donation to the school to be used at the principals discretion.
Survey Protocol
The survey consisted of an online survey, with requests to participate sent to 1191 principals in the states of Arizona, Colorado, Georgia, Massachusetts,
Michigan, and the District of Columbia. The survey consisted primarily of close-ended questions, but included several open-ended questions as well. Don Dillmans (Salant & Dillman, 1994) survey protocol was followed, but adjusted for use with an online survey and the availability of e-mail addresses. The following steps comprised this process:
The first contact with the initial round of requests consisted of letters mailed to the sampled principals explaining the research and requesting that they participate in the online survey. The URL for the survey included an identifier to track which schools returned the surveys, as well as to link test scores to survey responses.
A postcard reminder followed the initial letter.
I sent e-mails to those in the sample for which I could obtain e-mail addresses.
In all, I sent four separate requests to each potential respondent to participate in the survey, generally a mix of letters, postcards, and e-mails, depending upon whether I could obtain e-mail addresses for the respondents.

Interview Protocol
The interviews conducted at the sampled schools ranged from 33 minutes to one and a half hours in length. Participants typically included the principal of each school, but there were a few instances where a human resource officer, president, or similarly titled position participated in the interview in lieu of or in addition to the principal. The format was a semi-standardized interview, based on Bruce Bergs (2001) descriptions. A number of questions, with probes, were drawn up in advance and approved by the universitys institutional review board. Additional questions were asked during the interview to follow up on participants responses, as necessary. Whereas the survey was primarily used to obtain information on human resource management policies across the sectors and to test proposed hypotheses, the interviews provided more in-depth information on individual schools and uncovered some of the nuances of human resource practices at schools. These interviews further explored the principals perceptions of formalization, autonomy, and time required for decision making, as well as the effects of these issues on their management practices. This allowed for more descriptive information on how a few schools/principals manage their staffs, what types of constraints they operate under, and their impressions of the strengths and weaknesses of the current policies. It also generated further areas of interest for study.

The quantitative component of this study entailed multiple analyses, typically using SPSS software to conduct statistical analyses. A description of each type of statistical analysis follows.
Chi-square Analyses
I conducted chi-square analyses on the data to compare the differences in frequency of responses across school sectors, using SPSS software. Alpha was set at .05, so only analyses with p-values below .05 were deemed significant. The results chapter includes relevant statistical information, including p-values, chi statistics, and per cell percentages.
Adequate cell count judgments meant all cells in 2 X 2 tables had expected cell counts of at least five or at least 80% of the cells in larger tables had cell counts of at least five (with no cells having a zero expected cell count). In order to achieve sufficient cell counts for many of the analyses, I collapsed categories across similar responses. Within the survey, there were two primary types of questions designed to tap the dependent variables. One set of questions asked respondents to Please indicate how strongly you agree or disagree with the following statements, based on the scale below. Options for responses included the following: 1) strongly disagree, 2) somewhat disagree, 3) neutral, 4) somewhat agree, 5) strongly agree, and 6) N/A. Unless otherwise stated in the results section, these items were collapsed in the following manner: 1) strongly or somewhat disagree, 2) neutral, and 3) somewhat or

strongly agree. Similarly, a second type of question asked respondents to do the following: Using the scales below, how much ACTUAL influence do you think EACH group or person has on the following personnel decisions, followed by several types of personnel decisions and groups and individuals that would potentially influence such decisions. The options for responses on the scale were as follows: 1) no influence, 2) very little influence, 3) moderate influence, and 4) a great deal of influence. Unless otherwise indicated in the results section, these items were collapsed, combining no influence with very little influence and also combining moderate influence with a great deal of influence.
Independent t tests
Independent t tests enabled the analysis of the means for some of the survey items, including time for hiring and dismissal, as well as some of the overall scales, such as the summated formalization scales, the summated autonomy scale, and the school level autonomy scale. Again, alpha was set at .05 and relevant statistical data are included with the results in the next chapter.
Because t tests are not robust to conditions of nonhomogeneity of variance when the ns are not equal, I conducted and analyzed Levenes test for equality of variances prior to the analysis of the independent t test. The results of these tests were considered in the interpretation of the independent t tests.

Multiple Regression
Multiple regression analyses, in general, permit the identification of percentages of the variance in the dependent variable accounted for by each independent variable, resulting in the ability to predict the dependent variable based on the given independent variables (Bernard, 2000). In the case of the present study, the regression analyses treated publicness as the independent variable and formalization, time for personnel decisions, principal autonomy, and school level autonomy as the dependent variables to examine how sectors compare in terms of personnel practices. Since publicness is measured at the nominal level, a dummy variable was used to analyze the data. This has been a recognized method for comparing across sectors (Cirincione, 1999). The results allowed for the prediction of formalization, autonomy, and time based on the independent and control variables. For example, sector, school size, and free and reduced lunch eligibility each accounted for some variance in the given dependent variable, although not all had significant results.
The independent variables included in each regression were as follows: school size, free and reduced lunch eligibility, sector (using a public sector dummy variable and a charter sector dummy variable), and school location (using a rural location dummy variable and an urban location dummy variable). The public sector dummy variable had two levels, traditional public or not traditional public (including both charter and private schools). Likewise, the charter sector dummy variable had two

levelscharter or not charter, with the not charter level including both traditional public and private schools. The rural and urban dummy variables had similar formats, with two levels for rural (rural and not rural) and two levels for urban (urban and not urban). This model, using the enter variable selection method, was run on each of the following four dependent variables: 1) summated autonomy scale, 2) summated formalization scale, 3) weeks to fill teachers position, and 4) weeks to dismiss an ineffective teacher. The results chapter presents relevant findings.
Cronbach s Alpha
I combined multiple survey items to create scales during the analysis component of this research study: 1) summated formalization scales, 2) summated autonomy scale, 3) summated board and district influence scale, and 4) summated school level autonomy scale. The nature of these scales is discussed in more depth in the results section, including a description of the items included for each and which items required a reversal of polarization. In order to assess the internal reliability of these scales, I calculated a Cronbachs alpha for each scale with SPSS software. Interpretations of reliability were based on the following: >.8 as good, >.7 as acceptable, > .6 as questionable, > .5 as poor, and < .5 as unacceptable. Discussions of the scales analyses in the results chapter include information on the Cronbachs alpha calculations.

Problems and Limitations
There were clearly some limitations of this study. It was not meant to generate a definitive answer on the implications of charter schools for school human resource management. Rather, this dissertation was designed to contribute to the body of literature that examines the implications of school sector on human resource practices. As outlined in chapters 4 and 5, the study demonstrated such a relationship, but there are important limitations to the findings. Several main limitations of this study are discussed, including the potential influence of state charter laws, low response rates, and selection bias.
First, the charter laws of the state in which the research was conducted may have served as a constraint on the actions of charter schools. For example, since charter school laws vary by state and typically do not take into account all of Chubb and Moes (1990) recommendations, no ideal exists to test the results of their proposal. This issue was mitigated, however, by including education systems from more than one state and focusing on states with strong charter school policies, including Arizona, Michigan, Massachusetts, and the District of Columbia, which all received A grades from the Center for Education Reforms rankings of strength of charter school laws (Charter school laws across the states: Ranking and scorecard, 8th edition, 2004). Furthermore, the central focus of this study is on examining the implications of charter schools in practice, as a way to contribute to education policy as well as theory.

A second, more practical difficulty in this research involved generating adequate response rates. The raw response rate was approximately 20%, with 233 responses to 1191 requests to participate in the survey. Charter schools have been currently heavily researched, which probably negatively affected response rates. Low response rates were not as problematic for the primary research question, as it involved looking at formalization, autonomy, and time for decision making across sectors broadly. Low response rates, however, prevented meaningful analysis of secondary research questions concerning variability within the charter school sector that this study initially planned to address. Every effort was made to encourage survey completion, which included extensive effort to dig up e-mail addresses, as notices sent via e-mail seemed to generate more, as well as quicker, responses to the survey. As noted, four attempts were made to reach each principal, including a mix of mail and e-mail notices, where possible, and a $300 lottery for participants also served as an incentive for participation.
Selection bias was a third significant potential problem for this study. If students have choice in attending a charter school, there could be something about those students that makes it easier for a school to be organized effectively. This potential for creaming or other stratification is recognized and has been a major source of disagreement in the literature. Conversely, charter schools often have high percentages of minority and low-income students. To offset some of this potential

bias in either direction, this study controlled for free and reduced lunch eligibility (as a proxy for socio-economic status).
A principals choice in schools also has the potential to result in bias, particularly in light of the complexity in defining and measuring concepts such as formalization and autonomy. Although this study defined both terms, as outlined earlier in this chapter, many of the survey items rely on perception and responses can therefore be quite subjective. For example, formalization was defined as the extent to which policies are written and inflexible. Although it may be fairly clear cut as to whether a particular policy is written or not, there could be large variety in how detailed the written policies are, as well as the extent to which they are inflexible. Furthermore, when asking more general questions about the extent to which policies are formalized or the extent to which a principal has autonomy, there is the potential for a wide variety of perceptions across individuals. For example, what one principal sees as a formal policy, may not seem quite so formal to another principal. With the potential for principals to self-select their schools, this can be more problematic and may mask differences between the sectorsprincipals in more private settings may choose such schools because they prefer a more autonomous working environment and principals in more public settings may prefer a more structured environment guided by strict policies and procedures. Such individuals may have very different perceptions as to what constitutes a constraint on their autonomy. Previous work experience may also play a role in perception. A principal coming

from a public school and one coming from a private school into charter schools with similar levels of formalization and autonomy may have very different perspectives as to the extent of formalization or autonomy in the new setting. This issue is discussed at greater length in the conclusion, in light of the studys findings.
In summary, there are three major potential limitations to this study. First, the variety of charter laws across states prevent the testing of a more pure charter model, as outlined in Chubb and Moes (1990) work. Second, low response rates prevented meaningful quantitative analysis of the variety within sectors. Third, selection bias could have impacted the studys findings, both in terms of student body characteristics across sectors, as well as choice of sector by principals.

Again, the research question was as follows: Do differences in human
resource management exist between public, private, and hybrid (charter) schools?
The hypotheses based on this research question were as follows:
Hi: Personnel policies are more formalized in more public schools.
H2: Principals in more public schools have less autonomy than principals in
more private schools in terms of personnel decisions.
H3: From the time where a need to hire or fire personnel is recognized, such
personnel decisions take longer in more public schools.
Results are reported below, beginning with a description of the survey sample and
then a description of the qualitative sample. The remainder of the chapter discusses
the quantitative results in the areas of formalization, autonomy, and time; followed by
the qualitative results in these areas.
Descriptive Statistics for the Survey Sample
Descriptive statistics for the survey sample indicated that a wide variety of
schools across public, charter, and private sectors were included in the sample. This
section provides a description of the survey sample with regard to control variables,
as well as a few additional characteristics of interest to achieve variability among
respondents: state, location, free and reduced lunch eligibility, school size, and union

affiliation. A mix of public, charter, and private schools from each state returned surveys, with the largest number of responses coming from charter schools. Of the 233 surveys returned, 80 or 34% of the surveys were from public schools, 99 or 43% were from charter schools, and 54 or 23% were from private schools. This variety was essential for analyzing the hypotheses. The diverse sample included schools with a variety of locations, economic situations, and school sizes. There was also variety in terms of union affiliation, but this was clearly divided by sectors. The following provides an overview of the sample with regard to each of these areas.
State. As noted in Chapter 3, the sample primarily included schools within states with strong charter laws, as ranked and assessed by the Center for Education Reform. A smaller sample of schools from a weaker state, Georgia, was included in the unsuccessful attempt to assess variability within the charter sector, based on strength of state charter law. The respondents reflected this breakdown as well212 or 91% of the respondents were from schools in strong charter law states and 19 or 8% of respondents were from schools in the weak charter law state (Georgia). For the remaining 1% of respondents, state was unknown.
In the states with large numbers of charter schools, there was a good mix of responses among the three sectors. However, for the states with the smallest numbers of charter schools, and subsequently the smallest numbers of schools in the original sample, there were very small numbers of surveys in some categories. For example,

only one private school in Washington, D.C. and only two private schools in Georgia returned surveys. For more detail, please see Table 4.1.
Table 4.1. Sample descriptive statistics by state
Strong Weak State Unknown Totals
# Public Returned 18 4 29 9 13 7 0 80
# Charter Returned 8 7 28 7 38 10 1 99
# Private Returned 9 1 21 3 17 2 1 54
Totals 35 12 78 19 68 19 2 233
% of Total Returned 15% 5% 33% 8% 29% 8% 1% 100%
Location. Schools from a variety of locations, as measured by rural, suburban, and urban; returned surveys, including a mix of schools across sectors in these categories. Overall, the highest percentage, 39%, of the schools participating were located in suburban areas, with 32% located in rural areas and 30% located in urban areas. There was more variability within sectors than across the entire sample, as depicted in Table 4.2 below.

Table 4.2. Sample descriptive statistics by sector
Public Charter Private Total
N % N % N % N %
Location Rural 31 38.8% 26 26.8% 15 28.8% 72 31.6%
Suburban 26 32.5% 36 37.1% 26 50.0% 88 38.5%
Urban 23 28.8% 35 36.1% 11 21.2% 69 29.9%
Total 80 97 52 229
Free and Reduced Lunch Mean % FRL 42.63% 48.57% 11.71% 37.80%
Less than 33% 31 40.8% 34 37.4% 44 89.8% 109 50.9%
33% to Less than 75% 33 43.4% 32 35.2% 3 6.1% 68 31.2%
75% or More 12 15.8% 25 27.5% 2 4.1% 39 17.9%
Total 76 91 49 216
School size Median # of Students Enrolled 499 322 143 332
Union Affiliation No 15 19.7% 90 95.7% 51 98.1% 156 70.3%
Yes 61 80.3% 4 4.3% 1 1.9% 66 29.7%
Total 76 94 52 222
Free and reduced lunch. The percentage of students eligible for free or reduced lunch, as a measure of poverty, was a control variable. As expected, public and charter schools reported higher percentages of students eligible for free or reduced lunch. The mean percentage of students eligible for free or reduced lunch was 42.6% for public schools, 48.6% for charter schools, and 11.7% for private

schools. Across all schools, the mean percentage of students eligible for free or reduced lunch was 37.8%. The distinctions among the schools were further clarified by categorizing responses. Public and charter schools reported having less than 33% of students eligible for free or reduced lunch at similar rates of 40.8% and 37.4%, respectively. However, over a quarter of charter schools (27.5%) had three-quarters or more of their students eligible for free or reduced lunch, as compared with only 15.8% of public schools. As anticipated, private schools reported much lower percentages of free or reduced lunch eligibility, with 89.8% of private schools having less than one-third of students eligible for free or reduced lunch and 4.1% of private schools having three-quarters or more of students eligible for free or reduced lunch.
School size. There was great variability in terms of the size of schools in the sample, as measured by students enrolled in the school. The overall median for school size was 332 students across all sectors, but the range was from 13 to 2,650 students. The existence of some schools of much larger than average size brought the overall mean up to 398 students. Public schools were the largest, with a median of 499, a mean of 513, and a range between 72 and 1457 students. Charter schools were the next largest, with a median of 322, a mean of 414, and a range of 31 to 2,650. Private schools were the smallest in the sample and had a median of 143 students, a mean of 204, and a range between 13 and 661 students.

Union affiliation. As anticipated, union affiliation varied greatly by school sector, with public schools reporting unionization at the highest rates. One of the survey items asked the following question regarding union affiliation: Are teachers in your school affiliated with a union? In response to this question, 80.3% of public schools, 4.3% of charters, and 2.0% of private schools indicated that their teachers were affiliated with a union.
As described in this section, the study sample provided the variety sought in terms of school sector, state, school location, free and reduced lunch, school size, and unionization. As is explained in more detail later in this chapter, small numbers in some categories prevented meaningful analysis of some facets of the research questions. Relatively small numbers of private school respondents, as compared with the public and charter respondents, may have also impacted the power of some of the analyses of differences between charter and private sector schools. The relatively larger numbers for public and charter schools, however, was useful for the focus of this study on charter-public differences. The discussion will now turn to the descriptive statistics for the qualitative sample, but please see Table 4.2, above, for more detail on the variability of respondents in the survey sample.

Descriptive Statistics for the Qualitative Sample
A very diverse group of schools participated in the qualitative component of this study, consisting of sixteen interviews, primarily with the principal or head of school. In a few instances, administrators with other positionstwo human resources directors, a chief financial officer/chief operating officer, a president, and a manager of business and developmentjoined the interviews either in addition to or in lieu of the school principal. As noted in the methodology chapter, the qualitative sample consisted of schools that reflected a mix of characteristics, including the socioeconomic status and racial/ethnic diversity of the student body, as well as the location of the school (both in terms of urban/suburban/rural and the state).
Four public schools, six charter schools, and six private schools participated in the qualitative component of this study. At least one school from each sector served a low-income, high minority student body. The majority (10) of the schools participating in the interviews were located in the state of Colorado. Three schools from Massachusetts and three from Michigan also participated, one representing each sector from each of those two states. Nine schools were located in urban areas, five in suburban areas, and two in rural areas. Table 4.3 outlines the number of schools in each of these categories.

Table 4.3. Descriptive statistics for the qualitative sample

Total Interviews 4 6 6 16
State MA 1 1 1 3
CO 2 4 4 10
MI 1 1 1 3
SES High FRL Eligibility 3 2 1 6
Low FRL Eligibility 1 4 5 10
Racial/ Ethnic Diversity High Racial/ Ethnic Diversity 3 3 4 10
Low Racial/ Ethnic Diversity 1 3 2 6
Location Urban 3 2 4 9
Suburban 1 2 2 5
Rural 0 2 0 2
Schools comprising this qualitative sample also reflected a variety of other characteristics including different types of foci, curriculum, and awards. Many of the schools had been recognized for their outstanding performance; some under unlikely circumstances such as low-income student bodies. For example, one high-performing private school served a predominantly Hispanic population with a majority of English language learners. School sizes ranged from 30 to over 2,500 students. The following list describes some of the other interesting characteristics of at least one of the schools included in the qualitative component:
Existence of a for-profit EMO
National Blue Ribbon school
School charter held by a university
Bilingual education
Core knowledge curriculum
Focus on music

Focus on food, agriculture, and the environment
Focus on arts
Elite, nonsectarian private school
Catholic affiliation
Other religious affiliation
Grass-roots charter school
Taken as a whole, the diverse mix of schools included in the qualitative sample provided a rich variety of data to explore some of the issues concerning human resource practices in schools across the sectors. These qualitative findings will be discussed later in this chapter, following the discussion directly below of the quantitative results.
Quantitative Results
The following provides results for the three hypotheses based on the research question and captures some of the suggested differences in human resource management between public, private, and charter schools as identified in the literature. Summaries of statistical analyses for each hypothesis are included in the appendix.
Formalization refers to the extent to which human resources policies and procedures are written and inflexible. All statistically significant results supported Hi: Personnel policies are more formalized in more public schools. Generally, the differences between public and charter schools in these areas were significant, with public schools showing more formalization than charter schools. Only one

analysis, salary schedule, indicated significant differences between charter and private schools in these analyses. The following outlines the results for this hypothesis by analysis. Although the results in the second sub-section, which reports analyses with similar responses by public and private schools, were not significant, the lack of difference between public and private schools in those areas is interesting, so they are included in this discussion of results.
Formalization: Analyses in Support of Hi
Summated formalization scales I and II
Desire to adapt evaluation policies
Liberty to diverge from written personnel policies
Written policies for evaluation
Number of rules and procedures
Salary schedule
Formalization: Analyses with public and private schools showing similar levels of
Written guidelines for hiring
Specific criteria for salaries
Formalization: Analyses in Support of Hi
Analyses for the following items supported the hypothesis that personnel policies are more formalized in more public schools. In other words, all of the following indicated that public schools demonstrated more formalization than charter or private schools. The differences between charters and private schools were not significant, however. This discussion begins with results for the summated formalization scales, which combined multiple individual survey items. A discussion of the items contributing to at least one of these scales, and demonstrating

significance independently, follows. Additional analyses supporting this hypothesis conclude this subsection.
Summated formalization scales I and II. Two formalization scales were created to combine multiple items in the area of formalization. Summated formalization scale I combined seven items assessing formalization in areas such as evaluation policies, hiring guidelines, dismissal guidelines, salary criteria, and personnel policy, generally. Higher scores on the scale reflected higher levels of formalization. In analyzing the means with independent t test, public schools (M = 26.69, SD = 3.50, n = 68) demonstrated greater formalization than did charters (M = 22.45, SD = 4.47, n = 80), with the public-charter difference statistically significant, t (145) = 6.475, p < .001 (two-tailed). The means between public and private schools (M = 22.73, SD = 3.68, n = 49) were not significant
This scale was comprised of seven individual survey items that asked respondents the extent to which they agreed with statements addressing the formalization of human resource policies and procedures at the school. The seven items included are listed below, along with the corresponding statements to which respondents were asked to indicate levels of agreement or disagreement. One item required a reversal of polarization, which is indicated by a star (*):
Desire to adapt evaluation policies: I wish I could adapt our teacher evaluation policies to different teachers and circumstances, but the policies dont permit that.
Liberty to diverge from personnel policies*: I have the liberty to diverge from written personnel policies where needed.

Comprehensive and specific guidelines for dismissal: We have comprehensive and specific procedures for dismissing teachers.
Written guidelines for hiring: Our school follows specific written guidelines for hiring new teachers.
Specific criteria for salaries: Our teachers salaries are determined by very specific criteria and formulas.
Inflexible hiring procedures: My schools procedures for hiring teachers are inflexible.
Specific policies for evaluation: The schools written policies for teacher evaluations are very specific and comprehensive.
I also analyzed each of these items individually; descriptions and results for
individual analyses are also included in this results section. A Cronbachs alpha for
this scale assessed the internal reliability. With a score of .558, this scale
demonstrated poor internal reliability.
The hypothesis was further supported by multiple regression analysis. As
explained in the methodology chapter, regressions were run to predict scores on the
summated formalization scale with the following independent variables: school size,
free and reduced lunch eligibility, school sector (using a public sector dummy
variable and charter sector dummy variable), and school location (using dummy
variables for rural and urban locations.) These predictors accounted for about one
fifth of the variance in scores on the summated formalization scale (R2 = .208,
adjusted R2 = .182), which was significant, F (6,187) = 8.166, p < .001. The public
sector dummy variable (B = 3.457, p < .001) was the only variable that demonstrated
significant effects on the scale scores, indicating, that publicness had a strong,

positive effect on the dependent variable representing formalization. Table 4.4, provides a summary of this regression analysis.
Table 4.4. Regression analysis summary for variables predicting summated formalization scale I
Variable B Std. Error Std. Beta
School size 0 0.001 0.010
Free and Reduced Lunch Eligibility 0.004 0.012 0.028
Public Sector Dummy Variable*** 3.457 0.903 0.378
Charter Sector Dummy Variable -0.852 0.849 -0.097
Rural Location Dummy Variable -0.419 0.743 -0.045
Urban Location Dummy Variable
**p < .05, ***p < .01
Due to the poor internal consistency reflected in the Cronbachs alpha analysis, I created a second formalization scale, summated formalization scale II, to improve internal consistency. This scale included fewer items, but demonstrated greater internal reliability in terms of the Cronbachs alpha analysis. With a score of .627, this scale demonstrated questionable internal consistency, so results should still be interpreted with caution.

The summated formalization scale II combined three items assessing formalization in hiring, evaluation, and teacher dismissals. In analyzing the means with independent t tests, charters (M = 12.00, SD = 2.56, n = 83) had means extremely close to private schools (M = 12.02, SD = 2.27, n = 50), but significantly lower than public schools (M = 13.16, SD = 2.05, n = 69), with the public-charter difference statistically significant, t (150) = 3.042, p=.003 (two-sided). The potential range for means was between 5 and 15, with higher scores reflecting higher levels of formalization.
This scale was comprised of three individual survey items that asked respondents the extent to which they agreed with statements addressing the formalization of human resource policies and procedures at the school. The three items included are listed below, along with the corresponding statement respondents were asked about. No items required a reversal of polarization.
Comprehensive and Specific Guidelines for Dismissal: We have comprehensive and specific procedures for dismissing teachers.
Written Guidelines for Hiring: Our school follows specific written guidelines for hiring new teachers.
Specific Policies for Evaluation: The schools written policies for teacher evaluations are very specific and comprehensive.
I also analyzed each of these items individually; descriptions and results for
individual analyses are also included in this results section.

Full Text
On the other hand, respondents generally raised challenges regarding the
formalization of hiring policies with regard to specific policies. For example, two of
the public school respondents, as well as a private school principal with many years
of experience in the public system, raised the issue of policies involving placement of
teachers in schools, such as forced transfers. Charters also raised the issue of new
teacher qualification policies required by the No Child Left Behind legislation. As
these issues constrained a principals autonomy in hiring decisions, they will be
discussed at greater length in that sub-section.
Dismissal. The first hypothesis was also supported by the responses regarding
dismissal policies during the qualitative component of this research. As noted, all of
the public schools reported formal dismissal policies. The public school respondents
typically reported that these policies were very formal, were created at the district
level, and included a strict and lengthy timeline that must be followed. For example,
one public principal noted that it was about a year and a half process that required
strict documentation and specific evidence. Another public school principal
emphasized the specificity in the dismissal process:
You have to follow all the policies.. .you have to be very meticulous about that. You cant say that this teacher has to go, thats just not enough.. .you have to be very meticulous about that.
Like many personnel decisions, public schools reported that these policies were
created by the district, in conjunction with the teachers union. One public principal
also raised the issue of state tenure laws, reporting that their dismissal process was