Exceptional Proff 612

ACCOUNTABILITY AND MICROMANAGEMENT: DECENTRALIZED BUDGETING IN MASSACHUSETTS SCHOOL DISTRICTS DOUGLAS SNOW AIMEE WILLIAMSON Suffolk University

ABSTRACT

The tension between micromanagement and decentralization of decision making is a long-standing theme in the public and education administration literatures. Decentralized budgeting has become a system-changing political strategy to reform education by giving school principals and stakeholders more responsibility and flexibility for allocating resources to activities most likely to improve education outcomes. This study uses survey and archival data to identify and explain patterns of decentralized budgeting in Massachusetts school districts. Factor analysis confirms a political pattern of decentralization where school principals and school level stakeholders gain more influence over budget decisions when they have the ability to originate budget requests and control the disposition of budgeted funds. Political decentralization is accompanied by greater involvement in school level budget decisions by elected school officials. Political decentralization is more likely to be found in smaller, rural, and fiscally stressed districts where federal and state testing mandates have become salient in budget decisions. Factor analysis also confirms a bureaucratic pattern of decentralization where control over transfers among budget accounts is delegated to school superintendents and business officials and, to a much lesser extent, school principals. Wealthier, municipal school districts are more likely to conform to the bureaucratic decentralization pattern. With more financial resources, these districts are more likely to have stronger internal controls and adequate central office staff. Consistent with theory, school districts’ budget environments, controlling for other organizational characteristics, shape patterns of decentralized budgeting. Key Words: Micromanagement, decentralization, school-based

budgeting, budget reform, administrative reform

PAQ SUMMER 2015

221

INTRODUCTION

The purpose of this paper is to identify, describe, and explain patterns of decentralized budgetary control among Massachusetts school districts and thereby contribute to administrative, education, and budget reform theory. The development of strategies to reduce micromanagement has long been a major theme in public administration theory, with special applicability to budgeting (Behn, 1995; Rubin, 2010). As a critical, highly visible administrative function, budgeting is often micromanaged by elected officials in order to reduce the risk of embarrassment from shortfalls, misdirected resources, and waste, as well as to bind managers to policy (Schick, 1966; Rubin, 2010). Decentralization themes generally include strategies to increase efficiency and effectiveness through distribution and/or redistribution of responsibility for management decisions within an administrative hierarchy (Hayek, 1945; Anthony, 1965; Simon, 1976) and strategies to increase managerial creativity, entrepreneurship, and public value (Gulick, 1983; Osborne & Gaebler, 1992; Moore, 1995). These themes are also present in the education literature, although a major focus over the past thirty years has been on strategies to use decentralization as a means to achieve education reform (Brown, 1987). Decentralization reforms, known generally as school-based management and budgeting, stress the importance of citizen, parent, and teacher involvement in school-level management and budget decisions. Decentralization strategies that include external stakeholders in school level decisions are explicit system-changing, political strategies to redistribute decision- making to the school level and thereby solve the problem of micromanagement (Wohlstetter and Buffett, 1991; Goertz & Stiefel, 1998). Reform advocates maintain that school principals, in collaboration with parents, teachers, and community stakeholders, are in the best position to make decisions that will positively affect student performance and that to be effective, the ability to make budget decisions should be aligned with responsibility for educational outcomes. Decentralization does not, however, relieve school boards of their legal responsibilities and of their accountability to the

PAQ SUMMER 2015

222

electorate. While they may be supportive of reform goals in general, elected school board members may view decentralization as a risky proposition. As a result, they may elect to approach decentralization selectively rather than comprehensively (Brown, 1987). Selectivity, in turn, is likely to be shaped by the political, social, and economic environments in which officials make budget decisions (Schick, 1966; Rubin, 1996).

As normative theory, decentralization of budgetary control in any governmental organization, including school districts, raises important questions that can be addressed through empirical research. How much budgetary responsibility, control, and flexibility do elected officials delegate to middle or operating level managers? Does decentralization follow prescribed or unique patterns? How do elected officials ensure that managers with increased budget flexibility will pursue objectives consistent with policy? Does the inclusion of stakeholders and constituencies in decision-making reduce micromanagement? This article addresses these questions using Massachusetts school districts as cases in point. Research methods include correlation analysis, confirmatory factor analysis, and multiple regression. Data are derived from a survey of members of the Massachusetts Association of School Business Officials and from Massachusetts Department of Elementary and Secondary Education data banks.

School districts are local governments with elected boards and professional management. They employ some fifty- six percent of local government employees in the United States and consume twenty percent of total state and local expenditures (U.S. Census Bureau 2006 & 2007). As such, they represent an important, fertile field for public administration research. An examination of decentralization within the context of school districts thus serves as an important link between the fields of education and public administration. Our findings should be of interest to practitioners, policy-makers, and scholars in both public education and public administration. The high profile nature of education policy in national, state, and local politics hardly needs mention. Nevertheless, there is much to be learned

PAQ SUMMER 2015

223

about management practices that cut across the public education and public administration disciplines (Raffel, 2007).

LITERATURE

In his classic book, Planning and Control Systems,

Robert Anthony (1965) explores the distribution of decision- making authority through a framework consisting of three basic management functions: strategic planning, management control, and operational control. Following Anthony’s framework, responsibility for strategic planning should be located at the executive and policy levels of organizations. The strategic goals that emerge from the planning process shape the development of programs and short-term objectives. Line managers working at the lowest levels of the hierarchy are responsible for directing the routine activities of programs. Anthony refers to this function as operational control. As applied to governments, management control is a process that involves continuous interplay among elected officials, top administrators, and mid- level managers in assuring that program resources and objectives are aligned with overall strategic goals (Anthony & Young, 2007). Anthony, having served as Defense Department controller in the mid-1960s, was influential in the implementation of Planning, Programming and Budgeting systems which has, in turn, influenced more recent developments and prescriptions generally associated with new public management theory. For example, responsibility budgeting, a practice which falls within the PPB tradition, aligns responsibility for budgetary decisions with responsibility for program management and goals (Jones & Thompson, 1986; Barzelay & Thompson, 2003). Giving managers control over resources so that they can achieve goals as they see fit is also an important value in the reinventing government movement and in new public management theory (Osborne & Gaebler, 1992; Barzelay, 2001).

School reform prescriptions, most notably school-based budgeting, emphasize that principals, either implicitly or explicitly, are accountable for educational results. It follows that the ability to direct resources where they are most needed should

PAQ SUMMER 2015

224

fall to the person most familiar with a school’s needs, the principal. The individual school is conceptualized as a responsibility center, an appropriate unit of analysis in responsibility budgeting. To the extent that control over budget decisions remains within the school district’s hierarchy, school- based budgeting, as it is called in the education literature, is consistent with management control and responsibility budgeting theory (Barzelay & Thompson, 2003; Anthony & Young, 2007). When responsibility budgeting includes parents, citizens, and teachers in budget decisions, however, school-based budgeting takes on a political dimension (Goertz & Stiefel, 1998).

School-based Management and Budgeting School based-budgeting, the prescriptive education reform most relevant to this research, generally accompanies school-based management in the education literature. Definitions of each are generally quite broad but include strategy and policy, operational control, and budgeting (Caldwell, 1993; Hansen & Roza, 2005; Wohlstetter & Van Kirk, 1995). School- based budgeting is generally considered integral to school-based management (Candal, 2009; Cuban, 2007; Hadderman, 2002; Hansen & Roza, 2005; Kedro, 2003). Brown (1987) considers school-based budgeting to be the primary component of school- based management. More broadly, Roza (2008) asserts the importance of a good fit between budgeting methods and reform strategy. A comprehensive approach to school-based budgeting could include: • Control over all school level expenditures for staffing,

including the ability to determine the number and distribution of teachers, aids, clerical, custodial and maintenance employees;

• Control over purchasing instructional and building maintenance supplies and services,

• Procurement of professional development services (see Wohlstetter & Buffett, 1991; Goertz and Hess, 1998).

Implementing all of these features would be a dramatic change for most school districts. In traditional school districts, staff positions and assignments are controlled centrally and principals control only a few small budget items for instructional, office,

PAQ SUMMER 2015

225

and building supplies. School level control may also include accounts holding funds raised by parent-teacher organizations (Wohlstetter & Buffett, 1991).

As reforms, school-based management and budgeting are often prescribed as “system changing strategies” to shift the balance of power inside school districts (Wohlstetter & Buffett, 1991). Although the findings are mixed, supporting research suggests that decentralization improves student performance through stronger employee commitment and motivation, increased parent participation, and a reduction of bureaucratic barriers (Bandur, 2012). Referring to these reforms in the Chicago public schools, Goertz and Stiefel write: “The theory underlying this ‘democratic localism’ is that expanded local participation will create a political force for improvement, leveraging and sustaining the organizational and instructional changes needed to raise student achievement” (1998, p. 2). Local participation generally takes the form of a school council made up of parents, teachers, and other citizens, thus integrating stakeholder influence with responsibility budgeting. Empowering school-level stakeholders would presumably change the nature of control exercised by the central office and alter the ways that elected officials—school boards—exercise control. If this is not the case, one has to ask, “Why bother?” (Wildavsky, 1992, p. 594).

While any change in decision-making authority within an organization can change power relationships, the extension of influence to stakeholders at the school level allocates a measure of political accountability to individual schools, without, presumably, reducing the accountability of elected school board members. This distinguishes decentralization that takes place within the school district’s administrative structure from decentralization that empowers school constituencies in making budget decisions. Both types address the problem of micromanagement, but in very different ways. In a lengthy monograph on decentralization, Brown (1987, p. 5) identifies distinct bureaucratic and political types. Bureaucratic decentralization may occur within the organization and can be “revised” or “recalled” without requiring legislative action (Brown 1987, p. 5). In the political type, however, “groups of

PAQ SUMMER 2015

226

the public are given power to make decisions” (1987, 5). School-based budgeting prescriptions often include school councils, composed of teachers, parents, or community representatives, in school-level budget decisions (Goertz & Stiefel, 1998). The inclusion of school-level stakeholders in decision-making goes beyond bureaucratic forms where responsibility is distributed among employees. To state the obvious, the parents and community representatives on these councils are voters, not employees. Their role, as commonly prescribed, goes well beyond traditional parent-teacher organizations. The creation of school councils, by design, is intended to extend responsibility and accountability for school performance to stakeholders.

Barriers to Budget Decentralization

The school-based management and budgeting literature identifies a gap between theory and practice (Wohlstetter & Van Kirk, 1995). Reasons for the theory-practice gap include legitimate concerns of school board members, opposition of administrators, lack of training for principals, weak internal controls, public pressure for results measured under the No Child Left Behind Act of 2001 (NCLB), the potential for political conflict when parents and citizens are given budget influence at the school level, the need to centrally manage collective bargaining agreements, and fiscal stress. Perhaps the most significant of these reasons are the legal and political responsibilities of school boards which may lead them to delegate budgetary control to principals selectively rather than comprehensively (Brown 1987, p. 9). Alternatively, selectivity in decentralization may be due to a lack of political will, as observed by Moynihan (2009) in a study of administrative reform in state government. The result is a “constrained model” of decentralization, half measures that have the form of decentralization without the substance (Moynihan 2009, p. 165).

In writing about federal budget reforms in the 1960s, Allen Schick (1966) identifies weak internal controls as a reason for micromanagement. As applied to school districts, these controls are critical to assuring the effective use of school assets. Poor or ineffective controls may increase the risk of misuse of

PAQ SUMMER 2015

227

funds and assets for which school board members, and their appointees, are responsible. To reduce this risk, school boards may limit the discretion of principals in the use of funds and physical assets. Related to internal controls, Schick also notes that top level officials may lack confidence in the abilities of middle and operating level managers to execute stated policy, which, in turn, may be related to training and/or resources for organizational development (Clover, Jones, Bailey & Griffin, 2004). In order to “bind” lower level managers to stated policy, central control may extend to all or some aspects of budget decisions, including planning and programming, budget preparation, implementation, the annual audit, control over the number of authorized positions, transfers among accounts, and control over travel and purchasing (Schick, 1966, p. 244). In a similar vein, Irene Rubin (1996) suggests that municipal governments may increase central control over functions where public concern for poor performance is highest and, given scare administrative time and resources, reduce the attention paid to other functions. For example, the No Child Left Behind Act focuses accountability for student learning outcomes at the school level. Test scores are publicly reported and generate considerable public discussion. Decentralization of budgetary decision making could be a high stakes proposition for elected school boards and superintendents who lack confidence in the ability of principals and school councils to direct resources to those activities that will assure higher test scores. The extra attention paid to test scores could come at the expense of other education priorities. Existing research does explore the impact of standardized test scores on school budgeting, but it is recognizably limited (Chiang, 2009; Craig, Imberman, & Perdue, 2013) and focuses more on changes in allocation amounts than on budgetary reform strategies. Both Chiang (2009) and Craig, Imberman, and Perdue (2013), for example, find increases in school budgets resulting from the threat of sanctions and a reduction in accountability ratings, respectively. Such findings are not always consistent, however, even within studies, nor are budget increases permanent (Craig, Imberman, & Perdue, 2013).

PAQ SUMMER 2015

228

Selective micromanagement, the flip side of selective decentralization, is one way for school boards to exercise control over some types of budget decisions that are, at any given time, most salient. Both Schick (1966) and Rubin (1996) refer to this phenomenon as a “budget orientation.” The concept stems from the idea that governmental organizations typically lack the resources to perform all centralized functions equally well, which then leads them to orient themselves to those types of control most responsive to the environment in which budget decisions must be made. The orientations most relevant to this research are control (Schick, 1966) and accountability (Rubin, 1996).5

The involvement of school-level stakeholders in making budget decisions is a common reform prescription that generally accompanies school-based budgeting. School councils create a new level of responsibility and political accountability at the school level. These councils may present political risks to school boards by creating or exacerbating political conflicts within a district. Marlowe and Portillo (2006) suggest that an expansion of citizen participation in a local government could “exacerbate its pluralistic tendencies and inhibit compromise” (p. 180). While acknowledging the benefits of school-based budgeting, Fermanich, Odden, and Archibald (2000) observed increased conflict in a large district with school-based budgeting when economic conditions required budget reductions. Newly influential school councils and the teacher union represented new constituencies that had to be consulted in the budget cutting process. School boards may be justifiably reluctant to open up such conditions through decentralization.

Another important force that may drive micromanagement is responsibility for managing collective bargaining agreements. Since bargaining agreements cover a wide range over an entire district, decentralization of many types of management and budgetary control may not be feasible. State education department rules and regulations applying to the use of

5 Schick (1996), who coined the term budget orientation, also identifies planning and management orientations. Rubin (1996) also identifies a “prioritization orientation” linked to the salience of cutback management. Thurmaier and Gosling (1997) identify a “policy orientation” in state government budget offices.

PAQ SUMMER 2015

229

funds and purchasing could also be centralizing forces (Goertz & Hess, 1998). Fiscal stress may act alone or with other barriers to decentralization (Forrester & Spindler, 2001; Lu & Facer, 2004). Fiscal stress may force local government officials to reallocate their time and energy to reprioritization of their programs at the expense of other management activities. A major reprioritization is likely to be a centralizing force, especially if cuts involve whole programs or other organizational units, such as the need to close a school. The Massachusetts Legal Framework Massachusetts provides a particularly useful case to link public administration and education literatures. Massachusetts law delegates school district governance to elected school committees (with the exception of the City of Boston School Committee, which is appointed by the Mayor, Chapter 108, Acts of 1991).6 Statute defines a school district as either the school department of a municipality or a regional entity (Massachusetts General Laws 70: §2). All non-charter school districts are fiscally dependent on municipalities in Massachusetts. State school aid flows to municipalities first, then to school districts. Of 328 non-charter school districts in Massachusetts, 243 are municipal districts, which account for 85 percent of enrollment. School districts compete directly with other municipal departments for any dollars beyond the state mandated level of spending and for a place on the ballot for proposed property tax increases. Unused funds generally lapse to the municipality at the close of the fiscal year. The school finance system is based on the creation of foundation budgets for each school district, which are driven by enrollment and standardized costs that are adjusted for different labor markets within the state. State education aid makes up the difference between the foundation budget and local tax effort, which is computed by imputing a tax on both property wealth and, in recent years, adjusted gross income. Local tax effort is enforced by way of a minimum spending requirement. School districts may exceed the minimum spending requirement, but may not spend less 6 School district governing boards are called “school committees” in Massachusetts.

PAQ SUMMER 2015

230

except under special circumstances (Massachusetts General Laws 70: §11). Once the annual school finance plan becomes law, school districts may allocate funds within the district as they see fit (Massachusetts General Laws 70: §8), but in compliance with federal and state restrictions and union contracts. School committees do not have the final say on the revenue side of budgets, but must submit funding requests to the appropriate municipal body—either town meeting or city council—for approval (Massachusetts General Laws 71: §34). Regional districts must request approval of each member municipality’s share of the district’s budget. During budget development, school committees must allow input from local school councils (Massachusetts General Laws 71: §59C) and hold public hearings on the proposed budget (Massachusetts General Laws 71: §38N). School committees determine the amount of budget flexibility allowed to superintendents, business officials, and principals. While state law places a great deal of responsibility on school committees and establishes fiscal dependence on, and thus, political accountability to, municipal governments, school reform in Massachusetts has also sought to encourage school committees to decentralize administration to the school level. State law makes principals “the educational administrators and managers of their schools” with responsibility for “the operation and management of their schools and school property”, but subject, of course, “to the supervision and direction of the superintendent” (Massachusetts General Laws 71: §59B). The Massachusetts Education Reform Act of 1993 (MERA) stipulated the creation of school councils, whose members are parents, teachers, community members, and (in the case of high schools) students. MERA also decentralized decision-making authority in the areas of hiring and dismissing teachers from school committees to superintendents and principals (Anthony & Rossman, 1994; McDermott et al., 2001). There is, however, no requirement to delegate budget control to principals. Perhaps the most visible and salient feature of Massachusetts education reform is the Massachusetts Comprehensive Assessment System (MCAS), a set of exams that measure student, school and district level performance at

PAQ SUMMER 2015

231

specified intervals. Students must pass the 10th grade MCAS exams to graduate from high school. MCAS scores are reported to parents and in the media and have become both visible and controversial. The MCAS system is used by Massachusetts to comply with the assessment requirements of NCLB, although the MCAS system predates NCLB.

RESEARCH METHODS The extensive school reform and public management reform literatures and the Massachusetts legal environment, all of which have been part of the Massachusetts education environment for twenty years, suggest that school districts should be very much aware of decentralized decision making, broadly, and school-based budgeting prescriptions, specifically. The literature suggests that decentralization may follow bureaucratic and/or political patterns. Where decentralization exists, one would expect to observe these patterns in practice.

Hypothesis 1: Decentralized budget practices will follow distinct bureaucratic and political patterns.

Stakeholders—citizens, parents, and teachers— that are involved in budget decisions are more likely to be influential in districts where decentralization follows a political pattern. Statute gives a role in budget decisions to school level councils in Massachusetts and school-based budgeting generally prescribes such a role. These councils should be more influential in budgetary decision making in districts where principals have the ability to originate a budget request and exercise some control over the use of budgeted funds. And, by definition, one would expect principals to have more influence over budget decisions under the aforementioned circumstances.

Hypothesis 2: Principals and stakeholders will be more influential in the budget process in districts with higher levels of political decentralization.

Little is known about how school district characteristics and budget environments interact in shaping decentralization. Following Rubin’s (1996) observations about the centralizing

PAQ SUMMER 2015

232

effects of fiscal stress and public dissatisfaction with municipal performance, one would expect school districts to respond to both by centralizing control. On the other hand, reformers view decentralization as a way to unleash the knowledge and skills of principals and thereby improve both efficiency and educational outcomes. Schick (1966) suggests a linkage between weak internal controls and centralization of budget decisions. This could be especially true in two cases: 1) pre-auditing account balances prior to authorizing a transfer between accounts and 2) assuring that there is a budgeted position available before hiring a new employee. Internal controls could be weaker in school districts with lower fiscal capacity. This could be less important, however, in smaller districts where it is easier for central office officials to oversee fewer schools. The variables selected to measure individual school district characteristics and budget environments may have positive or negative effects on decentralization, as indicated in the above examples. These variables could also have differential effects on bureaucratic and political decentralization. Therefore, Hypothesis 3 must be expressed in broad terms.

Hypothesis 3: Decentralization is shaped by school district characteristics and the budget environment.

Data A survey administered to school business officials in April and May of 2009, with the sponsorship and support of the Massachusetts Association of School Business Officials (MASBO), provided most of the data for this study. Development of the survey instrument was informed by the literature and by pilot interviews with business officials in two urban, two suburban, and two rural districts. These data were supplemented by archival data available from the Massachusetts Department of Elementary and Secondary Education. Business official is the generic job title used by the association, but actual job titles vary and include, naming a few: Chief Financial Officer, Associate Superintendent for Human Resources and Finance, Director of Finance, and Director of Administration and Finance. Excluding respondent information, the survey

PAQ SUMMER 2015

233

included 162 items. Other data sets extracted from the survey have been employed in other research by the authors (Williamson & Snow, 2013, 2014). Charter schools were not included in the study since most of them are small, operate independently of the traditional school districts, and are already decentralized, by definition. Survey respondents represent 179 public school districts, 73 percent of the public, non-charter school districts with a MASBO member and 55 percent of the 328 Massachusetts non-charter public school districts. 7 Responding school districts are representative of all Massachusetts school districts in terms of enrollment, spending per student, and reliance on state aid. Municipal and regional districts are represented in the same proportions among responding districts and all others. The eastern part of the state is somewhat overrepresented among survey respondents: 64 percent of responding districts are located in Eastern Massachusetts compared to 56 percent of all other districts. The largest district in the state, Boston, which is an outlier in terms of both its size and governance, did not respond to the survey. However, urban schools are represented. Responding districts include the Worcester and Springfield municipal districts. Respectively, they are the second and third largest cities in the state with estimated 2009 populations of 182,000 and 156,000 and school enrollments of 27,000 and 24,000. More detail comparing responding districts with all districts can be found in the appendices.

These data collection methods bring both strengths and limitations. Although business officials offer the perspective of individuals uniquely positioned to participate and observe both financial decision-making processes and operations, their perspectives may be limited by their training and values.

7 The unit of analysis in this study is the individual school district, as viewed from the business official’s perspective. The response rate for individual MASBO members differs from representation of individual school districts in the data set. The survey generated 192 responses from 294 individual MASBO members who were invited to participate in the survey, a response rate of 65 percent. In the twelve cases where there were two responses from a school district, the response of the highest-ranking respondent, based on job title, was retained. One charter school business official who had been inadvertently invited to participate in the survey was also dropped from the data set. This left 179 usable surveys.

PAQ SUMMER 2015

234

Despite this potential bias, business officials are in an ideal position to report on financial and budget practices, and to provide responses to the wide range of questions in the survey. They play a leading role in financial and operations management and typically report directly to the superintendent. They interact with a wide range of other actors in the district, the State Department of Elementary and Secondary Education, and municipal government. School business officials have the “duty as stewards of the public’s money to make school finance more accessible, transparent, and easily understandable” (Willis, 2010, p. 17). Superintendents and school business officials have been described as “mutually interdependent” actors with varying responsibilities—superintendents bring leadership to the process, while school business officials bring financial expertise (Bird & Wang, 2010, 11). Furthermore, the focus on school business officials has helped generate the high response rate. The single- state focus also provides a natural control for the legal and policy environment. Like all states, however, Massachusetts must comply with federal education laws and regulations.

Selection of Variables and Measurement To identify patterns and types of decentralization among responding school districts, selected survey items focus on budget practices that 1) place greater control over budget decisions in the hands of principals, superintendents, or business officials, 2) increase controls exercised by elected school committees, and 3) increase the influence of stakeholders. Three survey items measure decentralization of budgetary decision making to school principals by asking business officials whether principals may: 1) formulate a complete budget request, including all grades, programs, salary, and non-salary expenses; 2) exercise direct control over all budgetary accounts related to their building or site, both non-salary and salary accounts; and 3) move (transfer) funds among budgetary accounts, without approval of the district business office. When combined, these practices constitute a substantial amount of budgetary responsibility in the formulation, adoption, and monitoring phases of the budget process. However, this study makes no attempt to specifically identify the existence of school-based

PAQ SUMMER 2015

235

management or budgeting as comprehensive management strategies or policies since definitions in the literature lack consistency and precision. To account for micromanagement by school committees, the study also includes three survey items that identify budget decisions that pilot interviews suggested could be exercised by school committees: 1) Control over transfers among budgetary accounts, 2) involvement in the details of developing school- level budgets and 3) control over the number of administrative, faculty, and staff positions. None of these items has a direct opposite among the three types of decision making exercised by principals. A school committee’s decision not to exercise a type of centralized control does not mean that control has been delegated to principals. Decentralization of budget decisions, as a system- changing, political strategy, should alter administrative relationships and the relative influence of actors in making budget decisions (Wohlstetter & Buffet, 1991). Survey items identify business officials’ perceptions of the influence of school councils, principals, teachers, and parents and citizens in general over budget decisions. Their influence is expected to be higher in districts where budgetary decisions are delegated to principals. The inclusion of teachers, citizens, and parents on school councils tends to make them representative in nature. Therefore, we include variables for stakeholders jointly, in the form of school councils, and individually. Every school should have a school council, as required by statute. Without the ability to originate a budget at the school level and some level of control over accounts once the budget is approved, however, one would not expect the council to be very influential in making budget decisions. The literature also suggests the possibility of greater union influence under decentralization (Straut, 1996). We included business officials’ perception of the influence of unions in budget decisions in our analysis.

PAQ SUMMER 2015

236

Table 1 Descriptive Statistics: Exercise of Budgetary Control Survey' Item' (1' =' Strongly'Disagree,' 2' =' Disagree,' 3' =' Neutral,' 4' =' Agree,' 5' =' Strongly'Agree)' N' Mean'

Std.' Dev.'

%'Agree'or' Strongly' Agree'

Principals' develop' and' submit' a' complete'budget'request' 160' 2.81' 1.31' 38' Principals' have' direct' control' over' all' budgetary'accounts' 160' 2.45' 1.19' 23' Principals' may' move' funds' among' budgetary'accounts' 160' 1.82' 1.04' 13' The' school' committee' retains' detailed' control'over'transfers' 163' 3.42' 1.19' 55' The'school'committee'is'involved'in'the' details' of' developing' schoolNlevel' budgets' 162' 3.06' 1.23' 44' The' school' committee' retains' detailed' control' over' the' number' of' administrative,' faculty,' and' staff' positions' 163' 3.19' 1.25' 52' The'superintendent'has'the'flexibility'to' make' transfers' among' budgetary' accounts' 163' 3.64' 1.24' 67' The' superintendent' delegates' control' over' minor' transfers' to' the' business' office' 162' 3.23' 1.33' 52'

Table 2 Descriptive Statistics: Influence of Stakeholders in the Budget Process Survey' Item' (1' =' No' Influence,' 2' =' Very' Little' Influence,' 3' =' Moderate' Influence,' 4' =' A' Great'Deal'of'Influence)' N' Mean'

Std.' Dev.'

%' Moderate' Influence'

%'Great' Deal'of' Influence'

Principals'' influence' in' the' budget'process' 158' 3.22' 0.66' 55.1' 34.2' School' councils’' influence' in' the'budget'process' 158' 2.27' 0.65' 36.7' 0.6' Union's'influence'in'the'budget' process' 158' 2.19' 0.74' 29.1' 3.2' Teachers'' influence' in' the' budget'process' 159' 2.45' 0.62' 48.4' 1.3' Parents''influence'in'the'budget' process' 158' 2.37' 0.67' 38.0' 3.2' Citizens'' influence' in' the' budget'process'' 159' 2.13' 0.65' 24.5' 1.3'

Note: This survey item was adapted from similar items on the US Department of Education’s Schools and Staffing Survey, Public School Principal Survey.

PAQ SUMMER 2015

237

To identify potential environmental determinants of decentralization patterns, this study includes three scales constructed from survey items and archival data: 1. MCAS/NCLB Budget Influence: A scale of five survey

items that measure business officials’ perceptions of the extent to which standardized exams required under both the Massachusetts Comprehensive Assessment System (MCAS) and No Child Left Behind Act (NCLB) influence budget decisions (Cronbach’s Alpha = .710)

2. Wealth Index: Composed of the mean of FY 2009: 1) school district spending as a percentage of the state minimum requirement, 2) the percentage of mandated spending from own source revenue, and 3) the percentage of students who are not eligible for free and reduced lunches (Cronbach’s Alpha = .738)

3. Fiscal Stress Index: The sum of “yes” responses (yes is coded “1”) to twenty-four survey items asking whether the district has experienced fiscal challenges in recent years, during the survey year, or expects to experience fiscal challenges over the next few years.

The influence of MCAS/NCLB mandated exams on budget decisions is a potential centralizing force. Rubin (1996) suggests that a salient issue may cause the public to demand greater accountability for its performance in that domain. Higher scores on this index, which indicate greater importance of NCLB/MCAS in budget decisions, could be the result of public demand for improved MCAS performance. Our exploratory analysis indicates that this index is weakly and negatively correlated with actual performance on MCAS exams (-.220). Higher scores on the index are driven by more than low test performance. Instead, higher MCAS/NCLB index scores may represent local aspirations or even concerns about property values, since shopping for a home based on the test scores of a community’s schools has been documented in Massachusetts (Bradbury, 1998). The wealth index is a proxy for fiscal capacity and based on the premise that wealthier districts have the resources to train principals (Clover, Jones, Bailey & Griffin, 2004) and invest in stronger internal controls (Schick, 1966). Fiscal stress is a potential centralizing force in that districts

PAQ SUMMER 2015

238

facing fiscal stress must orient budgetary decision making in the direction of reprioritization in order to balance their budgets (Rubin, 1996). As previously noted, however, these variables could easily have the opposite effects if one adopts a school reform perspective. They could also have differential effects on bureaucratic and political decentralization patterns. Details of scale construction are included in the appendices. Two additional variables measure progressively more severe sanctions imposed upon districts that fail to meet state determined performance targets required under NCLB. Meeting these performance targets is called adequate yearly progress (AYP). The failure of one or more schools to make adequate yearly progress after one year requires districts to allow parents in the attendance zone of a school that failed AYP to attend another school in the district (School Choice) and, if AYP is still unmet the following year, offer supplemental educational services (SES). To account for potentially different effects of these sanctions, we constructed separate regression models for School Choice and Supplemental Educational Services (Williamson & Snow, in 2014). Actual performance on MCAS exams has been excluded from the regression models due to colinearity with the wealth index (Pearson’s R = .720). We also included three characteristics of Massachusetts school districts as potential determinants of decentralization patterns: the number of schools, type of district (municipal or regional), and rural location. School committees with fewer schools to administer may be more likely to micromanage budgeting simply because of the narrower span of control. A dummy variable for municipal districts (regional district is the dropped category) accounts for potential differences in political accountability. Municipal districts are contained within a single political entity and are directly accountable to a single municipal, government. Regional districts are accountable to more than one municipality. A dummy variable was also created for rural districts where different political cultures and economic bases may be important to decentralization decisions. We excluded total district enrollment since it is collinear with the number of schools (Pearson’s R = .954). We also excluded dummy variables for urban and suburban schools because they are both

PAQ SUMMER 2015

239

collinear with the wealth index (Pearson’s R = - .587 and .502, respectively). Data Analysis

Data analysis was conducted in two stages. The first stage consisted of an analysis of zero order correlations. The decentralization, micromanagement, and budget influence variables listed in Tables 1 and 2 were included. To confirm the political and bureaucratic decentralization patterns observed in the correlation analysis and to obtain dependent variables for regression, confirmatory factor analysis was employed to extract two factors with the expectation that the results would follow the hypothesized political and bureaucratic decentralization patterns suggested by the literature. After conversion to standard scores, items with statistically significant correlations with one or more of the three types of budgetary control that could be delegated to school principals were included in the confirmatory factor analysis. This methodology is appropriate because the literature suggests that there are two distinct forms of decentralization, political and bureaucratic, as well as a critical distinction between the two, the influence of stakeholders in school-level decisions. The analysis is driven by theory as to the types and structure of the hypothesized latent variables, political and bureaucratic decentralization (Thompson, 2004). The extracted variables were retained in the data file as dependent variables for regression analysis. In the second stage of the analysis, the saved factor scores were regressed on measures of the budget environment and other school district characteristics.

FINDINGS

Frequencies Survey responses indicate substantial variation in the frequency with which various practices are decentralized. Among the budget decisions delegated to school principals, budget preparation is the practice with the highest frequency of responses in the agreement range (38 percent), followed by control over all budgeted accounts (23 percent), and ability to make transfers among budgeted accounts (13 percent). None of the business officials strongly agreed that all three practices

PAQ SUMMER 2015

240

applied in their districts. In sharp contrast, school committees exercise a substantial amount of control. Based on answers in the agreement range, a majority of business officials indicate that school committees retain the authority to approve transfers among budget accounts (55 percent) and exercise control over the number of administrative, faculty, and staff positions (52 percent). A large minority of respondents (44 percent) agree that the school committee is involved in the details of developing school-level budgets. In terms of influence in the budget process, 34 percent of respondents indicated that principals have a great deal of influence in the budget process. School councils, unions, teachers, parents, and citizens are much less influential in the budget process, as seen in Table 2, above. While survey items asking about the relative influence of school committees, superintendents, and business officials on budget decisions are not included as variables in our factor analysis (which was limited to the influence of stakeholders at the school level) we note that these actors are, as expected, very influential. Variability only occurs in the top two categories on the four point scale of no influence, very little influence, moderate influence, and a great deal of influence. For superintendents, 94 percent of responses fall in the “great deal of influence” category. Regardless of the amount of decentralization or school committee micromanagement, these actors exercise a great deal of influence in budget decisions. Bivariate Correlations Correlation analysis was conducted to 1) identify significant zero order correlations between each of the variables measuring delegation of control to principals, micromanagement by school committees, superintendents and business officials’ control over transfers, and stakeholder influence in the budget process and 2) to identify individual variables likely to be components of latent political and bureaucratic forms of decentralization (Table 3). The correlations suggest that decentralization follows both political and bureaucratic patterns, although the political pattern does not follow the ideal type identified in the literature. In the political pattern, the influence

PAQ SUMMER 2015

241

of stakeholders correlates positively with greater control exercised by principals in the areas of budget preparation and control over the use of funds. However, correlations also suggest increased micromanagement by school committees when these two types of control are delegated to principals. School committees appear to increase their involvement in school level decisions under decentralization, a potential effort to establish school level accountability to the school committee that will be explored further in the “Discussion” section, below. Components of the political decentralization variable are suggested by the significant positive correlations between two decentralization variables,

• principals develop and submit a complete budget request and

• principals have direct control over all budgetary accounts,

and each of the following variables: • principals' influence in the budget process, • school councils’ influence in the budget process, • teachers' influence in the budget process (significant for

the budget request variable only), • school committee retains detailed control over transfers, • school committee is involved in the details of developing

school-level budgets. Correlations also suggest a bureaucratic decentralization pattern that involves school committees’ delegation of authority to make transfers to superintendents. Business officials and principals also gain more control over transfers when school committees delegate control over transfers. Transfer control is also positively correlated with school councils’ influence. Components of this bureaucratic decentralization factor are suggested by the significant correlations between principals may move funds among budgetary accounts and:

• school committee retains detailed control over transfers (negative),

• superintendent has the flexibility to make transfers among budgetary accounts (positive),

PAQ SUMMER 2015

242

• superintendent reserves the right to approve major transfers but delegates minor transfers to business office (positive).

Table 3 Inter-Item Correlation Matrix

Delegated to School

Principals Micromanaged by School

Committee

Survey Item 1 2 3 4 5 6

1 Principals develop, submit complete budget request

1 .599** .016 .245** .304** -.027

2 Principals control all budgetary accounts .599 ** 1 -.008 .200* .333** -.021

3 Principals may move funds among accounts .016 -.008 1 -.223 ** -.072 -.087

4 School committee controls transfers .245 ** .200* -.223** 1 .314** .345**

5

School committee involved in details of developing school budgets

.304** .333** -.072 .314** 1 .291**

6

School committee retains control over number of admin., faculty, staff positions

-.027 -.021 -.087 .345** .291** 1

7

Superintendent has the flexibility to make transfers among accounts

-.036 -.025 .188* -.446** -.026 -.143

8

Superintendent delegates minor transfers to business office

.025 .036 .269** -.336** .053 -.010

9 Principals' influence in budget process .211 ** .201* .032 .116 .092 .126

10 School councils’ influence in budget process

.207* .170* -.130 .205* .235** .252**

11 Unions’ influence in budget process -.024 -.030 .008 -.057 .037 .196 *

12 Teachers' influence in budget process .160 * .101 -.068 .142 .234** .163*

13 Parents' influence in budget process .123 -.011 .088 .021 .095 .137

14 Citizens' influence in budget process .083 .033 .033 -.010 .120 .082

**Correlation is significant at the 0.01 level (2-tailed). *Correlation is significant at the 0.05 level (2-tailed). Listwise N=154: Columns are limited to the first six variables. The full fourteen-column correlation matrix is available from the authors.

PAQ SUMMER 2015

243

Position Control School committee control over administrative, faculty, and staff positions (position control) is the odd man out in this analysis. It is practiced by a majority of school districts and not correlated with other forms of decentralized budgetary control. Position control exercised by the school committee is, however, positively and significantly correlated with other forms of control exercised by school committees (transfers, .345; involvement in school-level budgets, .291). Position control has weak but statistically significant correlations with teacher union influence (.196), teacher’s influence (.163), and school council influence (.252). While position control is correlated with stakeholder influence, there is no correlation with the most commonly decentralized budget practices. It may be that stakeholder influence is high in some districts where budgeting is not decentralized. Confirmatory Factor Analysis Factor analysis of the statistically significant variables, summarized in Table 4, confirms the decentralization patterns suggested by the literature and observed in the analysis of zero order correlations. Pattern and structure coefficients (“loadings”) greater than .4 are reported. As hypothesized, decentralized budget practices follow distinct political and bureaucratic patterns (Hypothesis 1) and stakeholder influence increases with political decentralization (Hypothesis 2). The first factor extracted, which explains 27.75 percent of the variance among the included variables, confirms a latent political decentralization variable with which school committee involvement in school-level budgets, decentralization of budget development and control to principals, and influence of principals, teachers, and school councils in the budget process have positive loadings. The extracted factor scores were saved for regression analysis. The new variable is named “Political Decentralization.” Higher political decentralization scores indicate greater budgetary control at the school level, greater stakeholder influence, and increased school committee micromanagement.

PAQ SUMMER 2015

244

The second factor extracted explains 19.6 percent of the variance among the included variables and confirms a latent bureaucratic decentralization variable with which decentralization of control over transfers to superintendents and business officials as well as control over transfers exercised by school principals have positive loadings and control over transfers by school committees has a negative loading. These factor scores were also saved for further exploration. The new variable is named “Bureaucratic Decentralization.” Higher bureaucratic decentralization scores are indicative of less control over transfers by school committees and greater control over transfers by superintendents, business officials, and, to a lesser extent, by principals. Delegation of control over transfers to principals is neither widely practiced nor significantly correlated with the other two types of budgetary control exercised by principals that were measured in the survey, while transfer control by school committees is positively correlated with other types of school committee control. Determinants of Political and Bureaucratic Decentralization Regression analysis was performed to explore potential explanations of the variance in political and bureaucratic decentralization factor scores. Descriptive statistics for the regression variables are summarized in Table 5. Regression coefficients, reported in Tables 6 and 7, reveal distinct differences in the significant determinants of political and bureaucratic decentralization. Political decentralization scores are higher in school districts where the performance requirements of the No Child Left Behind Act and the Massachusetts Comprehensive Assessment System are more important to budget decisions, among districts that are not under a school choice sanction, where fiscal stress is greater, in rural districts, and in districts with fewer individual schools, i.e., a smaller span of control. Bureaucratic decentralization scores are higher in wealthy, municipal school districts.

PAQ SUMMER 2015

245

Table 4 Pattern and Structure Coefficients (Oblique Rotation)

Survey Item Political Decentralization

Bureaucratic Decentralization

Pattern Structure Pattern Structure School councils’ influence in the budget process .650 .671 Principals develop and submit a complete budget request .681 .667 Teachers' influence in the budget process .636 .649 Principals have direct control over all budgetary accounts .654 .637 Principals' influence in the budget process .617 .619 The school committee is involved in the details of developing school-level budgets .574 .572 The superintendent delegates control over minor transfers to the business office .891 .876 The superintendent has the flexibility to make transfers among budgetary accounts .883 .874 The school committee retains detailed control over transfers -.575 -.619 Principals may move funds among budgetary accounts .473 .471 Variation Explained 27.747 19.601 Total Variation Explained 27.747 47.348 Mean 0 0 Standard Deviation 0 0 Skewness 0 -.311 Kurtosis .202 -.631 Notes: Listwise n = 155. Oblique rotation selected due to significant correlations across variables in each of the two hypothesized types (Thompson 2004, 48). Correlation between political and bureaucratic decentralization = -.142. Loadings less than .4 omitted (absolute value). Kaiser-Meyer-Olkin Measure of Sampling Adequacy = .667; Bartlett's Test: Approx. Chi-Square = 424.488, df = 45, p < .001.

PAQ SUMMER 2015

246

Table 5 Descriptive Statistics: Regression Variables Variable Mean Std. Dev.

Dependent Variables: Political Decentralization .03 1.02 Bureaucratic Decentralization -.036 1.01

Budget Environment:

MCAS/NCLB Budget Influence 3.65 .66 Wealth Index 9.51 4.77 Fiscal Stress Index 85.48 13.11 School Choice .62 .49 Supplemental Educational Services .42 .495

School District Characteristics: Number of Schools 6.25 6.34 Municipal School District .28 .45 Rural School District .71 .45

N = 122 (Listwise): Political and Bureaucratic dimension descriptive statistics vary from Table 4 because of missing values, which reduces the cases available for regression analysis. Table 6 Standardized Regression Coefficients: Political Decentralization

Model 1 Model 2

Beta t Beta t

Budget Environment: MCAS/NCLB Budget Influence .220** 2.544 .222** 2.535 Wealth Index .-080 -.840 -.062 -.643 Fiscal Stress Index .150* 1.754 .149* 1.706 School Choice -.156* -1.832 Supplemental Educational Services .013 .150

School District Characteristics: Number of Schools -.232** -2.332 -.210** -2.066 Municipal School District .051 .555 .043 .459 Rural School District .177* 1.974 .189** 2.077 R .430 .402 R Square .185 .161 Adjusted R Square .135 .110 Std. Error of the Estimate .9515 .9653 F 3.702 3.134 Sig. .001 .005 N 122 122

***Significant at .01 level **Significant at .05 level; *Significant at .1 level (2-tailed)

PAQ SUMMER 2015

247

Table 7 Standardized Regression Coefficients: Bureaucratic Decentralization

Model 1 Model 2

Beta t Beta t

Budget Environment: MCAS/NCLB Budget Influence .091 1.021 .099 1.111 Wealth Index .276*** 2.785 .280*** 2.835 Fiscal Stress Index -.095 -1.077 -.106 -1.191 School Choice -.136 -1.546 Supplemental Educational Services -.129 -1.444

School District Characteristics: Number of Schools .086 .836 .078 .750 Municipal School District .169* 1.774 .157* 1.652 Rural School District .022 .240 .023 .247 R .360 .357 R Square .130 .127 Adjusted R Square .076 .074 Std. Error of the Estimate .9676 .9688 F 2.428 2.379 Sig. .024 .026 N 122 122

***Significant at .01 level; **Significant at .05 level; *Significant at .1 level (2-tailed) These regressions supply weak support to Hypothesis 3. The budget environment and school district characteristics do affect decentralization, but, clearly, there are other variables that are more important. This issue is addressed in the Discussion section, below.

DISCUSSION The majority of the school districts in the cross section micromanage the position and transfer control functions. Following Schick (1966), we speculate that they do so because they lack confidence that internal controls are sufficient to avoid deficit spending. Weak accounting and control mechanisms increase the probability of filling unfunded positions and transferring funds from over obligated accounts. Many districts do allow principals to originate a budget and control the disposition of funds in budgeted accounts, but the comprehensive decentralization advocated in the school-based budgeting literature is nowhere to be found. Decentralization

PAQ SUMMER 2015

248

does not follow ideal types, as we observed in the wide variation in the frequency with which each of the three budget practices has been decentralized. Instead, the Massachusetts school districts represented in this study are selective in the budget decisions that they delegate to principals (Brown, 1987). Interestingly, political decentralization includes both bottom-up and top-down accountability. Where budget development and control over accounts are delegated to school principals, school committees are more involved in school level budget details and, to a lesser extent, more likely to micromanage transfers. School councils and teachers are more influential in budget decisions. Explaining this complex set of relationships is not simple. Regression results indicate that higher political decentralization factor scores are driven, in part, by the influence of the No Child Left Behind Act and the Massachusetts Comprehensive Assessment System on budget decisions, by a smaller number of schools, and, to a lesser extent, by fiscal stress. School choice, as required by NCLB for schools that fail to make adequate yearly progress, is a centralizing force. NCLB appears to have differential effects. When the MCAS testing regime is salient in budget decisions, political decentralization scores are higher, but when individual schools are in trouble and the district must offer school choice, school committees are less likely to relinquish control. While the reform literature assumes that the benefits of influential school councils are positive, the risks of increased involvement by school level stakeholders could be high for school committees. Roza (2008), for example, asserts that the allocation of resources based on political demand can “undermine district strategy” and prevent the resources from being distributed to those it was intended. She also notes that accountability should accompany discretion to reduce this problem.

Rubin’s (1996) suggestion that an increased emphasis on accountability for results shapes budgetary decision making also finds support in these findings. Where political decentralization is strongest, higher levels of political influence and control can be found at both the school and the district level. To the question: “How can elected officials and chief executives ensure that managers with increased budget flexibility will pursue

PAQ SUMMER 2015

249

objectives consistent with policy and/or strategic goals?” this research suggests that they do the obvious. They involve themselves. Does the increased involvement of school committees observed in districts with higher political decentralization scores constitute micromanagement? There is no empirical answer to this normative question, especially if a discussion of the legitimate role of the school committee is left out of school- based budgeting prescriptions. Should budgeting be decentralized? Some school committees represented among the respondents seem to think so. They do look to principals and stakeholders to manage NCLB/MCAS mandates and, to a lesser extent, fiscal stress. But, given a manageable span of control, school committees stay involved. What reformers may perceive to be micromanagement, school committee members may perceive to be an ethical obligation. Prior practices and the incremental nature of change in the public sector could also result in a more “unintentional” or “de facto” strategy (Roza, 2008, p. 1). An alternative to viewing legitimate concerns for accountability as a force that shapes political decentralization is the more cynical view that some districts want to adopt school- based budgeting without confronting internal resistance to change, the constrained model suggested by Moynihan (2009). Superintendents and business officials would likely be the primary constituencies that would be opposed to increasing the budgetary powers of principals and increasing the influence of school councils. Case study research is needed to determine the extent to which opposition by these key actors is driven by their own propensity to micromanage or to legitimate concerns about the abilities of principals to manage in a decentralized budgeting environment. Bureaucratic decentralization, limited in the findings of this study to transfers among accounts, is more likely to occur in wealthier, municipal school districts. Following Schick’s (1966) logic, these districts may have better internal controls and larger and/or better qualified staffs in the central office, allowing elected officials to relax control. This type of decentralization primarily applies to superintendents and business officials, as

PAQ SUMMER 2015

250

indicated by the lower factor loading for principals. It is striking, however, that 55 percent of the responding business officials indicate that the school committee maintain control over transfers. Massachusetts school committees, it would appear, are careful about delegating control over transfers. Running a budget deficit could be a high stakes proposition for a school committee member who wants to be reelected. One must ask, as well, “Just how important is control over transfers to the school-based budgeting prescription? Would a transfer be more easily or quickly accomplished if the principal could initiate and approve the transfer?” A pre-audit of the transaction would still be required to determine sufficiency of funds in the source account, but, presumably, the transaction could bypass the signature of top officials, but it would be a poor internal control system indeed if the business official or a designee did not also have to sign off on the transfer. This study’s findings are generally consistent with the school-based budgeting literature in that stakeholders do become more influential under the political form of decentralization and, in this respect, decentralization may be a system changing strategy (Wohlstetter & Buffett, 1991). However, this study finds no reduction in central office budget influence under any form of decentralization. Admittedly, this may be due, in part, to the business official’s perspective. More research is needed to determine if there is a change in how superintendents and business officials exercise budget influence under decentralization. For example, respondents may fail to recognize the impact of “micro-decisions” made by principals, such as how a shared staff member distributes time, on resource allocation (Roza , 2008), which could result in an under- estimation of decentralized decisions. It is tempting to simply state that the human propensity for micromanagement is too powerful a force to be overcome by the prescriptions of reformers. Admittedly, superintendents and business officials may attempt to influence school boards to maintain centralized control over budgeting in order to protect their own turfs (Moynihan 2009), but these officials, along with school boards, have a stewardship role to play that they cannot abandon. Their concerns may be real. This should lead us away from

PAQ SUMMER 2015

251

prescriptions and to a practical focus on how elected officials and administrators perceive and exercise their stewardship. The best explanations for the variance in political and bureaucratic decentralization may be found in the relationships forged by professional administrators with school committees and the community, independent of the budget environment or school district characteristics. The very limited nature of decentralization among the school districts studied, which appears to occur in a kind of “safe mode,” may also help explain why the budget environment is not more important. School districts have already hedged against potential problems stemming from decentralization. One might also examine the adequacy of financial management systems and the experience of elected officials and administrator with the adequacy of these systems as factors in the propensity of elected officials to micromanage. Reforms have an uphill battle if they are inconsistent with the norms, values, and experiences that professionals and citizens bring to the organization.

PAQ SUMMER 2015

252

REFERENCES

Anthony, P. G., & Rossman, G.B. (1994). The Massachusetts Education Reform Act: What is it and will it work? Lanham, MD: Education Resources Information Center.

Anthony, R. (1965). Planning and control systems: A framework for analysis. Boston, MA: Harvard University.

Anthony, R. & Young, D. (2007). Management control in nonprofit organizations, 7th Edition. New York: McGraw-Hill.

Bandur, A. (2012). School-based management and development: Challenges and impacts. Journal of Educational Administration, 50(6): 845-873.

Barzelay, M. (2001). The new public management. Berkeley and Los Angeles: University of California Press.

Barzelay, M. & Thompson, F. (2003). Efficiency counts: Developing the capacity to manage costs at Air Force Materiel Command. IBM Center for the Business of Government Financial Management Series.

Behn, R. D. (1995). The big questions of public management. Public Administration Review, 55(4), 313- 324.

Bird, J. J. & Wang, C. (2010). Role interdependency between superintendents and school business officials in budget building leadership practices and strategies. The Journal of School Business Management, 22(1), 11-16.

Bradbury, K., Case, K., & Mayer, C. (1998, July). School quality and Massachusetts enrollment shifts in the context of tax limitations. New England Economic Review.

Brown, D. J. (1987). A preliminary inquiry into school-based management. Report to the Social Sciences and Humanities Council of Canada for Grant Number 410- 83-1086, Ottawa, Ontario.

PAQ SUMMER 2015

253

Caldwell, B. (1993, June). Decentralising the management of Australia’s schools: A discussion paper. Melbourne, Australia: National Industry Education Forum.

Candal, C. S. (2009, September). School-based management: A practical path to school district reform. Boston, MA: Pioneer Institute Center for School Reform Policy Brief.

Chiang, H. (2009, October). How accountability pressure on failing schools affects student achievement. Journal of Public Economics, 93 (9-10), 1045-1057.

Clover, M. W., Jones, E. B., Bailey, W. & Griffin, B. (2004). Budget priorities of selected principals: Reallocation of state funds. NASSP Bulletin, 88640, 69-82.

Craig, S. G., Imberman, S. A. & Perdue, A. (2013). Does it pay to get an A? School resource allocations in response to accountability ratings. Journal of Urban Economics, 73 (1), 30-42.

Cuban, L. (2007). No more magical thinking: Leading from top or bottom. School Administrator, 64 (3), 643.

Fermanich, M., Odden, A., & Archibald, S. (2000). A case study of district decentralization and site-based budgeting: Cordell Place School District. Wisconsin Center for Education Research, Consortium for Policy Research.

Forrester, J. P. & Spindler, C.J. (2001). Budgeting theory through relational learning. International Journal of Organization Theory and Behavior, 4(1 & 2), 107-131.

Goertz, M. E. & Hess, G. A. (1998). Processes and power in school budgeting across four large urban school districts. Journal of Education Finance, 23(4), 490-506.

Goertz, M. E. & Stiefel, L. (1998). School-level resource allocation in urban public schools. Journal of Education Finance, 23 (4), 435-46.

Gulick, L. (1983). The dynamics of public administration today as guidelines for the future, Public Administration Review, 43 (3), 193-198.

PAQ SUMMER 2015

254

Hadderman, M. (2002). School-based budgeting. Teacher Librarian, 30(1), 27-29.

Hansen, J. S. & Roza, M. (2005). Decentralized decision-making for schools: New promise for an old idea? Santa Monica: Rand Corporation.

Hayek, F. A. (1945). The use of knowledge in society. The American Economic Review, 35(4), 519-530.

Jones, L. R., & Thompson, F. (1986). Reform of budget execution control. Public Budgeting and Finance, 6(1), 33-49.

Kedro, M. J. (2003, January). Controlling the purse strings. Principal Leadership, 3(5), 19-23.

Lu, H. & Facer II, R. L. (2004, March). Budget change in Georgia counties: Examining patterns and practices. American Review of Public Administration, 34(1), 67- 93.

Marlowe, J. & Portillo, S. (2006). Citizen engagement in local budgeting: Does diversity pay dividends? Public Performance and Management Review, 30(2), 179-202.

Massachusetts Education Reform Act of 1993. 1993 Massachusetts Acts 159.

McDermott, K. A., Berger, J. B., Bowles, S., Brooks, C. C., Churchill, A., & Effrat, A. (2001). An analysis of state capacity to implement the Massachusetts Education Reform Act of 1993. Boston, MA: Massachusetts Education Reform Review Commission.

Moore, M. H. (1995). Creating public value: Strategic management in government. Cambridge, MA: Harvard University Press.

Moynihan, D. P. (2009). The new architecture of state government administration: Key questions and preliminary answers. Public Administration Quarterly, 33(2), 164-196.

PAQ SUMMER 2015

255

No Child Left Behind Act of 2001. Public Law 107–110, 107th Congress, Jan. 8, 2002.

Osborne, D. & Gaebler, T. (1992). Reinventing government: How the entrepreneurial spirit is transforming the public sector. Reading, MA: Addison-Wesley.

Raffel, J. A. (2007). Why has public administration ignored public education, and does it matter? Public Administration Review, 67(1), 135-151.

Roza, M. (2008). Allocation anatomy: How district policies that deploy resources can support (or undermine) district reform strategies. Seattle: Center on Reinventing Public Education. Retrieved from www.crpe.org/sites/default/files/pub_sfrp_aa_may08_0. pdf.

Rubin, I. S. (1996). Budgeting for accountability: Municipal budgeting for the 1990s. Public Budgeting and Finance, 16 (2), 112–132.

Rubin, I. S. (2010). The politics of public budgeting, (6th ed.) Washington, D.C.: CQ Press.

Schick, A. (1966, December). The road to PPB: The stages of budget reform. Public Administration Review, 26(4), 243-258.

Simon, H. A. (1976). Administrative behavior: A study of decision-making processes in administrative organization, Third Edition. New York: Free Press.

Straut, D. S. (1996). Full circle: A retrospective on labor relations and educational governance, Unpublished paper presented at the Annual Meeting of the American Educational Research Association, New York, NY, April 8-12.

Thompson, B. (2004). Exploratory and confirmatory factor analysis. Washington, DC: American Psychological Association.

PAQ SUMMER 2015

256

Thurmaier, K. & Gosling, J. J. (1997). The shifting roles of state budget offices in the Midwest: Gosling revisited. Public Budgeting and Finance, 17(4), 48–70.

United States Bureau of the Census, Local government employment and payroll, March 2006 and 2007 Census of Governments, Retrieved from www.census.gov.

United States Department of Education. Schools and Staffing Survey, 2007-08 School Year, Principal’s Survey. National Center for Education Statistics. Retrieved from http://nces.ed.gov/surveys/sass/question0708.asp.

Wildavsky, A. (1992). Political implications of budget reform: A retrospective, Public Administration Review, 52(6), 594-599.

Williamson, A., Snow, D. (2014). From accountability to decision making? Budgeting with mandated performance measures. International Journal of Public Administration, 37(4), 202-214.

Williamson, A., Snow, D. (2013). Bridging theory and practice: The landscape of public management reforms in local school district budgeting. Public Performance and Management Review, 37(1), 154-187

Willis, J. (2010). Seeing public engagement differently. School Business Affairs, 76(10), 16-18.

Wohlstetter, P. & Van Kirk, A. (1995). School-based budgeting: Organizing for high performance. Paper presented at the Annual Meeting of the American Educational Research Association, San Francisco, CA, April 18-22.

Wohlstetter, P. & Buffett, T. (1991). School-based management in big city districts: Are dollars decentralized too? Unpublished manuscript presented at Annual Meeting of the American Educational Research Association, Chicago, IL, April 3-7.

PAQ SUMMER 2015

257

Appendix 1. Comparison of Responding Districts to All Non- charter School Districts by Type

Type of School District Respondents All Districts Municipal 71.3% 74.2% Regional 28.7% 25.8% Total 100.0% 100.0%

Appendix 2. Comparison of Responding Districts to All Non- charter School Districts by Region Geographic Region Respondents All Districts Eastern Massachusetts 64.0% 55.9% Western Massachusetts 32.6% 36.8% Cape Cod and the Islands 3.4% 7.3% Total 100.0% 100.0% Appendix 3. Comparison of Responding Districts to All Non- charter School Districts by Enrollment, Spending, and State Aid

Respondents All Districts

ENRL

Expend. Per

Student State

Aid (%) ENRL

Expend. Per

Student State

Aid (%)

Mean 3,120 9,726 39.1 2,868 9,616 39.1 Std. Dev. 3,262 2,061 20.2 4,469 1,949 20.3 Minimum 134 7,985 13.4 7 7,854 13.4 Maximum 28,235 18,929 96.3 60,951 21,687 96.3 N = 179 N = 328

Independent samples t tests reveal no statistically significant differences. Variance in spending per student of respondents, however, is significantly different (p = .044) from non-respondents. Enrollment differences, while not statistically significant, are due to the very large City of Boston School Department, which is not among the responding districts.

PAQ SUMMER 2015

258

Appendix 4. Components of the MCAS and NCLB Index

Cronbach’s Alpha: .710. Appendix 5. Components of the Wealth Index Wealth Index Mean S. D.

Net school spending as percentage of requirement 115.63 10.95 Percentage of minimum spending from own source revenue (1 - state aid) 60.84 20.28 Percentage of students ineligible for free and reduced lunches 80.14 17.01 N = 175, Cronbach’s Alpha: .738. Appendix 6. Components of the Fiscal Stress Index

Fiscal Stress Index Mean S.D. Summation of survey responses for following items (potential range of 0-24): Thinking about fiscal stress, which of the following has your district experienced recently, is experiencing this year, or expects to experience over the next few years? (Please check ALL that apply. You may check more than one column for each item.). Difficulty Balancing the Budget, Layoffs, Decline in Local Contribution, Decline in State Aid, Mid Fiscal year Budget Cuts, Elimination of Academic Programs due to Fiscal Stress, Elimination of Extracurricular Activities due to Fiscal Issues, Inability to Provide Services at the Level or Quality Expected by the Citizens in our Community

9.32 4.87

N = 142, Cronbach’s Alpha: N/A

Importance of MCAS and NCLB in Budget Decisions N Mean S.D.

% Agree

No Child Left Behind requirements are a significant force behind budget decisions 158 3.54 .803 59.5 MCAS scores are a significant force behind budget decisions 160 3.40 1.041 53.1 Reallocation of FTE positions because of low MCAS scores 158 3.36 0.946 50.6 Importance of MCAS scores to the school committee in making budgetary decisions 165 3.97 0.946 74.5 Importance of MCAS scores to the superintendent/central office in making budgetary decisions 162 4.04 0.993 79.6

Copyright of Public Administration Quarterly is the property of Southern Public Administration Education Foundation and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.