Lotteries, litigation, and education finance.

• Author: Dee, Thomas S.

1. Introduction

   Historically, the finance of public education in the United States has been largely based on revenue raised locally through property taxation. However, this system has been routinely criticized on the grounds that the resulting distribution of educational opportunity is inequitable and that it may lead us to forgo socially productive investments in education. Over the past several decades, these concerns have motivated extensive litigation in almost every state, challenging the constitutionality of property-based education finance. To date, the supreme courts in 17 of these states have encouraged increased state aid to school districts by concluding that their current systems of education finance were indeed unconstitutional. However, over this same period, 37 states have also turned to new state lotteries as a way to increase state funding for services like education. There are two shared reasons that these disparate approaches to directing new resources to education might prove to be ineffective. First, it is by no means clear that these reforms would actually increase state or local educational spending. State legislatures may respond slowly, if at all, to a court ruling that encourages increased aid. Furthermore, states that earmark new lottery revenues for education may simply choose to then reduce their education appropriations from other sources. And, even if these reforms do increase state aid, the effect on educational spending may be undone by reductions in education revenues raised from local and federal sources. (1) Second, even if new state aid were to increase local education spending, it is not clear that these new resources would be allocated in ways that actually improved school quality. The extensive though controversial literature on educational productivity suggests that new resources would not be used effectively (e.g., Burtless 1996).

   This study provides novel, empirical evidence on these policy-relevant concerns by examining the consequences of two of the most recent and noteworthy education finance reforms. In 1993, Tennessee began court-ordered education finance reforms while the neighboring state of Georgia simultaneously initiated a lottery explicitly designed to promote educational spending. These policy experiments are partly of interest simply because of their narrow consequences for the patterns of educational finance and spending within these states. However, they are also likely to be of more general interest because they provide evidence on a question of importance to policymakers and voters everywhere. Specifically, the consequences of these reforms provide evidence on how plausibly exogenous increases in available educational resources are actually spent. Though there is an extensive literature on educational productivity, there is surprisingly little evidence on how school districts allocate resources. (2) Furthermore, these evaluations may also prove of more general interest because each of these states adopted rather unique and interesting strategies to earmark new state aid for specific educational expenditures. In Tennessee, the court-ordered education finance reforms were accompanied by broader educational reforms designed to ensure the efficacy of the new state aid. These measures included accountability measures for schools and school districts based on controversial, value-added measures of student test performance; the phasing in of mandatory class-size regulations; and a high-profile initiative to promote technological investments in classrooms. Georgia also attempted to earmark new lottery revenues for specific educational spending by promising new and highly visible educational initiatives (i.e., Helping Outstanding Pupils Educationally [HOPE] Scholarships, prekindergarten [pre-K] programs, and technological investments for classrooms). The absolute and comparative success of these earmarking attempts can provide evidence to other communities that hope to promote spending on specific educational functions.

   This study presents evidence on how these policies influenced per pupil revenues by source (federal, state and local) and expenditures by function. The panel data used in these evaluations are from the unified school districts in Georgia, Tennessee, and South Carolina and the annual "F-33" Survey of Local Government Finances combined for the 1990, 1992, 1994, and 1995 fiscal years. The F-33 survey divides expenditures into six broad categories: instruction, support services, noninstructional services, functions unrelated to elementary and secondary education, capital expenditures, and other expenditures. While more detailed data on the allocation of expenditures would have been welcome, this taxonomy still allows us to address the broad question of interest, namely, whether these reforms were effective in increasing the targeted functions of student instruction and capital improvements. (3) The empirical results presented here are based on two-way fixed effect specifications that exploit the panel nature of the available data to purge the potentially confounding determinants of revenue and expenditure outcomes that are specific to each school district and to each fiscal year. (4) The results indicate that both policy initiatives were successful at increasing state aid to school districts as well as district spending. (5) Furthermore, in both cases, this new state aid was targeted toward the poorest districts within each state, and therefore improved the equity with which state resources were distributed. Additionally, in both states, the new reform-driven aid led to statistically significant increases in spending on the targeted functions (e.g., student instruction, equipment purchases). However, significant amounts of the new aid were also allocated to functions that were not clearly targeted (e.g., support and noninstructional services). A comparative assessment of each state's earmarking efforts suggests that Georgia's were most effective, perhaps because they were linked to new and highly visible educational initiatives.

2. Lotteries and Litigation

   Over the past several decades, an extensive body of litigation has attempted to improve the equity with which educational resources are distributed. Beginning with the influential 1971 Serrano decision in California, court rulings in almost every state have assessed the constitutionality of systems of education finance based largely on local property wealth. To date, the supreme courts in 17 states have ruled in favor of the plaintiffs, deeming the systems of education finance unconstitutional. (6) These decisions have typically been based on interpretations of the equal protection clauses as well as the education articles of state constitutions. However, the rulings in recent years have addressed equity issues indirectly, focusing instead on the unique needs of students in poor communities and the inadequacy of locally available resources (Verstegen 1994). Recent empirical evaluations have demonstrated that the earliest court-ordered reforms were effective in encouraging states to direct new resources to poorer school districts (Evan, Murray, and Schwab 1997; Murray, Evans, and Schwab 1998; Card and Payne 2002). (7) For example, in a study based on district-level panel data from the 1972-1992 period, Murray, Evans, and Schwab (1998) conclude that the earliest 11 state reforms increased spending in the poorest districts by 11% while leaving spending in the wealthiest districts unchanged.

   This study assesses the effects of an important and more recent ruling, the 1993 Tennessee decision in Tennessee Small School Systems v. McWherter. (8) The litigation in Tennessee was initiated in 1988 by a group of small, largely rural school districts. Over the next four years, a combination of contradictory lower-court rulings and controversial legislative efforts to identify a tax base for new state funding left the issue largely unresolved. However, in March 1993, the Tennessee Supreme Court ruled unanimously that the state's foundation program for school funding was unconstitutional. Like other recent rulings, this decision emphasized the state's role in ensuring the equal availability of a quality education. The plaintiffs sought to have the state immediately improve funding levels. However, the state was instead allowed to implement the new Basic Education Program (BEP) created by the Education Improvement Act (EIA), which phased in $1 billion of new aid over the next five years (Green 1997; Goldhaber and Callahan 2001). The money for these increases came largely from a half-cent increase in the state sales tax. (9) The new BEP funding formula, one of the most complex in the nation, adjusted state aid in response to local fiscal capacity as well as cost differentials. However, the amount of aid generated by the formula was also based on the district-specific costs associated with 33 explicit classroom functions and 10 nonclassroom functions deemed necessary by the state for educational adequacy (Green 1997; Goldhaber and Callahan 2001). The state allocations associated with the BEP categories were explicitly designed to reallocate funds toward "direct student or classroom needs" (Green 1997). Interestingly, the increased funding created by the EIA was also bundled with other comprehensive educational reforms intended to target the new spending productively. The earmarking efforts included the phasing in of mandatory class-size reductions and an initiative to invest in high-technology "21st-century classrooms." However, the EIA also instituted several controversial accountability measures intended to promote the productive use of the new aid. These included exit exams for graduating students; performance standards for schools and districts based on test scores, dropout rates, and student attendance; and state authority to place underperforming schools and districts on probation.

   Over the...