School quality and Massachusetts enrollment shifts in the context of tax limitations

New England Economic Review, July-August, 1998 by Katharine L. Bradbury, Karl E. Case, Christopher J. Mayer

Like most states, Massachusetts underwent a large shift in public school enrollments between the 1980s and 1990s, requiring a number of sizable fiscal and educational adjustments by individual school districts. Between school years 1980 and 1989, the number of students in kindergarten through grade 12 fell 21 percent, from 1.04 million to 825,000. As children of baby boomers reached school age, the picture changed and enrollments grew more than 90,000 over the next seven years. These aggregate trends gloss over even more marked shifts at the local level.

Consider the communities of Brookline and Arlington, whose public school enrollments in school year 1980 were 6,246 and 6,245, respectively. Both are suburban communities located close to downtown Boston with little buildable land. The quality of Arlington's schools is considered slightly above average for the state, while Brookline's schools are perennially ranked among the top districts in the Commonwealth. By the mid 1990s the enrollment patterns for the two districts could not have looked more different. Arlington was closing schools. Despite the pickup in aggregate statewide enrollments, its 1996 enrollment was 4,059, a drop of more than one-third from 1980. Meanwhile Brookline experienced a much smaller decline in its number of students in the 1980s and faced an influx of students in the 1990s. By 1996, its enrollment was 6,039, barely 3 percent below its level in 1980.

This disparate pattern of enrollment shifts was not unique to these two communities. During the same 16-year period, almost one-quarter of Massachusetts communities lost more than 20 percent of students from their 1980 levels; at the other extreme, one-quarter gained more than 12 percent. These shifts in enrollment posed a significant fiscal challenge for communities struggling to provide facilities and teachers for the widely varying numbers of students. After all, educational expenditures represent almost one-half of the local budget for a typical community in Massachusetts.

The fact that households move is not surprising. Economists since Tiebout (1956) have recognized that households sort themselves based on their ability to pay and their preferences regarding both local public services and local housing characteristics, and these preferences can change over time as families begin having children or households decide to retire. Location models would predict a "flight to quality," for example, as households with children who reach school age choose to move to communities with higher-quality schools.

As documented below, this pattern of sorting greatly increased between the 1980s and the 1990s in Massachusetts, with a much higher percentage of households with children moving in recent years. Demographics may explain part of this pattern. In the 1980s aggregate school enrollments were declining as the tail end of the baby boom was exiting the public schools and many older baby boomers had delayed childbearing. In the 1990s school enrollments were again rising. In addition, baby-boomers who were having children in the 1990s had additional income to spend on housing based on gains made in the housing market from the 1980s.

As households change their desires based on life-cycle considerations, economic models also predict that communities would adjust the amount of public services (such as police, fire, and schools) in response to the changing desires of households. Thus, one might have expected cities and towns to respond to the demographically driven increase in demand for good schools in the 1990s by raising educational expenditures. However, a statewide property tax limitation measure, Proposition 2 1/2, raised strong barriers to providing desired services in some communities.

This article investigates the degree to which the constraints of Proposition 2 1/2, and other factors such as demographic and economic shifts and differences in school quality, affected the adjustments that both local governments and households in the Commonwealth made to a demographically driven turnaround in enrollment growth. The research accomplishes this task by comparing changes in enrollments in the first half of the 1980s to those in the first half of the 1990s. It relies on two sources of data to measure and analyze the mobility of students over time: Census estimates of the number of children living in each town in Massachusetts in 1980 and 1990, and annual public school enrollments from 1980 to 1995.

The study reports three major findings. (1) Net public school enrollment changes are positively related to differences across communities in school quality. (2) Shifts in enrollments were much more pronounced in the 1990s, when aggregate enrollments were rising and the economy was improving. (3) Proposition 2 1/2 appears to have significantly altered the pattern of enrollment changes, with families with students moving to districts less constrained by this property tax limit.

The article is organized as follows. Section I documents the large cross-sectional differences in public school enrollment changes across Massachusetts communities in the 1980s and 1990s and relates them to differences in test scores. The next section discusses the manner in which households make residential location choices, sorting themselves among localities, and describes the economic, demographic, and political changes that affected these outcomes in the Commonwealth over the sample period. Section III presents regression results that examine the relationship between various community characteristics related to this broad context and the difference between actual and demographically predicted changes in enrollments from 1980 to 1985 and 1990 to 1995. The conclusion explores the implications of the results.