Archived Working Papers

By Year 

2009 Papers

Working Paper #09-10 (November)

Why Do Some Schools Get More and Others Less? An Examination of School-Level Funding in New York City
Amy Ellen Schwartz, Ross Rubenstein, Leanna Stiefel

In the spring of 2007, the New York City Department of Education announced an ambitious plan to change the way it distributes resources across its more than 1,400 schools. The plan, known as the "Fair Student Funding" initiative, is intended to change funding methods in two ways: first, by allocating money based upon characteristics of the student body that capture differences in the cost of providing appropriate educational services; second, by allocating dollars rather than specific resources, primarily teacher positions, and allowing principals greater discretion in the deployment of those resources. The overarching goal is to improve equity, particularly vertical equity, in the distribution of resources and, ultimately, to improve the efficiency of how resources are used to promote student performance. These reforms will be implemented gradually. For 2007-2008, the new formulae will only be used to distribute a portion of newly available funds, and substantial hold-harmless provisions significantly limit the impact on overall funding.

Given this dramatic policy change, it is a particularly good time to examine the current distribution of resources across schools in New York City and to consider what we know about the intradistrict allocation of resources. The purpose is two fold. First, we hope to provide a context for understanding funding reform, both through empirical analyses of funding in New York City and by drawing on the lessons from previous reforms and research. Second, we hope to provide benchmarks against which we can assess the impact of Fair Student Funding in the future, as it is implemented. Before doing so, however, we consider why - and how - the intradistrict allocation of resources matters. We then review previous evidence on the distribution of school resources in New York City and elsewhere, including typical allocation methods and reform proposals. Next, we present new empirical analyses examining the distribution of resources by funding source across schools in New York City. We conclude with a discussion of policy implications. 

Working Paper #09-09 (August)
Does Title I Increase Spending and Improve Performance? Evidence from New York City
Meryle G. Weinstein, Leanna Stiefel, Amy Ellen Schwartz, Luis Chalico

Since its inception as part of the 1965 Elementary and Secondary Education Act(ESEA), Title I has provided the largest amount of federal funding aimed at improving the academic achievement of poor children. In this paper, we examine the impact of Title I on school spending and school performance, using New York City public school data. Based on a regression discontinuity design (RD) with panel data, and including separate analyses for elementary/middle and high schools, we estimate local average treatment effects of Title I.

Title I provides additional funding for schools serving high concentrations of poor children, but there are few curricular or programmatic constraints. Unfortunately it is possible that Title I funds supplant state or local funds, resulting in muted net impacts on spending and student outcomes. At the same time, the success of Title I in improving test scores offers important insight into the effectiveness of school-based compensatory funding in general, such as weighted student funding within districts or state compensatory aid across districts.

Overall, the results indicate that Title I changes the mix of spending, enabling high schools to significantly increase the amount of money they spend on direct services to students and to improve their pupil-teacher ratios (while reducing experienced teachers). Elementary and middle schools do not increase spending as much, which is consistent with our finding that state compensatory education funds may be supplanting some Title I funding in schools. Since schools just below the Title I cutoff are similar to those just above the cutoff, this finding may be an equitable, albeit unintended result.

Finally, additional Title I spending does not improve the achievement of students and may even reduce school-wide average test scores in elementary and middle schools. These effects for both spending and scores seem to increase with the length of time schools are Title I eligible and to be stronger for ones that are always Title I eligible compared to those that go in and out of eligibility.


Working Paper #09-08 (July)
Do Public Schools Disadvantage Students Living in Public Housing?

Amy Ellen Schwartz, Brian J. McCabe, Ingrid Gould Ellen and Colin Chellman

In the United States, public housing developments are predominantly located in neighborhoods with low median incomes, high rates of poverty and disproportionately high concentrations of minorities. While research consistently shows that public housing developments are located in economically and socially disadvantaged neighborhoods, we know little about the characteristics of the schools serving students in these neighborhoods. In this paper, we examine the characteristics of elementary and middle schools attended by students living in public housing developments in New York City. Using the proportion of public housing students attending each elementary and middle school as our weight, we calculate the weighted average of school characteristics to describe the typical school attended by students living in public housing. We then compare these characteristics to those of the typical school attended by other students throughout the city in an effort to assess whether public schools systematically disadvantage students in public housing in New York City. Our results are decidedly mixed. On one hand, we find no large differences between the resources of the schools attended by students living in public housing and the schools attended by their peers living elsewhere in the city; on the other hand, we find significant differences in student characteristics and outcomes. The typical school attended by public housing students has higher poverty rates and lower average performance on standardized exams than the schools attended by others. These school differences, however, fail to fully explain the performance disparities: we find that students living in public housing score lower, on average, on standardized tests than their schoolmates living elsewhere -- even though they attend the same school. These results point to a need for more nuanced analyses of policies and practices in schools, as well as the outside-of-school factors that shape educational success, to identify and address the needs of students in public housing.

Working Paper #09-07 (June)
From Front Yards to School Yards: Linking Housing Policy and School Reform
Amy Ellen Schwartz, Leanna Stiefel and Abigail Conover Carlton

Housing and education share strong ties in the United States. This relationship is shaped, in large part, by mobility. Students move to new schools, homes and neighborhoods as a result of planned and unplanned family relocations. Taxpayers move from one school district to another in a nation where school quality is closely tied to the district in which a family resides. Teachers weigh factors such as location, pay, and long-term career opportunities as they decide where to work and when to move within or between school districts. Despite the strong relationship between housing and education, policies that recognize and support this relationship are relatively rare. In this paper, we explore the mechanisms by which housing and education are related. We focus particular attention on disadvantaged students in urban areas, as these students often face a unique set of challenges that set them apart from their more advantaged and/or non-urban counterparts. First, we explore the ways in which a child's housing unit, his neighborhood, and the political economy of public schools might shape his educational outcome. We then turn to a discussion of the implications of these mechanisms for education and housing policy. Herein, we highlight recent efforts to strengthen the ties between education and housing policy and discuss how the lessons learned from these efforts might be brought to bear as policymakers consider new education and housing initiatives.

Working Paper #09-06 (May)
Accounting for Resource Use at the School-level and Below: The Missing Link in Education Administration and Policy Making
Dwight V. Denison, Leanna Stiefel, William Hartman, and Michele Moser Deegan                                            

A long standing debate among policymakers as well as researchers is whether and how funding affects the quality of education. Often missing from the discussion is information about the costs of providing education at the school level and below, yet such information could impart a better indication of the linkages between outcomes and resources than is available with more macro-level data. In addition, because No Child Left Behind (NCLB) and state accountability systems often require reporting of performance at the grade or school level, micro-level cost information would be useful to school administrators as they try to allocate resources productively.

In this paper, we analyze the challenges involved in establishing a system to track costs at the school, grade, and subject level that will fit the needs of both internal and external users. To begin, we review the literature on cost accounting that is relevant to micro-level costs and the research that analyzes sub-district level resources. Next, we describe general challenges that arise in reporting at the level of the school and below and we then discuss school-level reporting in practice. We follow with a case study of an improved reporting system that links resource use, student demographic characteristics, and student outcomes at the school, grade and subject level. We conclude with recommendations for states when constructing such systems.

Working Paper #09-05 (April)
Risk Aversion and Support for Merit Pay: Theory and Evidence from Minnesota's Q Comp Program
Carl Nadler and Matthew Wiswall                                                      

Recent research attributes the lack of merit pay in teaching to the resistance of teachers. This paper examines whether the structure of merit pay affects the types of teachers who support it. We developed a model of the relative utility teachers receive from merit pay versus the current fixed schedule of raises. We show that if teachers are risk averse, teachers with higher base salaries would be more likely to support a merit pay  program that allows them to keep their current base salary and risk only future salary increases. We test the predictions of the model using data from a new merit pay program, the Minnesota "Q Comp" program, which requires the approval of the teachers in each school district. Consistent with the model's predictions, we find that districts with higher base salaries and a higher proportion of teachers with masters degrees are more likely to approve merit pay.

Working Paper #09-04 (March)
Vouchers, Public School Respnse and the Role of Incentives: Evidence From Florida 
Rajashri Chakrabarti

In this paper, I analyze the behavior of public schools facing vouchers. The literature on the effect of voucher programs on public schools typically focuses on student and mean school scores. This paper tries to go inside the black box to investigate some of the ways in which schools facing the threat of vouchers in Florida behaved. Under the 1999 Florida program, schools getting an "F" grade for the first time were exposed to the threat of vouchers, but did not face vouchers unless and until they got a second "F" within the next three years. Exploiting the institutional details of this program, I analyze the incentives built into the system and investigate the behavior of the threatened public schools facing these incentives. There is strong evidence that they did respond to incentives. Using highly disaggregated school level data, a difference-in-differences estimation strategy as well as a regression discontinuity analysis, I find that the threatened schools tended to focus more on students below the minimum criteria cutoffs rather than equally on all, but interestingly, this improvement did not come at the expense of higher performing students. Second, consistent with incentives, they focused mostly on writing rather than reading and math. These results are robust to controlling for differential pre-program trends, changes in demographic compositions, mean reversion and sorting. These findings have important policy implications.

Working Paper #09-03 (March)
Impact of Voucher Design on Public School Performance: Evidence From Florida and Milwaukee Voucher Program  
Rajashri Chakrabarti

This paper examines the impact of vouchers in general and voucher design in particular on public school performance. It argues that all voucher programs are not created equal. There are often fundamental differences in voucher designs that affect public school incentives differently and induce different responses from them. It analyzes two voucher programs in the U.S. The 1990 Milwaukee experiment can be looked upon as a "voucher shock" program that suddenly made low-income students eligible for vouchers. The 1999 Florida program can be looked upon as a "threat of voucher" program, where schools getting an "F" grade for the first time are exposed to the threat of vouchers, but do not face vouchers unless and until they get a second "F" within the next three years. In the context of a formal theoretical model, the study argues that the threatened public schools will unambiguously improve under the Florida-type program and this improvement will exceed that under the Milwaukee type program. Using school-level scores from Florida and Wisconsin, and a difference-in-differences estimation strategy in trends, it then shows that these predictions are validated empirically. These findings are reasonably robust in that they survive several sensitivity checks including correcting for mean reversion and a regression discontinuity analysis.

Working Paper #09-02 (March)
Do Vouchers Lead to Sorting Under Random Private School Selection? Evidence From the Milwaukee Voucher Program
Rajashri Chakrabarti

This paper analyzes the impact of voucher design on student sorting, and more specifically investigates whether there are feasible ways of designing vouchers that can reduce or eliminate student sorting. It studies these questions in the context of the first five years of the Milwaukee voucher program. Much of the existing literature investigates the question of sorting where private schools can screen students. However, the publicly funded U.S. voucher programs require private schools to accept all students unless oversubscribed and to pick students randomly if oversubscribed. This paper focuses on two crucial features of the Milwaukee voucher program-random private school selection and the absence of topping up of vouchers. In the context of a theoretical model, it argues that random private school selection alone cannot prevent student sorting. However, random private school selection coupled with the absence of topping up can preclude sorting by income, although there is still sorting by ability.Sorting by ability is not caused here by private school selection, but rather by parental self selection. Using a logit model and student level data from the Milwaukee voucher program for 1990-94, it then establishes that random selection has indeed taken place so that it provides an appropriate setting to test the corresponding theoretical predictions in the data. Next, using several alternative logit specifications, it demonstrates that these predictions are validated empirically. These findings have important policy implications.

Working Paper #09-01 (February)
Local Demand for School Choice: Evidence from the Washington Charter School Referenda

Sean Corcoran and Christiana Stoddard

Many U.S. states provide public funding for charter schools-deregulated and privately managed schools operating in direct competition with government-run schools. While the impact of charter schools on student achievement and sorting has been intensely studied, less is known about the demand for these alternatives. Using precinct-level returns from three ballot referenda in Washington State, we assess the relative importance of school quality and community characteristics in explaining voter support for charter schools. We find that low student achievement predicts greater charter support across school districts, but is relatively unimportant in explaining variation across precincts within districts. Residents of districts with more highly qualified teachers and greater local spending were less likely to favor charters, as were districts with high teacher union membership. The strongest predictor at all levels was political partisanship: areas with more Republican voters were strongly and consistently more likely to vote in favor of charter schools.

2008 Papers

Working Paper #08-01 (January)
Do Immigrants Differ from Migrants? Disentangling the Impact of Mobility on High School Completion and Performance

Leanna Stiefel, Amy Ellen Schwartz and Dylan Conger
As immigrant students continue to enter U.S. schools in large numbers, policymakers, parents and school leaders have become intensely interested in their academic performance and educational attainment. While previous evidence has pointed to superior performance by foreign-born students in their elementary and middle school years, growing concern has centered around the education and life chance of immigrants who come to the United States in their high school years and pointed to a significant gap in the research literature. This paper takes a step toward filling the gap. We use data on a cohort of New York City public high school students to examine how the performance of immigrant students differs between students who enter in high school, middle school or elementary school, adjusting for the conventional student characteristics that may shape outcomes. We then compare these disparities to the disparities experienced by the native-born population in order to remove any differences in performance due merely to differences in mobility. Thus, we derive estimates of the "cost" in performance due to their entry in high school that has been purged of a range of possible confounding factors. Importantly, our difference-in-difference estimates suggests that, ceteris paribus, immigrant students do quite well and high school entrants even better than earlier entering immigrants.
Working Paper #08-02 (January)
New Directions in Measuring Racial Isolation in School

Dylan Conger
This article offers new directions in measuring racial isolation in schools. The most widely-used measurement approach is to examine the mean on the distribution of school percentage nonwhite across nonwhite students (the isolation rate) or the percentage of nonwhite students in schools with large shares of nonwhites (e.g. 90 percent or more) at a single point in time. Using data on New York City public school students, Conger discusses the complexity that is revealed when school officials and researchers consider the following three dimensions of racial isolation: that between classrooms, over time, and among nonwhite students.
Working Paper #08-03 (March)
Mission Matters: The Cost of Small High Schools Revisited

Leanna Stiefel, Amy Ellen Schwartz, Patrice Iatarola and Colin C. Chellman
With the financial support of several large foundations and the federal government, creating small schools has become a prominent high school reform strategy in many large American cities. While some research supports this strategy, little research assesses the relative costs of these smaller schools. We use data on over 200 New York City high schools, from 1996 through 2003, to estimate school cost functions relating per pupil expenditures to school size, controlling for school output and quality, student characteristics, and school organization. We find that the structure of costs differs across schools depending upon mission - comprehensive or themed. At their current levels of outputs, themed schools minimize per pupil costs at smaller enrollments than comprehensive schools, but these optimally-sized themed schools also cost more per pupil than optimally-sized comprehensive schools. We also find that both themed and comprehensive high schools at actual sizes are smaller than their optimal sizes.
Working Paper #08-04 (April)
Testing, Time Limits, and English Learners: Does Age of School Entry Affect How Quickly Students Can Learn English?
Dylan Conger
The No Child Left Behind Act requires schools to begin testing new English Learners (EL) in English language arts within three years after they enter school and holds schools accountable for their performance on these exams. Yet very little empirical work has examined exactly how long it takes EL students to become proficient in English and how the time to proficiency varies for different types of students. Linguistic theorists suggest, for instance, that the age at which students begin learning a second language may substantially influence their probability of obtaining proficiency quickly. Using panel data on English Learners (EL) in New York City public schools, I examine how long it takes students to become minimally-proficient in English and how the time to and probability of proficiency differs for students by their age of school entry. Specifically, I follow four entry cohorts of ELs and use discrete-time survival analysis to model variations in the rate at which different age groups acquire proficiency. I find that approximately half of the students become proficient within three years after school entry but that age of entry lowers the speed with which children can become proficient. Age of entry differences are robust to controls for differences in other student characteristic and the schools they attend. The results suggest that federal, state, and local policies regarding the testing of EL students in academic English should consider more flexible time limits.  
Working Paper #08-05 (June)
Why and How Does Source Country Matter?  The Effects of Home Countries and Immigrant Communities on Foreign-Born Student Achievement
Dylan Conger, Amy Ellen Schwartz and Leanna Stiefel
Objective: This paper explores the effect of the economic conditions of source countries and the human capital characteristics of coethnic immigrant communities on foreign-born students' reading and math achievement.
Methods: We use data on New York City public school foreign-born students from 39 countries merged with Census data on the characteristics of the city's immigrants, and United Nations data on the economic conditions of countries. We estimate regressions of student achievement on home country and coethnic immigrant community characteristics, controlling for student and school attributes.
Results: Children from middle income nations and nations where English is an official language have lower reading scores than students from other nations, though no such effects are observed for math. Children from immigrant communities with higher levels of income and educational attainment perform better in school than children from other communities. Yet children in highly English proficient immigrant communities test slightly lower than children from less proficient communities.