Research

Abstract: Fiscal accountability policies legislatively establish a process for states to intervene in local government affairs upon identification of fiscal distress. Despite their longstanding and growing presence in education, evidence on the adequacy and effectiveness of fiscal accountability policies to prevent or reduce distress for school districts remains elusive. Using an original dataset containing the universe of state statutes enacted since the late 1980s, I estimate the causal impacts of fiscal accountability in an event study framework with heterogeneity-robust estimators. I find that Monitoring laws, which focus on oversight and light-touch interventions, have a stabilizing effect for local districts. Whereas Sanctioning laws, which involve punitive measures like takeovers, have a destabilizing effect. Specifically, following the passage of Monitoring laws, districts, on average, experience an increase in total revenue, both from state and local sources. In the immediate short-term, districts are forced to cut spending, but spending stabilizes and even modestly exceeds prior levels once the additional revenues materialize. This results in a positive effect on multiple dimensions of fiscal health and a corresponding reduction in fiscal distress. In contrast, following the passage of Sanctioning laws, I find that revenues flatten, which forces districts to cut operating expenditures, largely directly from instruction. This translates to an improvement in short- (i.e., balanced budget) but not long-term (i.e., cash savings) fiscal health metrics.

Abstract: This paper asks whether increasing public school funding can be an effective long-run crime-prevention strategy in the U.S. Specifically, we examine the effect of increases in funding early in children’s lives on the likelihood that they are arrested as adults. We exploit quasi-experimental variation in public school funding, leveraging two natural experiments in Michigan and a novel administrative dataset linking the universe of Michigan public school students to adult criminal justice records. The first research design exploits variation in operating expenditures due to Michigan’s 1994 school finance reform, Proposal A. The second design exploits variation in capital spending by leveraging close school district capital bond elections in a regression discontinuity framework. In both cases, we find that students exposed to additional funding during elementary school were substantially less likely to be arrested in adulthood. We show that the social benefits of increasing school funding are greater than the costs, even when considering only the crime-reducing benefits.

Abstract: Using North Carolina as a lens to illuminate broader national developments, this paper examines how and why educational policy in the United States turned away from a civil rights agenda of opportunity and embraced test-based accountability as a way of promoting racial equality. We show that comprehensive desegregation, enforcement of the Civil Rights Act of 1964, and Great Society Programs expanded educational opportunities for African Americans, fueled significant increases in black educational achievement and attainment, and brought African Americans closer to equality with whites by the 1980s. We situate the turn to accountability in a political context shaped by an increasingly conservative political environment, and examine three overlapping waves of test- based accountability that began in North Carolina in the late 1970s and spread throughout the region and the nation in the decades that followed: the minimum competency movement of the late 1970s and early 1980s, the standards based reforms of the 1980s, and the more comprehensive and coercive forms of high stakes testing in the 1990s. We argue that the southern political leaders who shaped U.S. educational policy turned to test-based accountability as a politically expedient alternative to the task of equalizing educational opportunities for African Americans. Civil rights organizations endorsed test-based accountability, but we find little evidence that test-based approaches improved African American educational outcomes. Opportunity policies, we conclude, did more to promote racial equality in educational achievement and attainment than test-based accountability. 

Abstract: An extensive literature has examined the presence and effectiveness of privatization in education, but most research focuses on charter schools and voucher programs. I instead examine contracting out for support services, a form of privatization that has re-emerged as an alternative policy response for districts looking to re-organize the delivery of public education. Using Michigan as an illustrative and informative case, I employ discrete-time hazard models to test the relationship between fiscal, political, and racial and the decision to privatize food, transportation, and custodial services. I find that race, measured as the proportion of district enrollment for Black students, is a key independent and previous undertheorized predictor of switching to privatized services. My dynamic modeling approach also confirms previous work that economic factors matter, whereas we continue to lack support for political ideology. 

The National School Lunch Program (NSLP) is the second-largest food assistance program in the U.S, serving lunches to 30 million students each year (United States Department of Agriculture Food and Nutrition Service 2018). Despite its prominence as a federal nutrition program, assessing its causal impact on student achievement is quite challenging. This is due to a lack of exogenous variation in program generosity as well as non-random selection into program participation. To certify their eligibility for subsidized meals, students’ families must have income below 130% of the federal poverty line to receive free meals or 185% of the federal poverty line to receive reduced-price meals.  Students can “directly certify” for subsidized meals if they participate in means-tested programs such as food stamps or the cash welfare program. Economic disadvantage itself is negatively correlated with achievement, making it difficult to disentangle the causal effect of subsidized meals on student achievement.

In this project, we propose to investigate the causal impact of subsidized school meals on student achievement by exploiting a policy change in the state of Michigan that added Medicaid to the set of programs that directly certify students for subsidized meals. Because of this change, students who participate in Medicaid are no longer required to fill out paperwork to certify their eligibility for subsidized meals; part of a growing effort to reduce administrative burdens through data linkages across state agencies. Direct certification reduces the burden on families to fill out paperwork each year and provides access to subsidized meals for families who may not have known they are eligible for or do not participate in the program for stigma, or other concerns. As we will show, including Medicaid as a direct certification program resulted in a large increase in the number of students who were certified for subsidized meals in the year after the policy change was introduced, providing a unique opportunity to study the causal impact of gaining access to such meals on student achievement.

Aim 1: The effect of Medicaid direct certification on eligibility for subsidized meals

In the first part of the project, we investigate how including Medicaid as a direct certification program impacted eligibility for subsidized meals and the number and type (free, reduced-price, or full price) of meals served in schools. As part of this descriptive analysis, we will describe the characteristics of students who participate in Medicaid and became eligible for free or reduced-price meals due to the policy change, and how they compare to students who gain eligibility for subsidized meals through other means. We will also analyze how the number and type of meals served in schools changes as a result of the policy, to understand whether Medicaid direct certification resulted in more meals served, or merely shifted the price of meals that students pay. This portion of the analysis will shed light on potential mechanisms through which Medicaid direct certification impacted student outcomes, namely through changes in nutrition or changes in family income resulting from a decrease in the cost of food.

Aim 2: Estimating the effect of CEP on student achievement

In the second portion of the analysis, we examine how Medicaid direct certification impacts eligibility and participation in the CEP program, and its subsequent impact on student achievement using the policy change as an exogenous shock to school and district eligibility for CEP.

Aim 3: Estimating the causal effect of subsidized meals on student achievement

In the final portion of the analysis, we estimate the causal effect of subsidized meals on individual student achievement using a student fixed effects design and leveraging the policy change that occurred in the 2017-18 school year in the state of Michigan to include Medicaid as part of direct certification. In theory, all of the students who certify for subsidized meals through this policy change were already eligible for the program—only students whose income is below 185% of the poverty threshold will certify for subsidized meals under Medicaid direct certification—but in order to receive subsidized meals, they would have needed to fill out paperwork to verify their income. The policy change eliminated the need to fill out this paperwork, and as we will show, increased the number of students certified eligible for subsidized meals. This portion of the analysis will allow us to estimate the causal effect of gaining individual access to subsidized meals on student achievement.  

This study seeks to unpack the implications of fiscal accountability for racial equity. While some have argued that the rise of fiscal accountability was a natural reaction to the dramatic increase in state investment in education during the second half of the 20th century, others have argued that accountability policies were racially motivated. I first test the relationship between racial progressivity of state aid and the decision to adopt Sanctioning laws, using plausibly exogenous variation in progressivity induced by school finance reforms and discrete-time hazard models. I then ask, are districts serving higher proportions of racially and ethnically marginalized students more likely to be sanctioned, and does this vary by sanction type? I compile a second original dataset containing the universe of districts sanctioned for financial reasons, separately for bankruptcy, dissolution, takeovers, and other sanctions. I use these data to descriptively document national trends in sanctions and test whether there are differences in the probability of being sanctioned by race and ethnicity.