This dissertation explores a number of issues in the economics of education and health care. The first chapter, co-authored with Bassam Kadry, M.D., Joseph Orsini, Igor Popov, and Christopher Press, M.D., starts by noting that a large body of literature argues that a large fraction of the increasing health care costs in the United States is due to the increasing prevalence of obesity. Yet most of these studies focus only on the degree to which obesity is correlated with the number of procedures an individual undergoes. If obesity also increases the cost of a procedure, the previous estimates understate the medical cost of obesity. We illustrate this point in the context of hip replacement surgery. We find that obesity does increase the cost of surgery and that ignoring this margin underestimates the cost of obesity by approximately 5%. This suggests that potential savings from obesity reduction programs are larger than previously thought. In the second chapter, I use the fact that multiple elementary schools feed into the same middle school to demonstrate that the positive impact that teachers have on their own students spills over to affect their students' future peers. Although this indirect effect on any particular individual is small, a teacher impacts many more students indirectly than directly, so the indirect value is a sizable portion of a teacher's total value; I find that ignoring teachers' indirect effects underestimates their value by roughly 35%. Because the spillovers also affect teacher value added estimates, I develop a method of moments estimator of teacher value added that accounts for the spillovers and show that accounting for the spillovers does not have a large impact on the ranking of teachers in New York City. I conclude by showing that the spillovers occur within groups of students who share the same race and gender, which highlights the crucial importance that the social network plays in disseminating the effect. Finally, the third chapter, co-authored with Michael Dinerstein, attempts to shed some light on how incentivizing teachers to increase their students' test score affects the way teachers behave, since this is an extremely important input to many education policy decisions. Exploiting a dramatic change in the way teachers in New York City were granted tenure, we provide evidence that incentivizing teacher value added leads to small but statistically significant increases in the teachers value added scores. This increase, however, comes at a cost: the test score gains these teachers cause in their students fade-out more rapidly than similar gains caused by non-incentivized teachers, which suggests a change in the way incentivized teachers direct their attention. Yet, at this point, both of these effects appear small relative to the potential changes this policy could have on the teachers entering and exiting the profession.

In this report prepared for the New York City Chancellor's Office, we estimate short-term impacts of the New York City Renewal Schools Program based on the first two full years of program implementation. We utilize a novel method of multiple rating regression discontinuity design (MRRDD) that leverages the selection criteria used to select schools as Renewal or non-Renewal Schools. This method provides rigorous causal treatment effect estimates of the Renewal Schools Program, but only for the schools that barely qualified for the program. Our analysis suggests that the Renewal Schools program is helping to improve student attendance and reduce chronic absenteeism, while also increasing the amount of credits earned among high school students. These effects grew over the years of implementation. On the other hand, we find dropout rates worsened very slightly among Renewal Schools, but there was no systematic trend over the years in this effect. We find no statistically significant effects of the program on student achievement measures, nor do we find any statistically significant effects on any of the measures in the New York City Quality Review rubric. We also found evidence suggesting that among high school students the program impact was strongest at schools with greater levels of student economic needs.

There is an emerging consensus that teachers impact multiple student outcomes, but it remains unclear how to summarize these multiple dimensions of teacher effectiveness into simple metrics that can be used for research or personnel decisions. Here, we discuss the implications of estimating teacher effects in a multidimensional empirical Bayes framework and illustrate how to appropriately use these noisy estimates to assess the dimensionality and predictive power of the true teacher effects. Empirically, our principal components analysis indicates that the multiple dimensions can be efficiently summarized by a small number of measures; for example, one dimension explains over half the variation in the teacher effects on all the dimensions we observe. Summary measures based on the first principal component lead to similar rankings of teachers as summary measures weighting short-term effects by their prediction of long-term outcomes. We conclude by discussing the practical implications of using summary measures of effectiveness and, specifically, how to ensure that the policy implementation is fair when different sets of measures are observed for different teachers.

Researchers developed cost-estimating tools and estimated future construction costs for Puerto Rico's recovery from Hurricane María. This report documents their approach, data, findings, and recommendations.

Teachers, like the students they serve, never stop learning. In-service teacher professional development (PD) gives educators opportunities to learn more about pedagogy and improve their own instruction methods to boost students' academic and social and emotional outcomes. Districts make a significant financial investment to provide teacher PD, and research on the impact of teacher PD on teacher instructional practices and student education outcomes has been mixed. In addition, there are only a few studies that examine the impact of teacher PD using rigorous empirical evaluation designs (such as randomized control trials) and consider PD across multiple contexts (public versus charter schools). The authors evaluated the Chicago Collaborative, a teacher PD program that is aligned to Common Core Standards and implemented by Leading Educators, a national nonprofit organization that partners with districts and charter management organizations to help teachers develop the leadership skills that they need to successfully transition from leading students to leading their peers. The authors conducted a randomized control trial evaluation using data from 40 schools across three school districts in the Chicago area during the 2018-2019 and 2019-2020 school years. They examined how the Chicago Collaborative program was implemented and whether the program impacted student achievement. The authors found that the Chicago Collaborative was successfully delivered, despite the challenges posed by the coronavirus pandemic at the end of the research period in 2020. The authors also found robust evidence that the Chicago Collaborative increased student test scores.

The author discusses the use of value-added models (VAMs) to estimate teacher contribution to student progress, account for students' prior characteristics, and enable relative judgments, as well as the limitations of VAMs.

Nearly every public school teacher in the United States receives some form of feedback based on classroom observations, and these assessments are used by policymakers and administrators as indicators of teacher quality and input for professional development purposes. The TNTP Core Teaching Rubric, introduced in 2014, is unlike others in that it is scored based on student behavior rather than teacher actions. In this report, RAND researchers assess whether the TNTP Core Teaching Rubric produces scores that are representative of teachers' overall instructional practices and whether raters' content expertise influence scores on TNTP Core.For this study, volunteer teachers of mathematics and English language arts (ELA) from more than 20 districts or charter school networks across eight states had their classroom instruction videotaped three times. This sample was complemented with a sample of approximately 100 4th-and 5th-grade teachers drawn from the Measures of Effective Teaching study for whom videotaped instruction was available. Raters trained in the TNTP Core Teaching Rubric then scored the videos from both samples. Analyzing the raters' scores, RAND researchers found a modest relationship between teachers' TNTP Core scores and student achievement gains in ELA, and no statistically or practically significant relationships between TNTP Core scores and math achievement gains. Further, raters struggled to agree on their judgments about the quality of a lesson, suggesting a considerable amount of uncertainty about the extent to which TNTP Core scores represent teachers' overall instructional practices. The report includes a number of recommendations to improve the use of TNTP Core.

This report presents an evaluation of Neighborhood Court, a restorative justice diversion program run by the District Attorney in San Francisco. A randomized trial indicates that the program reduces recidivism (although this result is statistically insignificant), a discrete choice experiment shows demand and public willingness to pay for its features, and a qualitative analysis demonstrates positive perceptions of the program among stakeholders.

What happens when employers would like to screen their employees but only observe a subset of output? We specify a model in which heterogeneous employees respond by producing more of the observed output at the expense of the unobserved output. Though this substitution distorts output in the short-term, we derive three sufficient conditions under which the heterogenous response improves screening efficiency: 1) all employees place similar value on staying in their current role; 2) the employees' utility functions satisfy a variation of the traditional single-crossing condition; 3) employer and worker preferences over output are similar. We then assess these predictions empirically by studying a change to teacher tenure policy in New York City, which increased the role that a single measure -- test score value-added -- played in tenure decisions. We show that in response to the policy teachers increased test score value-added and decreased output that did not enter the tenure decision. The increase in test score value-added was largest for the teachers with more ability to improve students' untargeted outcomes, increasing their likelihood of getting tenure. We find that the endogenous response to the policy announcement reduced the screening efficiency gap -- defined as the reduction of screening efficiency stemming from the partial observability of output -- by 28%, effectively shifting some of the cost of partial observability from the post-tenure period to the pre-tenure period.

What happens when employers would like to screen their employees but only observe a subset of output? We specify a model in which heterogeneous employees respond by producing more of the observed output at the expense of the unobserved output. Though this substitution distorts output in the short-term, we derive three sufficient conditions under which the heterogenous response improves screening efficiency: 1) all employees place similar value on staying in their current role; 2) the employees' utility functions satisfy a variation of the traditional single-crossing condition; 3) employer and worker preferences over output are similar. We then assess these predictions empirically by studying a change to teacher tenure policy in New York City, which increased the role that a single measure -- test score value-added -- played in tenure decisions. We show that in response to the policy teachers increased test score value-added and decreased output that did not enter the tenure decision. The increase in test score value-added was largest for the teachers with more ability to improve students' untargeted outcomes, increasing their likelihood of getting tenure. We find that the endogenous response to the policy announcement reduced the screening efficiency gap -- defined as the reduction of screening efficiency stemming from the partial observability of output -- by 28%, effectively shifting some of the cost of partial observability from the post-tenure period to the pre-tenure period.