Search for EdWorkingPapers here by author, title, or keywords.
In education settings, treatments are often non-randomly assigned to clusters, such as schools or classrooms, while outcomes are measured for students. This research design is called the clustered observational study (COS). We examine the consequences of common support violations in the COS context. Common support violations occur when the covariate distributions of treated and control units do not overlap. Such violations are likely to occur in a COS, especially with a small number of treated clusters. One common technique for dealing with common support violations is trimming treated units. We demonstrate how this practice can yield nonsensical results in some COSs. More specifically, we show how trimming the data can result in an uninterpretable estimand. We use data on Catholic schools to illustrate concepts throughout.
Using detailed classroom-level data for North Carolina, we build on previous research to examine racial gaps in access to high-quality teachers. We calculate the exposure of White, Black and Hispanic students to teachers with various characteristics in 4th grade, 7th grade math and English, and 10th grade math and English. We find that across the state White students enjoy sizable advantages over both Black and Hispanic students in the form of higher exposure to teachers with strong credentials and lower exposure to teachers with weak credentials. Remarkably, we also find this pattern of White advantage in most individual counties, with the largest White advantage occurring in the largest counties by enrollment. A decomposition of the White advantages shows that the bulk of them can be attributed to differences across counties and differences between schools within counties. Only in 10th grade are differences across classrooms within schools important in explaining the White advantage.
Children exposed to Interparental Verbal Conflict (IPVC) exert negative spillovers on their peers. Using nationally representative survey data from middle schools in China, focusing on schools that randomly assign students into classrooms, and using both (1) within-school, across-classroom variation and (2) within-student, year-to-year variation to identify effects, we find that being assigned to classes where more classmates experience IPVC reduces mental wellbeing, diminishes self-confidence, lowers social engagement, and increases the likelihood of problem behaviors. Effects operate by damaging relationships between classmates. There is no evidence of impacts on test scores or teacher’s outcomes.
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.
Philanthropic investment in education has evolved considerably over the past several decades. This paper provides early evidence of another distinct adaptation, which we dub design philanthropy. In contrast to the macro-level structural reforms recently supported by large foundations, design philanthropy seeks to directly influence the instructional core. We describe the broad contours and characteristics of design philanthropy, which employs a centralized management and design system to support a decentralized approach to implementation. Through a case study of one design philanthropy’s reform initiative, we explore how participants experience this emergent process and manage a series of tensions inherent in the approach.
An administrative rule allowed students who failed an exam to retake it shortly after, triggering strong `teach to the test' incentives to raise these students' test scores for the retake. We develop a model that accounts for truncation and find that these students score 0.14 standard deviations higher on the retest. Using a regression discontinuity design, we estimate thirty percent of these gains persist to the following year. These results provide evidence that test-focused instruction or `cramming' raises contemporaneous performance, but a large portion of these gains fade-out. Our findings highlight that persistence should be accounted for when comparing educational interventions.
This study investigates whether a principal’s likelihood of hiring a teacher of color is sensitive to the racial composition of students in the school. We used an administrative dataset from Texas including 59,157 principal observations and 662,997 teacher observations spanning 2000 to 2017 in order to consider whether or not the disappearing diversity from a majority white school is a factor in principals’ decisions to hire teachers of color. We examined the hiring patterns of principals within schools where 50% of the students were white and compared the probability that a nonwhite teacher would be hired as the homogeneity of the student body increased (that is, as increasing proportions of the student body were white). We found that white principals were less likely to hire teachers of color as the proportion of white students approached 100%. This study provides initial evidence that teacher hires are not only sensitive to the principal’s race but also to the racial composition of the student body. Specifically, as the diversity of the student body disappears, so too does the principal’s likelihood of hiring a teacher of color.
Due to the COVID-19 Pandemic, the decision to reopen schools for in-person instruction has become a highly salient policy issue. This study examines what overall factors drive public support for schools re-opening in person, and whether members of the public are any more or less willing to comply with school re-opening decisions based on their own preferences and the level of government from which the order comes. Through two rounds of national surveys with an embedded experiment, I find consistent evidence that 1) trust in information from elites - not contact with COVID - best explain preferences for reopening, 2) political ideology and racial and class identification help explain preferences as well, and 3) the President of the United States is best positioned to generate compliance with a school reopening mandate. However, low public trust in the president makes the public significantly less likely to comply. This study suggests that politics - not public health - drives public support for schools reopening and compliance with government orders to reopen.
Millions of high school students who take an Advanced Placement (AP) course in one of over 30 subjects can earn college credit by performing well on the corresponding AP exam. Using data from four metro-Atlanta public school districts, we find that 15 percent of students’ AP courses do not result in an AP exam. We predict that up to 32 percent of the AP courses that do not result in an AP exam would result in a score of 3 or higher, which generally commands college credit at colleges and universities across the United States. Next, we examine disparities in AP exam-taking rates by demographics and course taking patterns. Most immediately policy relevant, we find evidence consistent with the positive impact of school district exam subsidies on AP exam-taking rates. In fact, students on free and reduced-price lunch (FRL) in the districts that provide a higher subsidy to FRL students than non-FRL students are more likely to take an AP exam than their non-FRL counterparts, after controlling for demographic and academic covariates.
Before the 2020-21 school year, educators, policymakers, and parents confronted the stark and uncertain trade-offs implied by the health, educational, and economic consequences of offering instruction remotely, in person, or through a hybrid of the two. Most public schools in the U.S. chose remote-only instruction and enrollment fell dramatically (i.e., a loss of roughly 1.1 million K-12 students). We examine the impact of these choices on public-school enrollment using unique panel data that combine district-level enrollment trajectories with information on their instructional modes. We find offering remote-only instead of in-person instruction reduced enrollment by 1.1 percentage points (i.e., a 42 percent increase in disenrollment from -2.6 to -3.7 percent). The disenrollment effects of remote instruction are concentrated in kindergarten and, to a lesser extent, elementary schools. We do not find consistent evidence that remote instruction influenced middle or high-school enrollment or that hybrid instruction had an impact.