Search EdWorkingPapers

Search for EdWorkingPapers here by author, title, or keywords.

K-12 Education

Di Xu, Qing Zhang, Xuehan Zhou.

This paper presents new experimental estimates of the impact of low-ability peers on own outcomes using nationally representative data from China. We exploit the random assignment of students to junior high school classrooms and find that the proportion of low-ability peers, defined as having been retained during primary school (“repeaters”), has negative effects on non-repeaters’ cognitive and non-cognitive outcomes. An exploration of the mechanisms shows that a larger proportion of repeater peers is associated with reduced after-school study time. The negative effects are driven by male repeaters and are more pronounced among students with less strict parental monitoring at home.

More →


Susana Claro, Susanna Loeb.

Existing research on self-management skills shows that measures of self-management predict student success. However, these conclusions are based on small samples or narrowly defined self-management measures. Using a rich longitudinal dataset of 221,840 fourth through seventh grade students, this paper describes self-management gaps across student groups, and confirms, at a large scale, the predictive power of self-management for achievement gains, even with unusually rich controls for students’ background, previous achievement, and measures of other social-emotional skills. Self-management is a better predictor of student learning than are other measures of socioemotional skills. Average growth in English language arts due to changing from a low to a high level of self-management is between 0.091 and 0.112 standard deviations, equivalent to almost 80 days of learning.

More →


Simon Burgess, Shenila Rawal, Eric S. Taylor.

This paper reports improvements in teacher job performance, as measured by student test scores, resulting from a program of (zero-) low-stakes peer evaluation. Teachers working at the same school observed and scored each other’s teaching. Students in randomly-assigned treatment schools scored 0.07σ higher on math and English exams (0.09σ lower-bound on TOT). Within each treatment school, teachers were further randomly assigned to roles: observer and observee. Teachers in both roles improved, perhaps slightly more for observers. The typical treatment school completed 2-3 observations per observee teacher. Variation in observations was generated partly by randomly assigning a low and high (2*low) dose of suggested number of observations. Benefits were quite similar across dose conditions. 

More →


Brendan Bartanen.

Using statewide data from Tennessee over more than a decade, this paper estimates the job performance returns to principal experience as measured by student, teacher, and principal outcomes. I find that principals improve substantially over time, evidenced by higher student achievement, higher ratings from supervisors, and lower rates of teacher turnover. However, improvement in student achievement as principals gain experience does not carry over when principals change schools. The returns to school-specific experience are largest for principals in high-poverty schools, highlighting the potential benefits of policies to improve the recruitment and retention of high-quality leaders in hard-to-staff environments.

More →


Corey A. DeAngelis.

I compare per pupil revenues, expenditures, and performance levels in public charter schools to district-run public schools in Texas for the 2017-18 school year. After controlling for several school and student characteristics, I find that public charter schools are funded around $1,700 (15 percent) less, and spend around $3,700 (28 percent) less, per pupil than district-run public schools. Public charter schools demonstrate cost-effectiveness advantages between 8 and 42 percent, depending on the model employed, over district-run public schools in Texas. I also find evidence to suggest per pupil spending is positively related to state testing outcomes for public charter schools, but not for district-run public schools.

More →


Carlos X. Lastra-Anadón, Paul E. Peterson.

Tiebout theorizes that local public services are provided more efficiently if costs are paid out of local revenues rather than by inter-governmental grants. But if local politics is not as pluralistic as Dahl has argued, citizens of higher socio-economic status will exercise greater influence, resulting in higher inequalities in service provision. We use administrative data to estimate the impacts of local revenue shares on individual performance of a nationally representative sample of over 140,000 U.S. eighth graders in math and reading. Causal effects are estimated with geographic discontinuity models and 2SLS models that use change in housing prices as an instrument. For every 10 percent increase in local revenue share, students perform about 0.05 standard deviations higher. Gains from local funding are less for disadvantaged students. Local financing affords better education for all but widens achievement gaps.

More →


Carolyn Abott, Vladimir Kogan, Stéphane Lavertu, Zachary Peskowitz.

We use close tax elections to estimate the impact of school district funding increases on operational spending and education outcomes across seven states. The analysis indicates that districts where tax levies passed spent $400-$500 more annually per pupil through 5-7 years after the election. They directed most of these funds toward increasing instructor salaries. These spending increases correspond to student test score gains of approximately 0.1 of a standard deviation and gains in graduation rates of approximately 3-4 percentage points. There is some evidence of diminishing returns, as these effects are driven by districts below the median in spending per pupil.

More →


Beth Schueler, Martin R. West.

Public support for school improvement policies can increase the success and durability of those reforms. However, little is known about public views on turnaround. We deployed questions and embedded experiments in a nationally representative 2017 survey (n=4,214) to uncover opinions regarding (a) which level of government should lead on turnaround and (b) state takeover of troubled districts. We find a large plurality prefers states play the greatest role in identifying and fixing failing schools. However, a substantial share prefers local governments increase their role. We find high levels of support for state takeover, yet support is greater in cases of financial mismanagement than academic underperformance. Those most likely to be directly affected express the least support for state takeover. 

More →


Hans Fricke, Susanna Loeb, Robert Meyer, Andrew Rice, Libby Pier, Heather Hough.

Recent attempts to measure schools’ influence on students' SEL show differences across schools, but whether these differences measure the true effect of schools is unclear. We examine the stability of school-by-grade effects on students' SEL across two years using a large-scale survey. Correlations among effects in the same grades across different years are positive but lower than those for math and English. Schools in the top or bottom of the effect distribution have more persistent impacts across years than those in the middle. Overall, the results suggest that these school effects measure real contributions to students' SEL. However, their low stability draws into question whether including school value-added measures of self-reported SEL in school performance systems would be beneficial.

More →


Oded Gurantz, Matea Pender, Zachary Mabel, Cassandra Larson, Eric Bettinger.

We examine whether virtual advising – college counseling using technology to communicate remotely – increases postsecondary enrollment in selective colleges. We test this approach using a sample of approximately 16,000 high-achieving, low- and middle-income students identified by the College Board and randomly assigned to receive virtual advising from the College Advising Corps. The offer of virtual advising had no impact on overall college enrollment, but increased enrollment in high graduation rate colleges by 2.7 percentage points (5%), with instrumental variable impacts on treated students of 6.1 percentage points. We also find that non-white students who were randomly assigned to a nonwhite adviser exhibited stronger treatment effects.

More →