Education Program Meets

November 15-16, 2012
Caroline M. Hoxby of Stanford University, Organizer

Guido Schwerdt, Ifo Institute for Economic Research, and Martin West, Harvard University

The Effects of Test-based Retention on Student Outcomes over Time: Regression Discontinuity Evidence from Florida

Many American states now require that students who do not demonstrate basic reading proficiency at the end of third grade be retained and provided with remedial services. Schwerdt and West exploit a discontinuity in the probability of third grade retention under Florida's test-based promotion policy to study the causal effect of retention on student outcomes over time. Although conventional OLS estimates suggest negative effects of retention on achievement, regression discontinuity estimates indicate large positive effects on achievement and a reduced probability of retention in subsequent years. The achievement gains from test-based retention fade out over time, however, and are statistically insignificant after six years.

Benjamin Castleman, Harvard University, and Bridget Long, Harvard University and NBER

Looking Beyond Enrollment: The Causal Effect of Need-Based Grants on College Access, Persistence, and Graduation

Gaps in average college success among students of differing backgrounds have persisted in the United States for decades. One of the primary ways that federal and state governments have attempted to ameliorate such gaps is by providing need-based financial aid to low-income students. Castleman and Long examine the impact of eligibility for the Florida Student Access Grant (FSAG) on a range of college outcomes. Exploiting the cutoff in the index used to measure a family's ability to pay for college and determine grant eligibility, they use a regression-discontinuity (RD) strategy to estimate the causal effect of being eligible for the grant. They investigate whether being eligible for a need-based grant increases the probability that students enter college, remain continuously enrolled, accumulate more college credits, and ultimately earn a degree. As have other researchers,they find that grant eligibility had a positive effect on attendance. Moreover, grant aid increased short-term persistence and the cumulative number of credits students earned over time. Most importantly, they find that FSAG increased the likelihood of a receiving a bachelor's degree within six years from a public college or university by 22 percent among students near the eligibility cut-off. The effects are largest for academically high-achieving students.

Lindsay Daugherty and Francisco Martorell, RAND Corporation, and Isaac McFarlin, University of Michigan

Percent Plans, Automatic Admissions, and College Enrollment Outcomes

In 1997, the Texas Legislature passed House Bill 588 – also known as the Texas Top Ten Percent Law – guaranteeing automatic admission to all state-funded universities for Texas students in the top 10 percent of their high school class. Automatic admissions policies remain controversial, and the effects of these policies on college enrollment and choice remain unclear. Using regression discontinuity methods and data on six cohorts of graduates from a large urban school district, Daugherty, Martorell, and McFarlin examine the effect of eligibility for automatic admission on college enrollment and persistence. They find that the Top Ten Percent Law has a substantial impact on enrollment at Texas flagship universities and increases the total number of semesters enrolled at a flagship university four years after high school graduation. This increase in flagship enrollment appears to displace enrollment in private or out-of-state universities, and they find no effect on college enrollment overall or on the quality of college attended. They find evidence of effects on flagship enrollment for both white and minority students. However, these effects are concentrated in schools that send large (relative to the district) fractions of graduates to college, suggesting that automatic admissions may have little effect on the outcomes of students in the most disadvantaged schools.

Sarah Cohodes and Joshua Goodman, Harvard University

First Degree Earns: The Impact of College Quality on College Graduation Rates

Cohodes and Goodman use a Massachusetts merit aid program to provide the first clear causal evidence on the impact of college quality on students’ post secondary enrollment decisions and rates of degree completion, where college quality is defined by a variety of measures, including on-time graduation rates. High school students with test scores above multiple thresholds were granted tuition waivers at in-state public colleges of lower quality than the average alternative available to such students. A binding score regression discontinuity design comparing students just above and just below these thresholds yields two main findings. First, students are remarkably willing to forego college quality for relatively small amounts of money. Second, choosing a lower quality college significantly lowers on-time completion rates, a result driven by high skilled students who otherwise would have attended higher quality colleges. For the marginal student, enrolling at an in-state public college lowered the probability of graduating on time by more than 40 percent. The low completion rates of scholarship users imply that the program had little impact on the in-state production of college degrees. More broadly, these results suggest that the critically important task of improving college quality requires steps beyond merely changing the composition of the student body.

Maria Fitzpatrick, Cornell University, and Michael Lovenheim, Cornell University and NBER

Early Retirement Incentives and Student Achievement

Given the prevalence of districts offering older teachers incentives to retire early in order to fill budget gaps, and the rising average age of the teachers, the composition of teachers will change dramatically in coming years. However, there currently is no information on how these changes in the teacher workforce will affect student achievement. Fitzpatrick and Lovenheim use exogenous variation in teacher retirement behavior driven by an early retirement incentive program in Illinois in the mid-1990s to identify the effects of large-scale teacher retirements on student achievement. They find that the program did not reduce test scores; likely, it increased them. The positive effects were most pronounced in low-SES and lower-performing schools. These results suggest that districts may be able to lower costs without damaging student outcomes through early retirement programs.

Caroline Hoxby, and Sarah Turner, University of Virginia and NBER

Expanding College Opportunities for Low-Income, High-Achieving Students

Joshua Angrist and Parag Pathak, MIT and NBER, and Christopher Walters, MIT

Explaining Charter School Effectiveness (NBER Working Paper No. 17332)

Estimates using admissions lotteries suggest that urban charter schools boost student achievement, while charter schools in other settings do not. Using the largest available sample of lotteried applicants to charter schools, Angrist, Pathak, and Walters explore student-level and school-level explanations for this difference in Massachusetts. In an econometric framework that isolates sources of charter effect heterogeneity, they show that urban charter schools boost achievement well beyond that of urban public school students, while non-urban charters reduce achievement from a higher baseline. Student demographics explain some of these gains since urban charters are most effective for non-whites and low-baseline achievers. At the same time, non-urban charter schools are uniformly ineffective. These estimates also reveal important school-level heterogeneity within the urban charter sample. A non-lottery analysis suggests that urban charters with binding, well-documented admissions lotteries generate larger score gains than under-subscribed urban charter schools with poor lottery records. Finally, they link charter impacts to school characteristics, such as peer composition, length of school day, and school philosophy. The relative effectiveness of urban lottery-sample charters is accounted for by these schools' embrace of the No Excuses approach to urban education.

Peter Arcidiacono, Duke University and NBER, and Cory Koedel, University of Missouri

Race and College Success: Evidence from Missouri

Conditional on enrollment, African-American entrants at four-year public universities are much less likely to graduate, and to graduate in STEM fields, than white entrants. Using administrative micro data from Missouri, Arcidiacono and Koedel show that the success gaps between African-American and white students in college can be explained by three factors: 1) racial differences in how students sort to universities and majors; 2) racial differences in high-school quality prior to entry; and 3) racial differences in other observed pre-entry skills. They decompose the success gaps between African-Americans and whites to identify the relative importance of these three factors. Even holding racial differences in high-school quality and pre-entry skills fixed, they find that a non-negligible fraction of the racial gap in graduation rates can be explained by differences in student sorting across universities and majors (10 to 20 percent). Differences in observed measures of pre-entry skills – primarily students' high-school class rankings conditional on high school of attendance – are consistently the most important determinants of the success gaps by race. Differences in pre-entry skills explain a larger share of the graduation gap for men than for women, and most of the racial gaps in STEM attainment (conditional on STEM entry) for both genders.

Timothy Bartik and Marta Lachowska, W.E. Upjohn Institute for Employment Research

The Short-Term Effects of the Kalamazoo Promise Scholarship on Student Outcomes

To study whether college scholarships can be an effective tool in raising students' performance in secondary school, Bartik and Lachowska use one aspect of the Kalamazoo Promise that resembles a quasi-experiment. The surprise announcement of the scholarship created a large change in expected college tuition costs that varied across different groups of students based on past enrollment decisions. This variation is arguably exogenous to unobserved student characteristics. They estimate the effects of this change by a set of "difference-in-differences" regressions where they compare the change in student outcomes in secondary school across time for different student "length of enrollment" groups. They find positive effects of the Kalamazoo Promise on Promise-eligible students, large enough to be deemed important—about a 9 percent increase in the probability of earning any credits and one less suspension day per year. They also find large increases in GPA among African-American students.

Kasey Buckles, University of Notre Dame; Ofer Malamud, University of Chicago and NBER; Melinda Morrill, North Carolina State University; and Abigail Wozniak, University of Notre Dame and NBER

The Effect of College Education on Health

Buckles, Malamud, Morrill, and Wozniak exploit exogenous variation in college completion induced by draft-avoidance behavior during the Vietnam War to examine the impact of college completion on adult mortality. Their preferred estimates imply that increasing college completion rates from the level of the state with the lowest induced rate to the state with the highest rate would decrease cumulative mortality by 28 percent relative to the mean. Most of the reduction in mortality is from deaths due to cancer and heart disease. They also explore potential mechanisms, including differential earnings, health insurance, and health behaviors, using data from the Census, ACS, an NHIS.

Karthik Muralidharan, University of California at San Diego and NBER

The Aggregate Effect of School Choice: Evidence from a Two-stage Experiment in India

Muralidharan presents experimental evidence on the impact of a program that provided economically disadvantaged children in the Indian state of Andhra Pradesh (AP) with a scholarship, allowing them to attend a private school of their choice. The design of the program featured a unique two-stage lottery, which creates both a student-level and a market-level experiment and allows us to study both the individual and the aggregate effects of school choice. Compared to teachers in government-run schools, private school teachers are paid much lower salaries and have lower levels of formal education and training. The mean annual cost per student in the private schools in this sample is less than a third of the costs in public schools. On the other hand, private schools have a longer school day, a longer school year, smaller class sizes, lower teacher absence, higher teaching activity, and better school hygiene. Private schools spend significantly less instructional time on Telugu (native language of AP) and Math, and instead spend more time on English, Science, Social Studies, and especially Hindi. At the end of four years of the school choice program, the author finds that lottery winners do not have higher test scores than lottery losers on tests of Telugu, Math, English, Science, and Social Studies, but they do score significantly higher in Hindi. There is evidence of heterogeneity of impact by medium of instruction in the private school, but not by most other demographic characteristics. There is also some evidence suggesting that the impact of the voucher may have been greater in areas with more choice and competition. There is no evidence of significant spillovers on students who do not apply for the voucher and who remain in the government schools, or on students who start out in private schools to begin with, suggesting that the program had no adverse effects on these groups.

NBER Videos

National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us