Education Program Meets
May 9, 2013
Katja Kaufmann and Matthias Messner, Bocconi University, and Alex Solis, Uppsala University
Kaufmann, Messner, and Solis estimate the marriage market returns of attending a higher ranked ("elite") university by exploiting unique features of the Chilean university admission system. This system centrally allocates applicants based on their university entrance test score. It therefore constitutes an ideal context for identifying causal effects by using a regression discontinuity approach. Moreover, the Chilean context also provides the necessary data on the long-run outcome 'partner quality'. The authors find that being admitted to a higher ranked university has substantial returns in terms of partner quality. While their estimates show that there are marriage market returns for both sexes, they also indicate that they are more pronounced in the case of female students. They also analyze how marriage market returns vary across different types of university-degree programs and between different socioeconomic groups. They find that they are highest for students in top degree programs and for students from a privileged socioeconomic background.
Justine Hastings, Brown University and NBER, and Christopher Neilson and Seth Zimmerman, Yale University
Hastings, Neilson, and Zimmerman use administrative data from Chile from 1985 through 2005 to estimate the causal returns to a post-secondary degree as a function of field of study, course requirements, selectivity, and student socioeconomic status (SES). The data link high school and college records to labor market earnings from federal tax forms. They exploit hundreds of regression discontinuities from the centralized, score-based admissions system to estimate the causal impacts of interest. Returns are positive and significant only among more selective degrees. Returns are highly heterogeneous by field of study, with large returns in health, technology, law, and social sciences, but small-to-negative returns in arts, humanities, and education. They do not find evidence that vocational curriculum focus increases returns for less selective degrees. Nor do they find differential outcomes for students coming from low-versus-high SES backgrounds admitted to selective degrees.
Peter Bergman, University of California at Los Angeles
Bergman uses a field experiment to answer how information frictions between parents and their children affect human capital investment, and how reducing these frictions can improve student achievement. A random sample of parents in Los Angeles received detailed information about their children's academic progress. As in a standard principal-agent model, having more information allowed parents to induce more effort from their children, which translated into gains in achievement. However, having additional information also changed parents' beliefs and spurred demand for additional information from the school. Bergman concludes that relative to other interventions, additional information to parents potentially produces gains in achievement at a low cost.
Peter Hinrichs, Georgetown University
Teacher quality is a pressing public policy concern, yet there is little evidence on how effective schools are at selecting teachers. Hinrichs reports the results of an experiment that involved sending schools fictitious resumes with randomly-chosen characteristics in an attempt to determine what characteristics schools value when hiring new teachers. The results of the study suggest that an applicant's academic credentials have little impact on the likelihood of success, that schools may display a slight preference for female applicants, and that schools display a strong aversion to out-of-state applicants. Interestingly, some of these results may be stronger at private and charter schools than at traditional public schools.
Maria Fitzpatrick, Cornell University and NBER, and Damon Jones, University of Chicago and NBER
Fitzpatrick and Jones present new evidence on the effects of merit aid scholarship programs on residential migration and educational attainment using Census data on 24-to-32-year-olds in the United States from 1990 to 2010. Eligibility for merit aid programs slightly increases the propensity of state natives to live in-state, while also extending in-state enrollment into the late twenties. These patterns notwithstanding, the magnitude of merit aid effects is of an order of magnitude smaller than the population treated, suggesting that nearly all of the spending on these programs is transferred to individuals who do not alter educational or migration behavior.
Peter Arcidiacono and V. Joseph Hotz, Duke University and NBER, and Esteban Aucejo, London School of Economics
The low number of college graduates with science degrees -- particularly among under-represented minorities -- is of growing concern. Arcidiacono, Aucejo, and Hotz examine differences across universities in graduating students in different fields. Using student-level data on the University of California system during a period in which racial preferences were in place, they show significant sorting into majors based on academic preparation, with science majors at each campus having stronger credentials on average than their non-science counterparts. Students with relatively weaker academic preparation are significantly more likely to leave the sciences and to take longer to graduate at each campus. The authors show the vast majority of minority students would be more likely to graduate with a science degree and to graduate in less time if they had attended a lower ranked university. Similar results do not apply for non-minority students.
Jason Fletcher, Yale University and NBER; Stephen Ross, University of Connecticut; and Yuxiu Zhang, Yale University
Fletcher, Ross, and Zhang examine the demographic pattern of friendships among young people and its impact on their educational outcomes. They use friendship network data in the Add Health, and develop and estimate a reduced-form matching model to predict the formation of friendship links and to identify parameters based on across-cohort, within-school variation in the "supply" of potential friends. Their model provides novel evidence on the impact of small changes in peer demographic composition on the pattern of friendship links. It suggests, for example, that increases in the number of students from college educated family backgrounds leads to a greater likelihood of friendship links with students of that type among students whose mothers are college educated and among high school graduates. They also observe that increases in the share of African-American or Hispanic students lead to reductions in the incidence of cross-race friendships. They use the predicted friendship links from their model in an instrumental-variable analysis of the effects of friends' socioeconomic status as measured by parental education on own-grade-point-average outcomes. Although the conditional correlation between friendship composition and grade point average suggests large associations between friends' characteristics and own grades, this effect is robust only for females in the instrumental variable analysis. The authors then present evidence that the GPA effects are driven by science and English grades and that the mechanism is likely through self-esteem.