Education Program Meeting

May 11, 2012
Caroline M. Hoxby of Stanford University, Organizer

Kalena Cortes, Texas A&M University; Joshua Goodman, Harvard University; and Takako Nomi, University of Chicago

Doubling Up: The Long-Run Impacts of Remedial Algebra on High School Graduation and College Enrollment

Success or failure in freshman algebra has long been thought to have a strong impact on subsequent high school outcomes. Cortes, Goodman, and Nomi study a remedial algebra policy implemented by the Chicago Public Schools (CPS) for cohorts entering high school in 2003 and 2004. Students scoring below the national median on an eighth grade exam were assigned in ninth grade to a double-dose algebra course, which doubled instructional time and increased tracking by ability. This assignment rule allows difference-in-difference and regression discontinuity estimates of average and local average treatments effects. Using longitudinal data that tracks students from eight grade to college enrollment, the authors confirm prior work showing positive short-run impacts on freshman GPA, passing rates, and math scores. They show long-run positive and substantial impacts of remediation on college entrance exam scores, high school graduation rates, and college enrollment rates. This is the first evidence they know of to demonstrate long-run impacts of remediation in an American urban high school setting.


Lesley Turner, Columbia University

The Incidence of Student Financial Aid: Evidence from the Pell Grant Program

The federal Pell Grant Program provides billions of dollars in subsidies to low-income college students to increase affordability and access to higher education. Turner estimates the economic incidence of these subsidies using regression discontinuity (RD) and regression kink (RK) designs. Turner shows that 16 percent of all Pell Grant aid is passed-through to schools in the form of higher effective prices. However, the extent and pattern of pass-through vary by institutional control and selectivity. While RK estimates suggest that schools capture Pell Grant aid through price discrimination, RD estimates imply the opposite result, that schools supplement Pell Grants with increases in institutional aid. Turner reconciles these disparate findings through a framework in which the treatment of Pell Grant aid is multidimensional: students receive an additional dollar of Pell Grant aid and are also labeled as Pell Grant recipients. RD estimates confound the effects of these dimensions, which have opposite impacts on schools' pricing decisions. Developing a combined RD/RK approach allows one to separately identify schools' willingness to pay for students categorized as needy and the pricing response to outside subsidies.


Brian Jacob, University of Michigan and NBER, and Brian McCall and Kevin M. Stange, University of Michigan

The Consumption Value of Postsecondary Education

Education provides both investment and consumption benefit: the former are realized after schooling is completed, but the latter accrue only while schooling is actually taking place. Jacob, McCall, and Stange quantify the importance of consumption-value considerations to schooling decisions in the context of higher education and then examine the implications for the demand-side pressure that colleges face in the market for students. They estimate a discrete-choice model of college demand using micro data from the high school classes of 1992 and 2004, matched to extensive information on all four-year colleges in the United States. They find that most students do appear to value college attributes, which the authors categorize as "consumption," including college spending on student activities, sports, and dormitories. In fact, students appear to be more willing to pay for these non-academic aspects of colleges than for typical academic aspects, such as spending on instruction. The estimates suggest that this taste for consumption amenities is broad-based among many student groups, whereas the taste for academic quality is confined only to the high achieving. Consequently, policies that reallocate financial resources away from these non-academic aspects to instruction would not enable most schools to attract more or better students, as some policymakers suggest. The implication is that most colleges face a trade-off: increases in instructional spending will attract high achieving students, but may deter enrollment from a broader student body. Increases in service spending, however, will attract all types of students (but disproportionately lower-achieving students). Because student preferences for college attributes are very different, however, colleges face different incentives for changing their characteristics depending on their current student body and those they are trying to attract.

David Deming, University of North Carolina at Charlotte; Steve Billings, Harvard University; and Jonah E. Rockoff, Columbia University and NBER

School Segregation Educational Attainment and Crime: Evidence from the end of Busing in Charlotte-Mecklenburg

Deming, Billings, and Rockoff study how the end of race-based busing in Charlotte-Mecklenburg schools ("CMS") affected academic achievement, educational attainment, and young adult crime. CMS was prohibited in 2001 from using race in assigning students to schools. School boundaries were redrawn dramatically to reflect the surrounding neighborhoods, and about half of the CMS students received a new school assignment. Using addresses that were recorded prior to the policy change, the authors compare students who lived on opposite sides of a newly drawn boundary in the same neighborhood and therefore experienced a change in school racial composition. They find that the re-segregation of CMS schools widened racial gaps in middle school and high school math scores and were accompanied by large increases in crime for poor minority males. They conclude that the end of busing widened racial inequality, despite efforts by CMS to mitigate the impact of increases in segregation.


Matthew Wiswall, New York University; and Basit Zafar, Federal Reserve Bank of New York

Determinants of College Major Choice: Identification using an Information Experiment

Wiswall and Zafar study the determinants of college major choice using an experimentally generated panel of beliefs, obtained by providing students with information on the true population distribution of various major-specific characteristics. Students logically revise their beliefs in response to the information, and their subjective beliefs about future major choice are associated with beliefs about their own earnings and ability. The authors estimate a rich model of college major choice using the belief data. While earnings are a significant determinant of major choice, tastes - which are heterogeneous - are the dominant factor in the choice of major. The authors also investigate gender differences in major choice.


Dan Goldhaber, University of Washington; and Duncan Chaplin, Mathematica Policy Research

Assessing the "Rothstein Test" Does It Really Show Teacher Value-Added Models Are Biased?

In a provocative and influential paper, Jesse Rothstein (2010) finds that standard value-added models (VAMs) suggest implausible and large future teacher effects on past student achievement, a finding that obviously cannot be viewed as causal. This is the basis of a falsification test (the Rothstein falsification test) that appears to indicate bias in VAM estimates of current teacher contributions to student learning. Rothstein's finding is significant because there is considerable interest in using VAM teacher effect estimates for high-stakes teacher personnel policies, and the results of the Rothstein test cast considerable doubt on the notion that VAMs can be used fairly for this purpose. Goldhaber and Chaplin show that the Rothstein test does show that students are tracked to teachers, but the tracking could be based on lagged achievement, the key control variable used in most VAMs. These results indicate that the Rothstein test does not appear to provide additional useful guidance regarding the efficacy of VAMs, suggesting a much more encouraging picture for those wishing to use VAM teacher effect estimates for policy purposes.