November 14-15, 2013
Rosario Maria Ballatore, Bank of Italy, and Andrea Ichino and Margherita Fort, University of Bologna
Ballatore, Fort, and Ichino exploit rules of class formation to identify and estimate the causal effect of increasing the number of immigrants in a classroom while keeping class size constant. Since principals use class size to neutralize the possible detrimental effects of immigrants' inflows, this composition effect is the relevant policy parameter but it has been neglected in the literature so far. The authors' empirical analysis is guided by a model of the educational production function that allows for two types of students, natives and immigrants, who have a non-negative probability of generating disruption or positive externalities across or within groups. They show that the pure composition effect is substantial and negative at age seven (-12 percent in language and -7 percent in mathematics), but vanishes when children grow up. The reasons for this vanishing of effects, even at constant class size, remain to be explored.
Jason Cook and Richard Mansfield, Cornell University
In this paper, Cook and Mansfield use administrative panel data to decompose worker performance into components relating to general talent, task-specific talent, general experience, and task-specific experience. They consider the context of high school teachers, in which tasks consist of teaching particular subjects in particular tracks. Using the timing of changes in the subjects and levels to which teachers are assigned to provide identifying variation, the authors show that much of the productivity gains to teacher experience estimated in the literature are actually subject-specific. By contrast, very little of the variation in the permanent component of productivity among teachers is subject-specific or level-specific.
Peter Hinrichs, Georgetown University
Hinrichs documents how segregation between blacks and whites across colleges in the United States has evolved since the 1960s. He also explores potential channels through which changes are occurring, and uses recent data to study the issue of segregation within colleges. There are four main findings. First, white exposure to blacks has been rising since the 1960s, whereas black exposure to whites increased sharply in the late 1960s and early 1970s and has fluctuated since then. Meanwhile, black-white dissimilarity and the Theil index fell sharply in the late 1960s and early 1970s and have fallen more gradually since. Second, there has been regional convergence, although colleges in the South remain more segregated than those in any other region when measured by dissimilarity, by the Theil index, or by black exposure to whites. Third, a major channel for the decline in segregation is the declining share of blacks attending historically black colleges and universities. Finally, although there is segregation within universities, most segregation across major by university cells occurs across universities.
Steven Hemelt, University of North Carolina - Chapel Hill, and Kevin Stange, University of Michigan and NBER
Hemelt and Stange examine the effect of marginal price on students' credit-taking behavior and progress toward degree attainment using rich administrative data on all in-state students attending public universities in the state of Michigan. They find that full-time students facing zero marginal price attempt and complete about the same average number of credits as students at institutions that charge for each additional credit above the full-time threshold. However, zero marginal (or "flat") pricing does induce about 8 percent more students to attempt up to one additional class (that is, three credits), but the authors find little evidence that these attempted credits translate into earned ones. The moderate impact on attempted credits is largest among lower-achieving students, students eligible for free or reduced-price meals, and upper division students. Overall, the authors find little evidence that eliminating the marginal price associated with credits above the full-time minimum translates into more accumulated credits, a greater likelihood of persistence, or an increased likelihood of meeting "on-time" benchmarks toward timely degree completion. The authors discuss implications of these findings for students and post-secondary institutions.
Scott Imberman, Michigan State University and NBER, and Michael Lovenheim, Cornell University and NBER
Value-added data are an increasingly common evaluation tool for schools and teachers. Many school districts have adopted these methods and released the results publicly. In this paper, Imberman and Lovenheim study the release by the Los Angeles Times of value-added data in Los Angeles to identify how measured value-added is capitalized into housing prices. This analysis is the first in the U.S. school valuation literature to examine property value responses to a value-added information shock, which is of interest because this measure is less correlated with demographics than typical school quality measures. Unique to this setting also is the release of both school and teacher-level value-added data, which allows the authors to examine how property values respond to both types of information. Using a difference-in-differences methodology surrounding the release, the authors find that neither school nor teacher value-added scores are capitalized into home prices. Their results suggest that despite the contentiousness following these data releases, homeowners do not consider value-added models as currently constructed to be a relevant school quality measure on the margin.
Felipe Barrera-Osorio, Harvard University; David Blakeslee, Columbia University; Matthew Hoover, RAND Corporation; Leigh Linden, University of Texas at Austin and NBER; Dhushyanth Raju, the World Bank; and Stephen Ryan, University of Texas at Austin
Barrera, Blakeslee, Hoover, Linden, Raju, and Ryan evaluate the effects of publicly funded private primary schools on child enrollment in a sample of 199 villages in ten underserved districts of rural Sindh province in Pakistan. The program is found to significantly increase child enrollment, which increases by 30 percentage points in treated villages. There is no overall differential effect of the intervention for boys and girls because of similar enrollment rates in control villages. The authors find no evidence that providing greater financial incentives to entrepreneurs for the recruitment of girls leads to a greater increase in female enrollment than does an equal compensation scheme for boys and girls. Test scores improve dramatically in treatment villages, rising by 0.67 standard deviations relative to control villages.
Hanley Chiang, Melissa Clark, Sheena McConnell, Tim Silva, and Kathy Sonnenfeld, Mathematica Policy Research, Inc.
Teach For America (TFA) is an important but controversial source of teachers for hard-to-staff subjects in high-poverty U.S. schools. Chiang, Clark, and McConnell present findings from the first large-scale experimental study of secondary math teachers from TFA. They find that TFA teachers are more effective than other math teachers in the same schools, increasing student math achievement by an average of 0.07 standard deviations over the course of one school year. Addressing concerns about the fact that TFA requires only a two-year commitment, the authors find that novice TFA teachers are more effective than more experienced non-TFA teachers in the same schools.
Catharine Hill, Vassar College
Hill presents a simple model to demonstrate that increasing income inequality can contribute to the trends seen in American higher education and to the financial challenges that many colleges and universities are currently facing. The model is most appropriate for the selective non-profit higher education institutions. Given their commitment to the socioeconomic diversity of their students, the model demonstrates how increasing income inequality leads to higher tuition, higher costs, and higher financial aid than would otherwise be the case. A simple numerical example is also presented that estimates how much lower tuition, spending (proxying for costs), and financial aid would have been if household incomes in the United States had grown by the same aggregate amount between 1971 and 2009, but with no increase in income inequality. The policy implications are discussed.
Kristin Butcher, Wellesley College and NBER, and Patrick McEwan and Akila Weerapana, Wellesley College
Average grades in colleges and universities are markedly higher now than they were in the 1960s: the average grade point average (GPA) was about 2.4 in 1960 and about 3 in 2006. Many critics express concern that grade inflation (and compression at the top of the distribution) erodes incentives for students to learn and gives students, employers, and graduate schools poor information on students' absolute and relative abilities. They argue there is an implicit quid pro quo between grades and student evaluations of their professors which puts upward pressure on grades, and that grade inflation erodes the value of higher education. Butcher, McEwan, and Weerapana present a quasi-experimental evaluation of an anti-grade-inflation policy where a cap of B+ was placed on the course average grade. This cap was binding for high-grading departments (in the humanities and social sciences) and was not binding for low-grading departments (in science and economics), allowing for a difference-in-differences analysis of the effect of the policy on grades, receipt of honors, student sorting, and students' evaluations of their professors. The authors find that professors complied with the policy by reducing compression at the top of the grade contribution and that this had little effect on receipt of top honors, but did affect receipt of magna cum laude designations. They also found that the effect of the policy was to expand racial gaps in GPA. There is evidence that enrollments shrank in the capped departments, but no evidence of a shift in majors. Finally, students in capped courses were less likely to "strongly recommend" or "recommend" their professors and were more likely to report that they were "neutral" or did "not recommend" their professors.
Karthik Muralidharan, University of California at San Diego and NBER, and Nishith Prakash, University of Connecticut
Muralidharan and Prakash study the impact of an innovative program in the Indian state of Bihar that aimed to reduce the gender gap in secondary school enrollment by providing girls who attended secondary school with a bicycle, which would improve access to school. Using data from a large representative household survey, the authors employ a triple difference approach (using boys and the neighboring state of Jharkhand as comparison groups) and find that being in a cohort that was exposed to the Cycle program increased girls' age-appropriate enrollment in secondary school by 30 percent and also reduced the gender gap in age-appropriate secondary school enrollment by 40 percent. Parametric and non-parametric decompositions of the triple-difference estimate as a function of distance to the nearest secondary school show that the increases in enrollment mostly took place in villages where the nearest secondary school was farther away, suggesting that the mechanism for program impact was the reduction in the time and safety cost of school attendance made possible by the bicycle. The authors find that the Cycle program was much more cost-effective at increasing girls' enrollment than comparable conditional cash transfer programs in South Asia, suggesting that the coordinated provision of bicycles to girls may have generated externalities beyond the cash value of the program, including improved safety for girls cycling to school in groups, and changes in patriarchal social norms that proscribed female mobility outside the village, which inhibited female secondary school participation.
Kate Ambler, University of Michigan; Diego Aycinena, Universidad Francisco Marroquín; and Dean Yang, University of Michigan and NBER
Ambler, Aycinena, and Yang study the intersection of two research areas: educational subsidies, and migrant remittances. They implement a randomized experiment offering Salvadoran migrants subsidies for education, which are channeled directly to a beneficiary student in El Salvador chosen by the migrant. The subsidies, in the form of matching grants, lead to increases in educational expenditures, higher private school attendance, and lower labor supply of youths in Salvadoran households connected to migrant study participants. The authors find substantial "crowd in" of household educational investments, particularly for female students: for each $1 received by female beneficiary students, educational expenditures on that student increase by close to $5. There is no evidence of shifting of educational expenditures from other students in the household to the target student, and the subsidy has no substantial effect on remittances sent by the migrant.
Joshua Angrist, MIT and NBER; Erich Battistin, University of Padua; and Daniela Vuri, University of Rome
Using a Maimonides Rule identification strategy based on class-size cutoffs around 25, Angrist, Battistin, and Vuri document a payoff to smaller classes in Italian primary schools. These gains are driven mainly by schools in Southern Italy, suggesting a substantial return to class size for relatively poor residents of the Mezzogiorno. In addition to low socioeconomic status, the Mezzogiorno is distinguished by pervasive teacher cheating on standardized tests, a fact established by an experiment randomly assigning school monitors. The authors use Italy's Maimonides Rule to show that small classes facilitate teacher cheating, providing an alternative explanation for the causal effects of class size on test scores in Southern Italy. This motivates a causal model for achievement with two endogenous variables: class size, and proportion cheating. The model is identified by a combination of class size cutoffs and randomly assigned classroom monitors. The resulting estimates suggest that the effects of class size on measured achievement in Italian primary schools are driven entirely by the relationship between class size and teacher cheating. Models that estimate class size and cheating effects jointly generate precise zeros for the former.