« Back to Results

CSMGEP Dissertation Session

Paper Session

Saturday, Jan. 5, 2019 2:30 PM - 4:30 PM

Atlanta Marriott Marquis, M102
Hosted By: American Economic Association & Committee on the Status of Minority Groups in the Economics Profession
  • Chair: Kalena Cortes, Texas A&M University

Who Does Aid Help? Examining Heterogeneity in the Effect of Student Aid On Achievement

Breyon Williams
University of South Carolina


Does financial aid impact achievement differently for low- and high-income students? I exploit the eligibility requirements of a state merit-aid program that awards additional aid to STEM majors. A triple-difference design, using administrative data from a large institution and exploiting differences over time, by merit-aid recipient status, and by major type (STEM or non-STEM), shows that student aid increases the GPAs and graduation prospects of low-income students but has little impact on high-income students. Additional analysis suggests that reduction in student part-time work may be a potential mechanism for this differential effect. These results suggest that merit aid programs could be targeted more effectively than most currently are.

Refugee Students and Peer Effects

Maria Camila Morales
Georgia State University


The number of refugees around the world is at a historical record high. Much of the recent political debate on whether to change the number of incoming refugees centers on the perceived unintended consequences of refugees, and policies that can help them assimilate into local communities. Approximately 40 percent of the refugees who enter the U.S. are under 19; therefore, much of the cultural assimilation takes place in local schools, making education a critical context to study the impact of refugee integration. I exploit natural variation in the share of refugees within schools and across grades to estimate the causal effect of having refugee peers on the test performance of native students in ELA and Math. I also examine non-academic outcomes such as attendance and disciplinary incidents. Results from this paper will serve as evidence on the externalities of refugee integration into local communities, particularly how this population affects native students. If refugee peer exposure matters, these findings can inform discussions on the optimal distribution of students in classrooms, in an effort to improve overall achievement. Moreover, results may also shed light on the spillover effects that can be obtained from improving the academic outcomes of refugee students.

The (Perceived) Cost of Being Female: An Experimental Investigation of Strategic Responses to Discrimination

Mackenzie Alston
Texas A&M University


Occupational gender segregation and wage gaps may be caused by women applying less than men or women being discriminated against during the hiring process. I estimate how much discrimination job applicants expect to face by giving them the opportunity to include or exclude gender from their resumes in an online experiment where subjects perform a stereotypically male task. On average, men were willing to forfeit 9.4% of their earnings from being hired to use resumes that included gender while women were willing to forfeit 10.0% of their earnings to use resumes that excluded gender. However, workers did not earn significantly more when they included rather than excluded their gender. Therefore, both men and women overestimated the cost of being female in this study, suggesting that men (women) apply for stereotypically male jobs at higher (lower) rates than optimal.

Democracy, Genes and the Male Survival Disadvantage

Marie Christelle Mabeu
University of Ottawa
Roland Pongou
University of Ottowa


Boys have a survival disadvantage in most societies. This paper examines whether and how improvement in the quality of democratic institutions affects sex differences in infant mortality. Using data on more than 3 million births from sub-Saharan African countries, our identification strategy exploits within-mother variation in the quality of institutions. The main finding demonstrates that the male survival disadvantage in infant mortality falls by 0.31 percentage points, 20% of the sample mean, when we move from autocracy to democracy. Analyzing the channels through which this effect operates, we find that better democratic institutions expand the likelihood of tetanus immunization, access to prenatal care services, breastfeeding practices, and normal birth weight, all of which are associated with stronger health benefits for boys than for girls. Moreover, using twins, we find that better democratic institutions significantly constrain genetic influences on male mortality in early ages. We empirically reject the hypothesis that hormonal (testosterone) transfer could be driving our findings.
Susan Dynarski
University of Michigan
Marie T. Mora
University of Texas-Rio Grande
Christine Piette Durrance
University of North Carolina-Chapel Hill
Melanie Guldi
University of Central Florida
JEL Classifications
  • I0 - General
  • J0 - General