UC Santa Barbara

This dissertation consists of three essays in applied microeconomics that address topics of human and social capital accumulation. The first essay addresses the topic of women and STEM (science, technology, engineering and math.) Women are substantially less likely than men to graduate college with a STEM degree. This paper investigates whether class composition can help explain why women are disproportionately more likely to fall out of the STEM ``pipeline''. Identification comes from a standardized enrollment process at a large public university that randomly assigns freshmen to different mandatory introductory chemistry lectures. Using administrative data, I find that women who are enrolled in a class with higher ability peers are less likely to graduate with a STEM degree, while men's persistence in STEM is unaffected by class composition. I also show that the decline for women is most pronounced for those in the bottom third of the ability distribution. I rule out the possibility that this is driven solely by grades because both men and women receive higher grades in classes with higher ability peers. Overall, these results suggest that class composition as an important factor in determining STEM persistence for women and provide a novel explanation for part of the STEM gender gap in post-secondary education.

The second essay, co-authored with Daniel Argyle, examines the effect of the four-day school week in rural Colorado on juvenile crimes. Four-day school weeks are becoming more common nationwide especially in rural areas. Those affected by the policy spend the same number of hours in school each week as students on a typical five-day week, however the four-day week schedule essentially reallocates unstructured time into larger blocks creating a three-day weekend every weekend, since treated students for the most part have Fridays off. Our difference-in-difference estimates for rural Colorado indicate that switching all students in a county from a five-day week to a four-day week increases juvenile arrests for property crimes, in particular larceny, by about 80%.

In the third essay, co-authored with Christiana Stoddard, we study the academic achievement of American Indians. The academic achievement of American Indians has not been extensively studied. Using NAEP supplements, we find that the average achievement relative to white students resembles other disadvantaged groups. However, there are several differences. Family characteristics explain two times as much of the raw gap as for blacks. School factors also account for a larger portion of the gap than for blacks or Hispanics. The distribution is also strikingly different: low performing American Indian students have a substantially larger gap than high performing students. Finally, racial self-identification is more strongly related to achievement, especially as American Indian students age.