I study an affirmative action policy aimed at improving gender balance within college fields. The policy boosts the application scores of individuals whose gender is underrepresented in the program they apply to. Since the admission design is strategy-proof, these extra points should only affect the probability of admission conditional on applying, not college applications themselves. Using a novel dataset covering all college applications in Norway, however, I find that candidates of the underrepresented gender are much more likely to apply when they receive the boost. When programs introduce (end) the policy, they receive about 15 percent more (fewer) applicants of the underrepresented gender. This response exists along the entire distribution of application scores, including among individuals who would have been admitted absent affirmative action had they just applied. I subsequently estimate how the policy affects academic and labor market outcomes of those shifted in versus those shifted out of their preferred program by the policy under different assumptions about application behavior.
Prenatal health insults have been shown to have negative consequences on adult outcomes, while postnatal policies have been shown to benefit treated children. In this project, combining two orthogonal difference-in-differences designs, we document that access to health care center at ages 0-1 reduces the negative effects of prenatal influenza exposure on completed years of education, lifetime earnings, and adult health. This suggests that health interventions early in childhood can compensate for the adverse health shocks experienced in-utero.
Please get in touch for the draft from NBER SI