I study the effects of improvements in transportation infrastructure on informality. To deal with endogeneity issues, I implement two complementary identification strategies. First, taking advantage of the staggered rollout of highways, I apply a generalized difference-in-difference regression model. Second, I conduct an instrumental variable strategy by exploiting the fact that municipalities along the route of important cities in Brazil were more likely to be connected to the federal highway system. I find that, once a municipality is connected to the federal highway network, there is a reduction in informality, measured by the self-employment rate among non-agricultural, low-skilled workers. Moreover, I show that connected municipalities have higher GDP per capita, larger firms in the formal sector, and higher demand for formal employment. These results suggest that better transportation infrastructure induces the growth and development of the formal sector.

In this paper, I empirically analyze the effect of the disclosure of corruption cases on citizens’ compliance with the law. To do so, I use data on corruption cases generated by the Brazilian anti-corruption plan, “Programa de Fiscalização em Entes Federativos,” which audits municipalities for their use of federal funds. The random selection of municipalities to be audited provides a straightforward empirical strategy. I measure non-compliance with the law by citizens using data on traffic offenses at the municipal level. My main results indicate that the disclosure of corruption cases at the municipal level increases per capita traffic offenses by 1.2% and an additional case of corruption disclosed increases traffic offenses per capita by 0.4%. These estimates are small and not statistically different from zero. Therefore, I cannot conclude that the disclosure of corruption impacts citizens' compliance with the law.

We examine the contribution of automation to occupational gender segregation and to the gender gap in college education. First, we document that women were more likely to be displaced by automation. Then, exploiting cross-commuting zone variation in the risk of automation, we show that women were much more likely than men to transition out of routine task intensive occupations to occupations requiring higher levels of skill, for a given shock in the risk of automation. Local labor markets that were more affected by automation experienced greater occupational integration by gender. Potential mechanisms are the growing demand for social skills that favor women and their greater ability to upskill. Consistent with these mechanisms, we find that local labor markets more susceptible to automation saw larger increases in the share of young women completing college relative to men and a greater movement of women into occupations with high math and high social skill requirements.

Draft coming soon!