This dissertation introduces three essays on the effects of different public policies on crime, education, and labor outcomes using quasi-experimental research designs. These policies include targeting high-ranked members of criminal organizations to fighting organized crime and extending the school day. In the first essay "Kingpin Approaches to Fighting Crime and Community Violence: Evidence from Mexico's Drug War," a joint work with Jason Lindo, we consider the effects of the kingpin strategy, an approach to fighting organized crime in which law-enforcement efforts focus on capturing the leaders of criminal organizations, on community violence in the context of Mexico's drug war. Newly constructed historical data on drug-trafficking organizations' areas of operation at the municipality level and monthly homicide data allow us to control for a rich set of fixed effects and to leverage variation in the timing of kingpin captures to estimate their effects. This analysis indicates that kingpin captures cause large and sustained increases to the homicide rate in the municipality of capture and smaller but significant effects on other municipalities where the kingpin's organization has a presence, supporting the notion that removing kingpins can have destabilizing effects throughout an organization that are accompanied by escalations in violence. In the second essay "The Short and Long Run Effects of Full-Time Schools on Academic Performance," I study the effect of extending the school day on student achievement in Mexico, where more than 23,000 schools have extended their school day from 4.5 to 8 hours since 2007. I use the variation in the timing with which schools extended their school day to estimate the impact of this intervention on students' math and reading test scores. I find evidence that extending the school day does not affect student achievement the year of adoption; however it improves math and reading test scores by 5 percent of a standard deviation one year after adoption and the effect grows over time to 15 percent of a standard deviation four years after adoption. I also find that the effects are more pronounced in schools located in high-poverty communities and for students in lower grade levels. In the third essay "The Effect of Children's Time in School on Mothers' Labor Supply: Evidence from Mexico's Full-Time Schools Program," a joint work with Francisco Cabrera-Hernandez, we examine the effect of the time children spend in school on female labor supply. In particular, we investigate the degree to which extending the school day by three and a half hours, in elementary schools, affects labor force participation, the number of weekly hours worked, and the monthly earnings of females with elementary-school-age children. To do so, we exploit within-individual variation in access to full-time schools and a rotating panel of households that contains individual-level data on labor outcomes and sociodemographic characteristics. Results from long-difference models show that extending the school day increases mothers' labor supply at the extensive and intensive margins, increasing mothers' labor force participation by 7 percentage points and the number of weekly hours worked by 2.4. Moreover, these increases are accompanied by a 47 percent increase in monthly earnings.
