Research

This paper develops and estimates a dynamic model of employment, schooling, and crime to study how the adult outcomes of adolescents from disadvantaged backgrounds are impacted by accumulating skills in the illegal sector, i.e., criminal capital. The estimation is based on 7 years of observations from a sample of juvenile offenders from Phoenix and Philadelphia, and shows that criminal capital accumulated during adolescence increases the likelihood of adult reoffending and permanently decreases educational attainment. Moreover, individual heterogeneity significantly explains enrollment, labor market attachment, and crime desistance in the long run. Using the estimates, I conduct counterfactual policy experiments to discuss the optimal allocation for long-run crime reduction of two policy instruments, school and wage subsidies. Simulations show that strategies to minimize crime and public expenditure should combine school subsidies before the age of majority and later wage subsidies—or, alternatively, school subsidies just before individuals turn 18.

We evaluate the returns to signaling occupation-specific skills using unique administrative data from a nationwide certification program in Colombia. The program certifies skills and issues three certificates: basic, intermediate, and advanced. We use regression discontinuity methods to compare workers' earnings around certificate-assignment thresholds. Signaling advanced occupation-specific skills yields significant returns: 9.7%, on average, within two years of certification. Instead, we find no effects from signaling basic or intermediate occupation-specific skills. Our analysis reveals that the primary mechanism behind the observed income effects associated with the advanced certificate is the ability to signal occupation-specific skills to potential employers.

This paper examines whether a policy promoting job stability among disadvantaged sales agents (lower-educated, young females) affects the firm's and workers’ performance. We define job stability as signing a permanent contract protected by the law. Unlike previous studies, we use a precise measure of output (daily sales) among comparable sales agents from a large retail firm in Southeast Asia. Our findings suggest that sales agents increase their daily sales by 7% in the next 180 days after getting a permanent contract. The findings are robust to store characteristics, seasonal trends, and peers’ quality. We analyze the distributional effects of stability by gender. Our estimates show no difference between the behavioral responses of men and women.

We investigate the relative effectiveness of two distinct advertising campaigns, self-contribution focus (SC) and peer-enforcement focus (PE), in promoting pro-social behavior. A large-scale field experiment in Singapore focused on reducing littering is the basis for evaluating the campaigns. The results show that both SC and PE campaigns effectively reduce littering, with PE being particularly successful among those who are less inclined to adopt pro-social. Two waves of door-to-door surveys were conducted to collect microdata on residents in areas where campaigns were launched. Results suggest that both campaigns were successful in raising awareness about anti-littering norms. Both campaigns had an impact on people's perception of being publicly scrutinized if caught littering. However, our estimates indicate that the effect was bigger in areas where the PE campaign was implemented.

This paper develops and estimates a dynamic model in which adolescents jointly decide to engage in crime, school, work, and consumption of illegal drugs. The model accounts for different dimensions of individual heterogeneity that affect the evolution of criminal engagement, human capital formation, and addiction. It also features multiple channels for studying the reciprocal effects of drug consumption on crime and human capital formation. Estimation is based on seven years of observations from a sample of juvenile offenders from Maricopa County (Phoenix) and Philadelphia County (Philadelphia). Estimates show that drug consumption increases illegal earnings, creating a feedback effect that reinforces criminal participation over time: crime rewards people with high addiction capital, facilitating future drug use, increasing illegal earnings, and making crime more appealing over time. We also find evidence of negative returns to repeated drug consumption in the legal sector. According to our estimates, the accumulation of addiction capital decreases the discount factor, making it less likely that individuals respond to incentives placed in the future.

Using the model parameters, we perform different simulations. Results indicate that limiting drug consumption at early ages reduces crime by 6.4% and the joint decision of schooling and crime by 8.4%. In addition, reducing criminal engagement before age 13 decreases addiction capital by 20% at age 25. We use the model estimates to explore the consequences of introducing policies implementing dynamic incentives to encourage ongoing education and disrupt the dependence cycle caused by prior engagement in criminal activities. Our findings indicate that utilizing dynamic incentives is significantly more effective in influencing long-term behavioral changes in adolescents.

When sellers can recommend services that are more expensive than necessary, how do trust signals from consumers affect their behavior? One might argue that trust signals could imply a lower level of ex post monitoring by consumers, lowering any potential reputational losses when sellers overtreat. However, trust signals may also induce sellers to be guilt averse towards violating the consumer’s expectations of honesty, limiting overtreatment. We shed light on the theoretical debate on the effects of consumer trust signals using two field experiments. We also investigate how trust signals operate when consumers might also be misinformed about the cor- rect service required, specifically when they believe that a more expensive service might be needed. The first experiment, set in the context of cellphone repairs, finds that when consumers signal trust only, they receive lower price quotes. However, when consumers also provide signals that they are misinformed, trust signals back- fire significantly, exacerbating overtreatment. A second field experiment confirms the latter — we find that taxi drivers are more likely to choose longer routes when passengers are misinformed and send trust signals.

Using data from a survey of juvenile offenders, I estimate the effects of incarceration, by facility type, over mental health and non-cognitive skills. I find evidence that effects vary by age. Younger offenders placed in adult facilities experienced the most significant decrease in mental health and non-cognitive skills. Also, my results suggest that such effects are long-lasting. I later discuss the implications of poor mental health over reoffending and high school graduation.

This paper describes and analyzes different forms of equilibrium within the market of standard form contracts where information is asymmetric and the agreement drafters decide which type of contract to offer considering that information is expensive for those who adhere. Under such conditions, equilibriums are met in which contracts exist with the minimum possible quality associated with a price. The results show that the market is unable to correct inefficient agreements and this produces a limitation to freedom of contract. This situation requires the implementation of controls over the contents of the standard form contracts in order to ameliorate such outcome. Finally, there is an experimental analysis of the fundamental theoretic results.