Research

Working Papers

(JMP) Power to the teens? A model of parents' and teens' collective labor supply

[draft] 

Teens make life-changing decisions while constrained by the needs and resources of the households they grow up in. Household behavior models frequently delegate decision-making to the teen or their parents, ignoring joint decision-making in the household. I show that teens and parents allocate time and income jointly by using data from the Costa Rican Encuesta Nacional de Hogares from 2011 to 2019 and a conditional cash transfer program. First, I present gender differences in household responses to the transfer using a marginal treatment effect framework. Second, I explain how the gender gap from the results is due to the bargaining process between parents and teens. I propose a collective household model and show that sons bargain cooperatively with their parents while daughters do not. This result implies that sons have a higher opportunity cost of attending school than daughters. Public policy targeting teens must account for this gender disparity to be effective.

You are the father! Effects of Costa Rica's Responsible Paternity Law on families 

[slides]  [draft] 

Costa Rica's Responsible Paternity Law made it easier for unmarried women to declare the father of their newborn child and thus obtain child monetary support. This paper assesses the impact of the law on household decisions. I estimate the law's effects using the law as a natural experiment and a fuzzy differences-in-differences setting. I find that the law had a negative impact on male labor participation as well as female and male weekly labor supply. Using a collective household model with matching, I argue that the law strengthens women's bargaining power in household decision-making. This has two consequences: a couple selection effect and an intra-household allocation effect. Structural estimates show that both effects exist in households. These findings demonstrate how child-related laws help us better understand household formation and decision-making.

Work in Progress

Attrition and time-variant unobserved heterogeneity in an education-labor dynamic discrete choice model (with François Poinas)

Discrete choice dynamic programming (DCDP) models allow to estimate education-labor decisions with individual heterogeneity that cannot be seen on the data. The most common way is to assume that this unobserved heterogeneity is time-invariant. However, some results point out that, in ages from 16 to 30 years old, life experiences vary this term. We present a DCDP model for educational and labor choices allowing for time-variant unobserved heterogeneity to include the effect of life decisions and life events that changes individuals' unobserved characteristics. With the model it is then possible to estimate counterfactuals for public policies like a raise in tuition fees and see the potential differences compared to a time-invariant setting. The data we use is Génération 98 which has a severe problem of attrition. We also present a selection mechanism on observables and unobservables to incorporate the decisions and wages of the individuals made before the decision to leave the sample. Correcting for attrition bias has not been done before in the DCDP literature, even when most of the data sets used have this problem.