Work in progress

The figure above shows dropout rates from school-age children at different timing in health shocks. For example, Δ=3 indicates the treatment group in t and the control group in t+3, and so on. Baseline (t=-1) in all estimations equals 1.


Household response to health shocks and its intergenerational consequences [Draft coming soon]

whith Diogo Britto and Breno Sampaio

We investigate the effects of health shocks in families of low socioeconomic status (SES). To provide causal evidence, we combine rich administrative data and explore the timing of health shock occurrence to define treatment and control groups. There is a severe income loss from sick individuals, with an attempt to compensate for increasing criminal activities related to economic issues. Spouses have a smaller income decrease, which seems to be explained by the necessity of caring for their sick partners. Children suffer greatly from parent health conditions, increasing dropout rates, and age-grade distortion. An increase in enrollment in high school classes at night, besides the insurance mechanism due to massive income reduction, suggests the dropout rates should be even worse. In addition, we also show an increase in young fertility and crime. Finally, we provide novel evidence of the effectiveness of financial assistance policy to mitigate these effects. Using reduced form through judge randomization, we show that sick individuals increase sick leave requests and labor participation.


More public school transport for less dropout rates: evidence from Brazil

whith Diogo Britto, Bladimir Carrillo and Breno Sampaio

In 2008, the Brazilian government developed a public transportation program for municipalities, named Caminho da Escola (CE), for children of school age who live in rural areas and are enrolled in public schools. The CE aims to provide facilities to ease access to school and decrease dropout rates. As a first stage, we show that when the municipality joins CE, the report of the use of public transportation for children to go to school increases. The CE’s primary goal is achieved, as we can show a substantial decrease in dropout rates when the program began in the municipality. We use Callaway and Sant’anna (2020) (CS) estimator to validate our results, as part of the brazilian municipalities was never treated on all panel years. The baseline results and CS estimator results are very similar.


Policy work