Re-evaluating the behavioral change from conditional cash transfers: Evidence from the Avancemos National Program in Costa Rica

Conditional cash transfers (CCT) have become a common solution aiming to reduce poverty by increasing the human capital of children. In 2006, the Government of Costa Rica introduced Avancemos, a nationwide CCT program, with the premise that a monetary subsidy, conditional on school attendance, would offset adolescents' opportunity cost from working and increase secondary school completion. In this paper, we assess whether the program achieved its main objective of increasing secondary schooling and reducing adolescent labor, and we disentangle the mechanisms generating these effects. We used data from the Costa Rica National Household Survey (2006-2014) collected by the National Institute of Statistics and Censuses. The program effects are estimated through a propensity score matching technique. We find that for program beneficiaries, Avancemos led to increases of about 27% in attendance and 0.72 year of schooling, and a decrease of 6% in the likelihood of labor participation. We also present evidence that economic vulnerability is not the only or main cause for school dropout and argue that efforts complementary to CCT, such as improving instruction quality and offering more information on the returns to education, are needed to increase attendance and completion and to reduce child labor outcomes.