Marcelo Barbosa Ferreira
Marcelo Barbosa Ferreira
Ph.D. Candidate in Economics, Princeton University
Job Market Candidate 2025/26
Fields: Public Finance and Labor Economics
Embedded Files
References
References
Professor David Lee - davidlee[at]princeton.edu
Professor Owen Zidar - ozidar[at]princeton.edu
Professor Simon Jäger - simon.jager[at]princeton.edu
This paper provides estimates of income effects on labor supply by studying a large, permanent, and unconditional monthly cash flow. I leverage a unique setting in Brazil where adult daughters of military personnel receive a lifetime pension after their parent's death. This transfer is exceptionally large, the median pension payment is 282% of the national median earnings, and truly unconditional, available regardless of their employment, earnings, or marital status. Using administrative pension records matched to linked employer-employee data, I find that this transfer induces a 10% decrease on the extensive margin of labor supply after seven years, with no effect on the intensive margin, and a marginal propensity to earn (MPE) of -0.08 after seven years. Labor supply adjusts gradually, driven by a large temporary increase in the hazard of separation and a persistent moderate reduction in the job finding rate. The marginal propensity to earn is greater for larger pension payments, consistent with adjustment costs, and larger for older daughters, consistent with binding liquidity constraints and desire for early retirement. Using a sufficient statistics model, our MPE estimate of -0.08 implies that the redistributive gains from a UBI in Brazil would outweigh its efficiency costs, resulting in net welfare gains, as long as the uncompensated elasticity of labor supply is less than 0.38, consistent with microeconomic evidence.
This paper provides a theoretical framework and empirical application for the welfare analysis of tenure requirements for unemployment insurance (UI) eligibility. The model incorporates the central trade-off: a lower tenure requirement provides more consumption insurance by covering more workers, but creates moral hazard by shortening employment spells and lengthening unemployment spells. I apply this framework to a 2015 Brazilian reform that substantially increased the eligibility tenure requirement by up to 200% for some workers (from 6 to 18 months). Using administrative linked employer-employee data (RAIS) covering the universe of formal Brazilian workers, I implement a triple-difference design. This design exploits the reform's differential impact on workers based on their prior UI claim history, as workers with two or more claims served as a natural control group. Decomposing the policy's fiscal impact, I estimate that the efficiency loss was small: each dollar of mechanical transfers generated only 11 cents of moral hazard. This moral hazard is shown to be driven primarily by changes in the layoff hazard ratio rather than by changes in the job finding rate. The findings suggest that the 2015 reform that increased the tenure requirements for eligibility for unemployment insurance reduced social welfare.
Google Sites
Report abuse