Firms’ Responses to a Dismissal Tax: Evidence from Brazil

Advisor: Renata Narita

Co-advisor: Gustavo Gonzaga

I study the empirical effects of a dismissal tax over firms’ behavior. By using a Brazilian law that created a 10% tax over dismissals without cause, I analyze how the tax influenced the firms’ layoff and collusion decisions. I do this in the Brazilian context, allowing for fake layoff agreements. To this extent, I develop a conceptual framework to understand the incentives created by the legislation. Guided by this framework, I estimate event-study and difference-in-differences models that exploit cross-firm variation in exposure to the reform. I find that more exposed firms reduce layoffs, particularly among higher-tenure workers, while quits increase and recall rates decline, consistent with a reduction in collusive separations. Additional evidence shows that these effects vary with informality, enforcement intensity, and spell-level exposure. Overall, the results indicate that dismissal taxes affect different separation margins, highlighting the role of institutional context and strategic behavior in shaping firms’ responses to labor market policies.