Can exposing corruption have unintended negative consequences? I tackle this question in the context of labor unions in the U.S., where the U.S. Senate McClellan Committee (1957-1960) publicly exposed corruption and organized crime infiltrations in their ranks. Using a difference-in-differences identification strategy and novel data, I examine the consequences of the Committee’s investigation on unionization, the capacity of unions to mobilize voters during elections, and their ability to influence public policy. I study both the direct effects of the investigation in areas where investigated union locals were present and the indirect effects (or spillovers) in areas where no investigated union locals were present. First, I find that the negative spillover effects on unionization were stronger than the direct effects. Second, I show that the investigation caused a persistent decrease in the capacity of unions to foster voters' political participation in presidential elections. Finally, I provide evidence suggesting that the spillovers are at least partially explained by a large-scale news and reputation shock that had negative consequences on the entire American labor movement.

Governments often implement large-scale redistribution policies to gain enduring political support. However, little is known on whether such policies generate sizable gains, whether these gains are persistent, and why. We study the political consequences of a major land reform in Italy. A panel spatial regression discontinuity design shows that the reform generated large electoral gains for the incumbent Christian Democratic party. The electoral effects persist over four decades. We explore several channels and find that clientelist brokering and patronage are plausible mechanisms for this persistence.

We study whether work practices adopted in recent history in agriculture influence the political empowerment of women in a European country. Focusing on Italy, we study the case of female rice weeders and their successful history of unionization and mobilization for better working conditions. Relying on an instrumental variable strategy to predict quasi-exogenous variation in rice production, we test whether the historical presence of female rice weeders predicts differences in measures of political empowerment during the second half of the XX century. We find that towns where rice production was historically relevant had higher support for divorce in the 1974 referendum, have more women in politics, and are more likely to have a nursery school. Our results suggest the importance of collective action of working women to achieve persistent female political empowerment. (Draft available upon request)