Kinship, Capitalism, and Democracy: Evidence from Brazilian Firms

• In Event: Firms and their political connections

Abstract

How does kinship affect the political behavior of economic actors? The founders of social theory considered kinship to be incompatible with economic and political development. The dissolution of kinship ties was therefore crucial for Europe’s take-off (Banfield, 1958; Fukuyama, 2011; Goody, 1983; Mitterauer, 2010; Weber, 1905). We ask a related question that has received little empirical attention: Whether and how kinship ties affect a firm’s political behavior. We do so in the context of Brazil, the world's fourth largest democracy and Latin America's largest economy. A pervasive feature of Latin American capitalism is the prevalence of family-based economic groups (Schneider, 2013). This has proved an enduring characteristic, withstanding liberalization pressures in recent decades. Family ownership and control is considered to make Latin American capitalism more hierarchical, potentially hurting competitiveness and productivity.

We go above and beyond previous studies of family firms by employing fine-grained network data on kinship ties across all Brazilian public companies. We use a large-scale web-scraped dataset collected from Brazil's securities regulator containing detailed information on firm's ownership structure and all kinship ties at both the board of directors and top-management levels and link it to the universe of individual and corporate-level campaign contributions.

Our main finding is that family firms are more likely to make campaign contributions and that individuals with family ties are also more likely to contribute to politics as private citizens. We argue that kinship ties act as a coordination device, facilitating collective action and lowering transaction costs. We test for this role empirically by implementing a peer effect identification strategy to study whether individuals linked by family ties but sitting in different boards are more likely to be influenced by a shock affecting their relatives.
We also uncover novel empirical patterns. First, we find that family firms consistently contribute more to parties than to individual candidates, suggesting that they might have longer time horizons. Second, family firms are also more likely to contribute to right-wing parties and less to incumbent parties, suggesting that, at least in our context, campaign contributions are more ideological than strategic, and, contrary to conventional wisdom, that family firms are not purely motivated by rent-seeking. Consistent with this idea, we also document that family firms are not more likely to make a contribution if they obtain state-subsidized loans. Finally, we ask whether this equilibrium can be altered and find that the introduction of institutionalized forms of ownership can crowd out the presence of family ties.

Finally, we leverage a recent change in Brazilian electoral law banning corporate campaign contributions and find evidence of a substitution effect -- after the ban individuals with family ties boost their campaign contributions, suggesting that they are more capable of adapting to the new political environment. This entails a warning for policy makers: Institutional reforms that ignore key features of social structure can backfire.

Authors

• Pablo Balan, Tel Aviv University

• Ignacio Puente, NA

• Juan Dodyk, Harvard University