The role of networks in political economy

Abstract

This dissertation investigates the different roles that networks play in political economy. In the first chapter, I study how a political party uses electoral data to monitor and incentivize the political brokers who control its clientelistic networks. I study networks organized around rural communal lands in Mexico, which are largely controlled by the Institutional Revolutionary Party (PRI). I use the fact that the level at which brokers operate (the communal land) does not necessarily coincide with the level at which the electoral data is disclosed (the electoral section). Guided by a simple model, I compute a measure of how informative the available electoral data is about the performance of the PRI's political brokers, as a function of the degree of overlap between communal lands and electoral sections. I compare the vote share for the PRI in communal lands where the electoral data is more or less informative, both when the PRI does and does not have access to resources to fund and incentivize brokers. The results suggest that clientelistic networks contribute significantly to the enforcement of clientelistic transactions. In the second chapter, which is co-authored with Joana Monteiro, we study the role of media in compensating political biases. In particular, we analyze how media presence, connectivity and ownership affect the distribution of federal drought relief transfers to Brazilian municipalities. We find that municipalities that are not aligned with the federal government have a lower probability of receiving funds conditional on experiencing low precipitation. However, we show that the presence of radio stations compensates for this bias. This effect is driven by municipalities that have radio stations connected to a regional network rather than by the presence of local radio stations. In addition, the effect of network-connected radio stations increases with their network coverage. These findings suggests that the connection of a radio station to a network is important because it increases the salience of disasters, making it harder for the federal government to ignore non-allies. We show that our findings are not explained by the ownership and manipulation of media by politicians. In the third chapter, which is co-authored with Arun Chandrasekhar and Emily Breza, we shed light on the relationship between network characteristics and investment decisions through a lab experiment in the field. We focus on the role for third parties to act as informal contract enforcers. Our protocol builds on a basic two-party trust game with a sender and receiver, to which we introduce a third-party to serve as either a monitor or punisher. The ex-ante benefits of a third party judge are ambiguous. On one hand, a third party may result in larger sender transfers due to her ability to punish. On the other hand, the punisher might act in a way to build reputation or may crowd-out intrinsic motivation. Importantly, these costs and benefits of a punisher might vary with her centrality in the network. Our findings are consistent with both the role for the punisher to induce efficiency and to crowd out intrinsic motivation. They are also consistent with the effects of reputation-building by the punisher. Importantly, we find that very network-peripheral punishers are detrimental to efficiency, while network-central individuals may improve outcomes when given the technology to punish. We also show that these results cannot be explained by either the fact that the punisher also acts as a monitor, or by the punisher's characteristics such as elite status, educational attainment, caste, or proxies for wealth.