UC Berkeley

In many countries, clientelist parties (or political machines) distribute selective benefits, especially to the poor, in direct exchange for electoral support. Many scholars view clientelism as a political strategy, but fail to distinguish between substantively different patterns of machine politics. Conflating distinct strategies of clientelism poses a serious threat to descriptive and causal inference. This study seeks to increase analytical differentiation of clientelism, building on fieldwork in Brazil, formal modeling, and econometric analyses of survey data.

A fundamental, yet frequently overlooked, distinction lies between strategies of electoral and relational clientelism. Whereas electoral clientelism involves elite payoffs to citizens during campaigns, relational clientelism involves ongoing relationships beyond campaigns. Electoral clientelism — the primary focus of this study — delivers all benefits to citizens before voting, and involves the threat of opportunistic defection by citizens. By contrast, relational clientelism delivers at least some benefits to citizens after voting, and involves the threat of opportunistic defection by both citizens and elites.

Scholars often conflate vote buying with other strategies of electoral clientelism. Much of what scholars interpret as vote buying (exchanging rewards for vote choices) may actually be turnout buying (exchanging rewards for turnout). This study advances research on clientelism by specifying and testing a mechanism by which parties can distribute benefits to mobilize supporters. Formal modeling suggests that turnout buying is incentive-compatible, and also provides several testable predictions: (1) machines focus rewards on strong supporters, (2) they target the poor, and (3) they offer rewards where they can most effectively monitor turnout. Although both strategies coexist, empirical tests suggest that Argentine survey data are more consistent with turnout buying than vote buying. Two other strategies of electoral clientelism are also frequently conflated with vote buying: double persuasion (exchanging rewards for vote choices and turnout) and negative turnout buying (exchanging rewards for abstention).

Formal analysis in Chapter 4 — coauthored with Jordan Gans-Morse and Sebastian Mazzuca — suggests that political machines are most effective when combining multiple strategies of electoral clientelism. Machines adapt the size of clientelist benefits to citizens' political preferences and inclination to vote, and are willing to pay relatively more for vote buying because unlike other strategies it both adds votes for the machine and subtracts votes from the opposition. The model also suggests that machines tailor their mix of electoral clientelism to five characteristics of political environments: (1) compulsory voting, (2) machine support, (3) political polarization, (4) salience of political preferences, and (5) strength of ballot secrecy. The model's predictions are consistent with qualitative evidence from Argentina, Brazil and Russia.