Peer-reviewed article · 2022
Brokers, Social Networks, Reciprocity, and Clientelism
Nico Ravanilla, Dotan Haim, and Allen Hicken · American Journal of Political Science 66(4): 795-812
In brief
How vote-buying brokers choose their targets in the Philippines: in dense village networks they pick well-connected voters they can monitor; in sparse networks they pick voters likely to reciprocate on their own.
Abstract
Although canonical models of clientelism argue that brokers use dense social networks to monitor and enforce vote buying, recent evidence suggests that brokers can instead target intrinsically reciprocal voters and reduce the need for active monitoring and enforcement. Combining a trove of survey data on brokers and voters in the Philippines with an experiment-based measure of reciprocity, and relying on local naming conventions to build social networks, we demonstrate that brokers employ both strategies conditional on the underlying social network structure. We show that brokers are chosen for their central position in networks and are knowledgeable about voters, including their reciprocity levels. We then show that, where village social networks are dense, brokers prefer to target voters that have many ties in the network because their votes are easiest to monitor. Where networks are sparse, brokers target intrinsically reciprocal voters whose behavior they need not monitor.
Research annotationQuestion, variables, design, findings, and mechanism at a glance▾
- Research question
- How do clientelist brokers decide which voters to target, and when do they need to monitor them?
- Independent variable (X)
- Village social network density, voters' centrality within those networks, and voters' intrinsic reciprocity.
- Dependent variable (Y)
- Whether a broker's campaign offered a voter money or in-kind goods, voters' perceptions that brokers can learn their vote choice and will sanction defection, and voters' self-reported vote for the campaign's mayoral candidate.
- Identification strategy
- Original survey of the full roster of 199 brokers employed by a non-incumbent mayoral campaign, paired with a survey of 701 randomly sampled voters, in an anonymized rural municipality in Southern Luzon during the 2016 Philippine local elections. Complete family networks for every barangay are built from shared surnames on the 2016 Certified Voter Lists, and intrinsic reciprocity is measured by one-shot anonymous play in a lab-in-the-field trust game. The targeting results are OLS at the broker-voter dyad level, regressing whether the broker reported offering the voter money on the voter's reciprocity, degree centrality, and betweenness centrality, each interacted with village family-network density, with broker fixed effects and standard errors clustered at the barangay level.
- Main findings
- Brokers are selected on betweenness centrality but are no more likely than ordinary voters to have high degree centrality, and they identify voters in their barangay about 20 percentage points more accurately than the average citizen. In the full sample brokers do not target intrinsically reciprocal voters; they target voters with high betweenness centrality. Conditioning on village network density reverses this by strategy: moving a voter's reciprocity from the 10th to the 90th percentile raises her probability of being targeted by 3 percentage points in a low-density village but lowers it by 6 points in a high-density village, while the same move in degree centrality lowers targeting by 10 points in a low-density village and raises it by 17 points in a high-density village. Being offered money is associated with a 28 percentage point increase in voting for that campaign's candidate, roughly a doubling, concentrated among central voters in dense networks.
- Mechanism
- Monitoring and intrinsic reciprocity are substitutes for solving the broker's enforcement problem, so village network density determines which is cheaper. What the data show is perceived rather than verified monitoring: in a high-density village, moving a voter's degree centrality from the 10th to the 90th percentile raises the probability she thinks a broker knows her vote by 18 percentage points and the probability she expects to be cut off from future handouts if she defects by as much as 35 points.
- Why it matters
- Reconciles the canonical monitoring-and-enforcement model of clientelism with newer evidence on reciprocity by showing both operate, conditional on network structure.
- Speaks to
- clientelism and vote buyingbrokers and machine politicssocial networksreciprocity and behavioral political economy
Cite
@article{ravanilla2022brokers,
title = {Brokers, Social Networks, Reciprocity, and Clientelism},
author = {Nico Ravanilla and Dotan Haim and Allen Hicken},
journal = {American Journal of Political Science},
year = {2022},
doi = {10.1111/ajps.12604},
}