Research

Invited to Revise and Resubmit

Abstract: How do African governments distribute state benefits to farming communities? Many accounts emphasize the role of ethnicity and traditional elites in shaping a politician’s distribution of state benefits. This paper instead argues that politicians supply rural producer organizations (RPOs), formal organizations uniting farmers, with benefits. These organizations simplify targeting decisions by complementing a politician’s incentives to supply benefits to supporters while also enabling vulnerable rural households to press governments for support. I consider this argument using nearly 100,000 survey responses from 22 nationally representative surveys in 14 countries across the continent. RPO members are consistently more likely to receive input subsidies and agricultural extension services. Additional models probe why this relationship exists, showing that RPO members are particularly likely to receive benefits when they grow cash or tree crops, live in incumbent strongholds, experience adverse production shocks, and during election years. Finally, when RPO members receive benefits, incumbents do better at the polls, substantiating the electoral mechanism motivating this hypothesized exchange. This article illustrates the underexplored political role of RPOs in African politics and shows how both supply and demand-side mechanisms underpin the role of organizational brokers.

Currently Under Review

Abstract: When do political intermediaries succeed in securing benefits for their subordinates? In this article, I argue that organizations with stronger downward ties to members and upward ties to politicians capture more nonprogrammatic state benefits. Specifically, an organization’s ability to overcome collective problems creates more credible promises of future political support, while ties between organizations and politicians ease transaction costs. I consider this argument by analyzing rural producer organizations, formal organizations bringing together agricultural producers. I begin by analyzing an in-depth panel survey from Ethiopia’s Amhara region using regression analysis and a matching approach relying on official, programmatic targeting criteria. Members of organizations demonstrating higher levels of collective action, organizations providing a higher quantity and quality of member services, and organizations with explicit ties to local politicians are more likely to receive state benefits. Furthermore, downward and upward ties potentially interact, with members of organizations both offering a higher quantity of services and featuring a local politician benefiting the most. Given the unique nature of the Ethiopian case, I replicate my analysis using survey data from eight West African cases. Members of organizations providing a larger quantity of services are more likely to receive agricultural input subsidies and government extension services.

Last Presented at the 2024 Annual Meeting of the Midwestern Political Science Association

Abstract: Ethnic models of voting behavior struggle to explain why individuals sometimes vote for non-co-ethnic candidates. Given the potential economic and psychological advantages of having a co-ethnic in power, and the possible costs of an ethnic rival’s victory, it is puzzling that voters in highly ethnicized polities often fail to show near-unanimous support for their “own” candidates. In many such contexts, cross-ethnic voting occurs frequently enough to alter national election outcomes. Rather than offering post-hoc rationalizations for why candidates fail to mobilize their co-ethnic base, we argue that cross-ethnic voting results from the interaction of two key variables: (1) a voter’s evaluation of candidate performance and (2) the social pressure individuals feel from their own ethnic group to support an in-group candidate. The first variable reflects the logic of retrospective voting, voters reward candidates they believe have performed well in office. The second captures the social pressure often assumed but rarely measured in studies of ethnic voting: the extent to which interaction with co-ethnics constrains individual choice. We hypothesize that citizens most likely to “cross the ethnic line” both evaluate their co-ethnic candidate’s performance poorly and interact less frequently with fellow group members. Using original exit poll data from Kenya, we find strong support for this claim. We then extend the argument to a non-ethnic context, where group-based pressure stems not from ethnicity but from partisanship. Using exit poll data from Uganda, where a dominant party rather than ethnic groups exerts pressure on voters, we observe the same dynamic: dissatisfied party members less embedded in their party networks are the most likely “crossers” at the ballot box. Our explanation integrates retrospective evaluations of candidate performance with the social pressures generated by group membership.