Working Paper: NBER ID: w24194
Authors: Jessica Leight; Dana Foarta; Rohini Pande; Laura Ralston
Abstract: Community targeting of vote payments — defined as the saturation of entire neighborhoods with cash prior to elections — is widespread in the developing world. In this paper, we utilize laboratory experiments conducted in the U.S. and Kenya to demonstrate that, relative to individual targeting, a vote-buying regime that distributes payments widely renders voters more tolerant of politician rent-seeking, and increases the level of politician rent-seeking observed in equilibrium. The most parsimonious model of preferences consistent with these patterns is a model in which both politicians and voters are characterized by multifaceted social preferences, encompassing reciprocity, altruism, and inequality aversion.
Keywords: vote-buying; politician accountability; community targeting; social preferences
JEL Codes: O1; P16
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| community targeting of vote payments (D72) | voters' tolerance of politician rent-seeking (D72) |
| payments distributed widely (D39) | voters become less likely to punish politicians for expropriation (P26) |
| voters who receive payments (D72) | increase their reelection thresholds (D72) |
| voters who do not receive payments (D72) | lower their reelection thresholds (K16) |
| multifaceted social preferences (D71) | greater responsiveness to community-targeted payments (H23) |
| community-targeted payments (H87) | politicians expropriate more resources without facing electoral punishment (P26) |
| payments (J33) | leniency towards politicians among recipients (D73) |
| payments (J33) | increased punishment among non-recipients (K49) |