Buying an Influencer with Binding Promises

Abstract In addition to campaign promises and clientelistic practices via particularized (non-programmatic) benefits to boost their chances of winning an election, politicians may also seek the endorsement of influential members of the society in exchange for benefits provided before the election or conditional on electoral success. As a starting point to analyze such exchanges, we construct a simple full-information contractual bilateral exchange game where particularized benefits are exchanged for an influencer’s costly action and derive conditions under which different forms of payment would be used. Using a framed laboratory experiment that exogenously varies the politician’s gain from winning and the influencer’s impact on the chance of success, we find that the number of exchanges is significantly lower than predicted by a model of purely self-interested behavior. When exchanges do occur, they overwhelmingly take place with winning-conditional payments. Our results are consistent with a model of inequity aversion, whereby influencers’ fairness concerns constrain influence buying, particularly when candidates’ gains from winning are high. We argue that knowledge of the politician’s potential rents, combined with other-regarding preferences among influencers, may limit influence buying.