Campaign Spending and Rents in a Probabilistic Voting Model

How the levels of corruption and embezzlement of a political system are influenced by electoral campaigns? How rent extraction can be reduced with anti-corruption policies? We answer these questions in the context of a probabilistic voting model characterized by the absence of political pressure groups and by the presence of ideological voters whose preferences can be manipulated by political candidates through campaign spending. Our main innovation is the introduction of an analysis of candidates’ campaign choices in the literature on the agency costs of political delegation. Moreover, we contribute to the literature establishing a direct link between campaign expenditures and the utility of the voters. We find that campaigning choices are orthogonal to decisions about rent extractions and that candidates always invest a significative amount of their resources of their expected rents in advertisements. As the electoral competition itself does not suffices to reach an efficient outcome, we then study how welfare policies can reduce the inefficiencies of the electoral competition. We show that limitations of campaign expenditures are, in absence of lobbies, always welfare decreasing for voters. Indeed, our main policy suggestion is to introduce an anti-corruption, i.e. a policy that reduces the ability of candidates to extract rents by abating the incentives to rent accumulation. We show that the introduction of such tax can make the citizens better off. Surprisingly, it may also make the candidates better off if the policy is not sufficiently funded. Finally, we establish the conditions under which a policy of this kind can achieve the popular support required for an effective implementation and we show that these conditions are difficult to achieve in countries with large income inequalities.