What aspects of corruption increase the electoral punishment? Evidence from a conjoint experiment

Why do voters not hold corrupt politicians accountable? Evidence shows limited electoral punishment of corruption. However, observational and experimental data reflect citizen’s clear rejection of corruption and low levels of intention to vote for corrupt candidates. Lack of information, partisanship and performance considerations have been found significant at explaining the actual low electoral punishment of corrupt candidates. However these previous researches use a single and compact treatment for corruption and can therefore only depict the outcome of the manipulation as a whole. Hence they are unable to determine what components of corruption cause a higher or lower electoral punishment. This leads to the question whether all cases of corruption, no matter who is the beneficiary, what payoff the actor gets or the scope of the case, have limited electoral consequences. Our research design aims at providing a more detailed account of the complexity of corruption. It will contribute to the current research on corruption voting as this is the first study that attempts to determine the impact that different components of corruption have on the vote. We use a conjoint experiment embedded in an online survey representative sample of the Spanish population (N=800). Respondents are confronted with the profiles of two candidates that have a set of dimensions with randomly assigned attributes and are asked to choose the candidate they would rather vote for. Each respondent is asked to repeat the same task three times. This procedure allows us to assess the relative influences of each attribute on the electoral punishment and determine what combination of variables increases the electoral punishment of corrupt candidates. This paper focuses on five dimensions of corruption: the type of actor involved, the scope of the case of corruption, the beneficiary of the corrupt act, the payoff gained by the actor and the impact the case might have had in society. We expect that corruption cases with a negative impact in society, where a female candidate with left ideology is benefiting a private company and getting in return an immediate economic payoff will have a greater electoral punishment. We also expect that this electoral punishment will decrease when the alternative candidate is placed farther away in ideological distance. This experimental evidence on a detailed account of corruption sheds some light on why democracy has not been successful in curbing corruption.