Attitudes towards corruption and their consequences on political behavior

Abstract

Although vehement condemnation of corruption is widespread, the reelection of corrupt governments or politicians is all too frequent. This situation still represents a paradox in the social sciences. According to standard democratic theory, elections are expected to serve as an instrument to hold politicians to account; however, studies conducted in multiple countries indicate that voters’ punishment of malfeasant politicians is rather limited. The aim of this thesis is to assess citizens’ attitudes towards corruption and its relative importance on their voting intentions, in order to provide a better understanding of corruption accountability. A recurring explanation for why voters do not sanction corruption more severely is that they are either insufficiently informed about the wrongdoings (ignorant voters) or that they are actually not that worried about malfeasance (indifferent voters). This dissertation provides compelling evidence that voters do indeed care about corruption and that ideally, they would like to punish the corrupt politician. Nevertheless, holding politicians to account is not as simple as it may seem. Besides integrity, voters consider many other important aspects when casting a ballot. Voting is a multidimensional decision and electors may trade integrity against other characteristics that they value. In line with the indifferent voter argument, an interpretation of the overwhelming disapproval of corruption in surveys is that these answers are plagued with social desirability bias. According to this position, the rejection of corruption that citizens express in surveys is driven by their will to express socially accepted attitudes. Chapter 2 of this dissertation shows that respondents’ intentions to vote for a corrupt candidate from their preferred party does not increase when the question is formulated in an unobtrusive way. However, a respondent’s intention to vote for the corrupt politician does increase when the question is formulated as a tradeoff. Therefore, the main problem of standard survey questions that ask about attitudes towards corruption is not social desirability bias, but their inability to replicate the multidimensionality of real elections. Keeping in mind the complexity of making decision in elections, Chapter 3 uses a conjoint experiment to reflect this multidimensional scenario and to thus measure the relative importance of corruption on voting intention. This chapter provides clear-cut evidence that, voters are indeed willing to trade off corruption for other valued characteristics such as partisan identity or economic performance. The results show that co-partisanship determines voting choice to the same extent as corruption. Moreover, both co-partisanship and, to some extent, economic performance, moderate the negative effect corruption has on the vote. Besides focusing on the tradeoffs that voters face when casting a vote, this dissertation also aims to increase our understanding of the tradeoff argument. This has been carried out by (i) identifying the causal mechanisms that lead voters to (not) vote for a malfeasant politician (ii) exploring what individual characteristics of voters increase the probability of them trading integrity against representation or competence and (iii) examining what characteristics of the alternative candidates cause an increased punishment of the corrupt politician. Chapter 4 shows that a drop in the level of trust felt towards the corrupt politician in question explains why voters may decide not to vote for her. Chapter 5 identifies some modest but potentially relevant heterogeneities in citizens’ responses to corruption, while Chapter 6 shows that voters punish the corrupt politician by switching to the alternative option when this is an attractive candidate.

Date of Award	28 Jul 2020
Original language	English
Supervisor	Eva Anduiza Perea (Director) & Jordi Muñoz Mendoza (Director)