Essays on the Political Economy of Local Corruption

[eng] This dissertation analyses the effects of information on corruption cases on citizens’ electoral behaviour and the media coverage of those scandals. Corruption, defined as the abuse of public office for private gain, has lately become a very prolific research field in both academic and policy areas. Considering the main factors driving corruption, some studies have identified democratic systems as a hurdle to political scandals. Advanced democratic institutions tend to be associated with higher transparency and better political accountability mechanisms, which are the channels through which they accomplish lower levels of corruption. Factors such as an independent judiciary, press freedom, and free elections are key elements that define an advanced democracy. This thesis is composed by three empirical studies. The study presented in Chapter 2 analyses how information on local corruption affected local electoral outcomes in Spanish municipalities between 1999 and 2007, a period characterised by the surge in local scandals. We use a novel database on those corruption cases to estimate an incumbent's vote share equation, accounting for the omission of popularity shocks, something that is lacking in prior studies. As an additional enrichment to the literature we have into consideration the degree of attention that the media devoted to each case and when the judiciary was involved in the scandal, analysing whether voters react to the amount of information and to information regarding the seriousness of the case. Thus, we account for the complementarity of these institutions in the fight against corruption. Chapter 3 studies how corruption affects voter turnout using information on local scandals occurring in Spain between 1999 and 2007 and survey data. This analysis has the advantage over the previous literature as it relies on a research strategy for differentiating between the ‘mobilisation’ and ‘disaffection’ effects of corruption on voter turnout. To the best of our knowledge this is the first study that is able to analyse empirically how these two effects are influenced by partisan leanings or corruption at different times, untangling the conclusions of earlier studies. Chapter 4 studies the media coverage of 165 Spanish local scandals spanning between 2004 and 2007 by national and regional newspapers. It analyses the incentives that media outlets may have to bias the information they report on those scandals. The literature has identified ideological slant and capture on the part of the government as two political elements that may bias media coverage of scandals. The study presented in chapter 4 is an improvement respect previous papers since it analyses both ideological slant and media capture as complementary factors rather than independent drivers of media bias. As an additional contribution we also consider the role of government’s popularity on the coverage of scandals. The three empirical studies that compose this thesis provide strong evidence that, even under a biased provision of news, Spanish voters are willing to electorally punish corrupt practices. Together with the significant number of cases recently unveiled by media and investigations undertaken by individuals or citizens’ organizations through different digital platforms, we can be optimistic about the evaluation of practices to control corruption. The promotion of policies that endorse media freedom and independence would also reduce the influence of political powers on Spanish media. Taken together, these factors would have a clear positive effect on electoral accountability, allowing citizens to obtain the impartial information they need to use elections as a way to constrain corrupt practices