• Trade-off Between Public Goods and Corruption: What Do We Learn When Comparing Survey Experiment with Traditional Survey?, with Gabriel Cepaluni and Gustavo Araújo

  • Is the use of survey experiment more efficient than traditional surveys in responding to social sciences' research demands? Or are we learning different things when using these methodologies? Few authors have compared results from either natural experiments with survey experiments or from list and endorsement survey experiments. Additionally, there is no empirical work comparing results from a survey experiment with a non-experimental survey. For theoretical reasons, the literature just assumes that the former is more efficient. And for that reason, we propose to empirically evaluate the topic embedding experimental and non-experimental survey questions on the provision of public goods by a hypothetical corrupt politician in Brazil. Using data from a national survey experiment with 1,500 respondents, we evaluate how Brazilian citizens behave when facing a trade-off between the provision of a specific public good – public health – and the possibility of  corruption involved in the case with both a survey experiment and a traditional survey. Our research presents a simple design, consisting of one baseline and two treatment questions and we found large differences between the direct traditional survey question and the indirect survey experimental question. We believe that the results are not only due to social desirability bias. Voters are also reacting to two distinct scenarios: for the survey experiment, they know all the strategies put forward by the corrupt politician, while they are uncertain about these strategies when answering survey experimental questions. Substantively, our study contributes to the large literature on the trade-off between the provision of public goods and corruption in developing countries (Winters, Testa, and Fredrickson, 2012). More specifically, we contribute to the growing experimental literature on the trade-off between the provision of public goods and corruption in Brazil (Winters, Testa, and Fredrickson, 2012; Winters, and Weitz-Shapiro, 2013, Ferraz, and Finan, 2008; Figueiredo, Hidalgo, and Kasahara, 2011).