Does respondent reticence affect the results of corruption surveys ? evidence from the world bank enterprise survey for Nigeria

A potential concern with survey-based data on corruption is that respondents may not be fully candid in their responses to sensitive questions. If reticent respondents are less likely to admit to involvement in corrupt acts, and if the proportion of reticent respondents varies across groups of interest, comparisons of reported corruption across those groups can be misleading. This paper implements a variant on random response techniques that allows for identification of reticent respondents in the World Bank’s Enterprise Survey for Nigeria fielded in 2008 and 2009. The authors find that 13.1 percent of respondents are highly likely to be reticent, and that these reticent respondents admit to sensitive acts at a significantly lower rate than possibly candid respondents when survey questions are worded in a way that implies personal wrongdoing on the part of the respondent.