Are Perceptions of Corruption Matching Experience? Evidence from Microdata

The efficacy of corruption perception indices to truly capture and accurately measure corruption behaviours has been often criticised. In fact, perceptions about corruption may not match actual experience and could represent distorted beliefs. Motivated by this criticism, we investigate the difference between perceived and experienced corruption (i.e., bribery) in public services in Europe by means of a theoretical model and an empirical analysis. Firstly, we model perceived corruption as a function of experienced corruption and a perception bias. Then, we employ a generalised setting of structural equation models to derive two distinct measures of perceived and experienced corruption from microdata on the public administration sector in Europe. The indices we obtain allow us to compare countries according to both measures of public corruption. Finally, our results suggest that perceptions of corruption may be affected by sources of media bias.