摘要：Studies on corruption deal with the issue both from a positive point of view, especially to understand its causes, and from the normative, with emphasis on ways to reduce corruption. An important motivation for these studies is to contribute to the economic and social development of several countries. Empirical literature has contributed to the study of corruption by confirming, or not, several theories. Empirical analyzes mostly employ perception-based corruption measures, indirect measures susceptible to subjectivity bias and measurement errors, leaving reasonable doubts about their validity and reliability. This article fits into that literature by verifying whether the Corruption Perception Index shows bias associated from the level of access to information in a country, measured by the percentage of the population with access to the internet, and the volume of queries for the term corruption in a given country measured by the Google Trends tool. Using a panel of 79 countries and six years, a fixed effects model was estimated. The results indicate statistical significance for the estimated parameters. The estimated coefficients indicate that the marginal effects are quite small. However, the explanatory power of the specific component for the individual in the fixed effects model indicates that the perception of corruption would be explained by reputational aspects such as image, fame, notoriety or even importance of the country. This hypothesis, however, must be analyzed in future studies.
