Abstract

The Corruption Perceptions Index (CPI) by Transparency International measures perceptions of corruption in the public sector in different countries around the world. Upon invitation of the CPI team, the JRC assessed the new methodology in the CPI 2012 and analyzed the consequences that come with this change. The statistical assessment of the Index was done along three main avenues: an evaluation of conceptual/statistical coherence of the index structure, an interpretation of the rankings based on significance tests, and an evaluation of the impact of key modelling assumptions (imputation and normalisation) on countries’ scores and ranks. The CPI 2012 passes all the statistical filters of the quality control. The main recommendation for the CPI team is to adjust the formula for the standard errors for the small population size (errors that are currently overestimated) and for policy makers to consider the statistical significance (by means of effect size for example) when comparing the CPI scores. The results make clear that even when differences in the CPI country scores are statistically significant they should be carefully interpreted.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call