This paper addresses the methodological problems of the measurement of corruption, which is an important aspect of anti-corruption policies that produces numerical results that are increasingly used in economic, political and other types of research focused on the development of modern states. The key methodological problem here is that corruption is a latent phenomenon, and reliable information on corrupt activities is unavailable. Consequently, one needs to use indirect numerical estimates of the level of corruption that are produced using expert evaluations and polls. All existing corruption indices are based on these methods. This article analyzes the methodologies behind a wide range of corruption indices including the Corruption Perception Index by Transparency International, the Index of the Control of Corruption by the World Bank, the Corruption component of the Political Risk by the International Political Risk Guide, the Bribery Index from the World Competitiveness Yearbook by the International Institute for Management Development, and the Rule of Law Index by the World Justice Project. We propose to make a validity analysis of all the studied indices we study in order to evaluate their measurement characteristics.
measurement of corruption , latent phenomenon , methodological problems , corruption index