Divining the Level of Corruption. A Bayesian State-Space Approach.

This paper outlines a new methodological framework for combining indicators of corruption. The methodology of the World Governance Indicators is extended to fully make use of the time-structure present in corruption data. The resulting state-space framework is estimated using a Bayesian Gibbs sampler algorithm. The state-space framework holds many advantages from a practical, an estimation and a theoretical point of view. Most importantly, the indicator significantly increases data availability while at the same time addressing the selection bias issues that plague the CPI and WGI indexes. It produces estimates that are more stable and reliable. Because the estimation framework is transparent and data is entered without any manipulations, the resulting indicator should also be more objective.