Bringing the facts back in: measuring corruption for evaluation purposes, in Handbook of Research Methods for Corruption Studies, Edited by Michael Johnston

The discussion on corruption measurement has long outlived its utility. Non-specific, undertheorized aggregates of corruption perceptions are already too many, all suffering from the same problems: lack of specificity, time lag due to data collection, high confidence errors due to statistical noise created by aggregation, and so forth. While a major trade-off exists between coverage and specificity, in the end we need less to compare the reputation of countries than to improve the situation in those countries. Any simple measure that allows to trace change after an anticorruption intervention is what policymakers should endorse and fund. The new indicators, such as the T-Index and the Corruption Risk Forecast, should encourage the development of a new generation of fact-based, action-able corruption measurements.