As it does every year, the release of the Corruption Perception Index (CPI) from Transparency International is making headlines. And rightly so. There is no better occasion for the political opposition and the critical media to rally against the government—that is, in countries blessed to enjoy a free press and a political opposition, which are fewer every year.
This shows us why the average of subjective rankings (gathered from various banks and agencies) known as CPI has been a media success and has endured so well over time since it was unexpectedly raised to fame by the friendly encounter between an intern in the early days of TI (Johann Graf Lambsdorff) and an Economist journalist. Of course, an average of subjective ratings is not more objective, but it does clearly indicate the reputation of a country. And reputation matters. CPI has managed so far great civic work in naming and shaming governments. And as anyone in the anti-corruption business knows, there is hardly a more effective instrument than naming and shaming.
This is not to say that CPI has not been misunderstood and misused daily. Most laypeople do not grasp the concept of a statistical aggregate and do not click on methodology entries. They compare the indicator across time, although the change in the number of countries and the statistical noise created by aggregation make this an imprecise if not altogether misleading affair. When we read “South Africa’s ranking on the 2023 Transparency International Corruption Perceptions Index has fallen to 83 out of 180 countries, marking a decline of 11 spots compared with the previous year’s rankings” we do not know actually where precisely South Africa is compared to where it had been a year ago (since other countries may have changed, dropped out, etc.). Nor, certainly, do we know why. If anything, disaggregated indicators of causes associated with good control of corruption in the research profession show that South Africa has rather progressed on the whole, although its judicial independence may be under strain. But those who perceive it as backsliding also perceive something. And what are these perceptions based on? The CPI does not tell us.
These are the kinds of issues that the next generation of research – such as the Horizon Project BridgeGap, in which Transparency International is an essential member and of which I am the principal investigator – will hopefully clarify. Beyond naming and shaming, which have been and remain laudable actions, we need more actionable and specific indicators that allow us to measure results before and after anti-corruption policies are implemented. Indeed, we need tools that help us evaluate if and how regulation is leading to progress – or not.