In a recent FCPA Blog post, William Weaver argued that “actual corruption is essentially unmeasurable,” and that “our imperfect measurements of corruption, even more accurate random forced response surveys, have found only a noisy relationship between perceptions of corruption and its reality.”
Because these bold statements implicitly reject substantial masses of empirical data from a variety of sources, anti-corruption experts need to analyze them closely.
The fundamental problem with Weaver’s thesis is his failure to define what he means by “actual corruption” and “unmeasurable.” Instead, he sidesteps the definitional problem by asserting that “[d]ifferent observers view corruption in different ways”: a “boon” or ‘self-described necessary evil” for corrupt individuals, “a premodern form of patronage,” or “various operations of a premodern system of communication now described as criminality.”
The broader and more diverse the list of attributes a concept has, of course, the more difficult it becomes to define or measure that concept. But even if there is no authoritative definition that covers every category of corruption, there are various methods by which countries, businesses, and non-governmental organizations can and do measure important dimensions of corruption.
At the grass-roots level, for example, there is now a broad array of experiential surveys that gather voluntary responses from representative samples of people in everyday life about their actual encounters with corruption. While barometer surveys can differ in their scopes and purposes, it is fair to say that the Global Corruption Barometer, Afrobarometer, and the World Justice Project Rule of Law Index (to name a few) provide important current and longitudinal data about actual and perceived corruption in jurisdictions around the world.
Weaver briefly disparages data relating to perceptions of corruption. Yet resources such as Transparency International’s Corruption Perceptions Index should not be summarily dismissed. Even taking into account the possibility that respondent and country characteristics may bias corruption perceptions, there is still value in systematically sampling the views of informed businesspeople and country experts about the levels of public-sector corruption.
Certainly at the grand-corruption level, there is no reliable contemporaneous measurement of the total number and amounts of bribes that corrupt officials around the world are taking. On the other hand, as numerous corruption investigations have demonstrated, in particular cases law enforcement agencies can document, with a fair degree of precision, how much in bribes was paid to officials and how much business the companies paying the bribes received in return.
Establishing the orders of magnitude and patterns of bribery by companies yields data relevant for corporate-compliance teams as well as anti-corruption advocacy groups.
In addition, Weaver’s commentary takes no note of the increasing use of sophisticated data analytics and artificial intelligence by companies and government agencies. Those technologies make it much more possible to conduct ongoing monitoring of transactions and accounts at all levels of a company, and identify anomalous activity that may be indicative of corruption.
In short, if the standard to be adopted for measuring corruption is perfection, then corruption is no more measurable than the length of the coast of Britain or the volume of water in the Mississippi River. But perfection can never be the standard for anything we seek to measure, regarding corruption or any other phenomenon. Reasonable people can and should work with reasonable estimates of corruption, so long as they recognize the need to keep improving the quantity and quality of those data.
Jonathan J. Rusch, pictured above, is Principal of DTG Risk & Compliance, and former Head of Anti-Bribery & Corruption Governance at Wells Fargo. The views expressed are his own. He can be contacted here.