We still don’t have a good way to measure global corruption

Corruption is an urgent global problem, one that costs the developing world dearly and badly slows its development down. But until recently, figuring out just how big the problem is, and what to do about it, has been fiendishly difficult.

Now, the ONE Campaign has launched an important report calling on the G20 nations to address the “trillion dollar scandal” of global corruption. In it, the campaign estimates that a web of corrupt activity robs developing nations of at least US $1 trillion every year.

Similar figures have appeared in other estimates: the World Economic Forum, for instance, has estimated the cost of corruption to be more than 5% of global GDP (US $2.6 trillion), while the World Bank believes more than $1 trillion globally is paid in bribes each year.

Of course, while there can be little doubt that corruption costs the global economy dearly, the secretive nature of corrupt exchanges makes uncovering their true extent a tricky task indeed.

Given the staggering corruption figures that are being floated, it’s hardly surprising there is a serious impetus to develop accurate ways to measure corruption – to show not only how much corruption exists in the world, but also where it occurs and with what frequency, and ultimately to provide guidance on how to stop it.

Unfortunately, because currently available corruption measures are so beset by conceptual, methodological and political problems (sometimes all at once), they are just not very good tools for effective anti-corruption policy.

Same old story

This is especially true of the most widely used measure, Transparency International’s Corruption Perceptions Index (CPI). Published yearly, usually to extensive media coverage, the CPI tries to quantify national corruption by measuring how experts, business leaders, and the public perceive it in their countries.

The CPI has been politically extremely influential; it has done a lot to raise awareness about the extent of corruption in different countries, and on its face, it provides a reasonably useful broad-brush picture of the overall extent of national corruption. But despite the way CPI is almost cherished as a data source, it’s still seriously flawed.

There are fundamental problems with the core assumption on which the CPI rests. In practice, the relationship between perceptions and experiences of corruption is both inconsistent across countries (there’s no reason why corruption should be perceived just as accurately in Bangladesh as in Burundi) and non-linear (there’s no reason why perceptions of corruption should ebb and flow with levels of corruption themselves).

But the CPI also presents an implausible consistency in individual countries' scores. In a recent article, we analysed this consistency over an 11-year period in CPI scores. The results are shown in the chart below.

Click to enlarge

This level of consistency across years we found is so strong that it clearly makes no sense to treat each annual score as separate. Given its lack of variation, the CPI might as well be published every five or ten years, since there would be very little loss in precision – hardly a resounding recommendation for such a widely used metric.

We repeated this analysis for another widely-cited perceptual measure of corruption, the World Bank’s Control of Corruption Index, and found essentially identical results: extreme consistency even across an 11-year time period.

These findings show that as a way to measure corruption, perceptual measures have serious – and still underappreciated – limitations.

Cooking the books

In response to concerns like these, there have been many attempts to create new corruption measures that rely on (seemingly) more objective sources – crime statistics or accountancy data, for example – in hopes that these data provide information about the objective rate of corruption.

That’s a plausible enough assumption; accounts might highlight where money is being spent with no return, and crime data might record the number of people actually convicted of a corruption-related crime.

But even these promising efforts still face big problems. Crime statistics, for instance, are particularly susceptible to subjective judgements about what corruption actually is: whether and when bribery, fraud or theft count, for example.

Using accountancy data, meanwhile, demands similar judgements about how to classify “missing” resources. Is all money that has disappeared indicative of corruption? Could it be a result of theft – and could that theft still count as corruption if so? Is chronic inefficiency in some areas necessarily a result of corruption?

A different tack

And most of all, many attempts to measure corruption are undercut by a core assumption: the idea of corruption as “one thing”, an indivisible property that can be captured by a single number or score – which in turn is applicable to the whole of an arbitrarily selected territory. As a result, many measurements of corruption are presented in league table format, purporting to show which countries are most or least corrupt.

In practice, of course, corruption actually takes place in concrete settings and specific places, in ways that do not easily map onto the nation-state. There may be significant variance at local level when dealing with corruption in particular sectors of an economy, corruption involving trans-national or cross-border networks.

Every country, meanwhile, has regions with conspicuously worse reputations for corruption than the national average would imply – think of Illinois in the US, nearly 90% of whose voters apparently believe corruption is “typical in state government.”

Stopping the flow

A major virtue of the approach the ONE Campaign has taken is that instead of constructing an international ranking of some kind, it highlights the scale of the problem across the developing world as a whole – and shows how the G20 nations can make a concrete difference. This is a big advance on the CPI-centric idea that simply comparing countries' individual corruption records is a serious way of addressing the problem.

As the report points out, the fact remains that many of the proceeds of corruption in the developing world are still allowed to flow through established financial centres in the developed world – London, New York, Hong Kong – via anonymous shell companies, who in turn funnel them to offshore tax havens.

These corrupt flows, still insufficiently understood (never mind policed), are the real corruption scandal – and cracking down on them is beyond the capacity of the developing nations they exploit.

The report’s research estimates that US$20 trillion of money illegitimately removed from developing countries is now held in offshore tax havens around the world. As the way global corruption is actually practised starts to be exposed, moving on from superficial and compromised ways of measuring the problem is the first step to ending it.

Authors

1. 

   Paul Heywood

   Sir Francis Hill Professor of European Politics at University of Nottingham

2. 

   Jonathan Rose

   Research Fellow, Faculty of Social Sciences at University of Nottingham