Comment: Illicit Financial Flows from Developing Countries 2001-2010
December 19th, 2012
December 19th, 2012
This week Global Financial Integrity released their periodic estimate of worldwide illicit financial flows authored by Dev Kar and Sarah Freitas. The report finds the developing world exported an estimated US$859 billion in illicit financial flows in 2010. In case you like moving words, here’s a presentation I put together on what that number means.
The GFI model of illicit financial flows includes two components: (1) an estimate of money that exits developing countries via trade channels (called trade mispricing) and (2) an estimate of money that leaves developing countries through other, private capital flows. Traditionally GFI has estimated the second type of flows using the World Bank Residual Method, which uses a country’s balance of payments (a bit like a balance sheet) to calculate leakages by comparing a country’s source of funds to its use of funds. By definition, if a country’s source of funds is greater than its use, there must have been an outflow. But as GFI lead economist Dev Kar notes, this is not necessarily an illicit outflow. This model does, to some extent, capture licit funds as well.
To correct for this fact, in this round Kar has modified his traditional model, instead using the Hot Money Narrow Method. In its classic sense, hot money involves the flow of capital between financial markets to earn a short term profit. It’s “hot” because it can move quickly and sometimes results in instability, particularly for developing countries. The Hot Money Narrow method calculates flows of illicit capital by going back to the balance sheet, but this method focuses on the “net errors and omissions” line item. Net errors and omissions (or NEO) is catch-all line to correct for the discrepancies between a country’s current account and capital account (which must be inversely equal). To the extent that the NEO captures flows of money (and not statistical errors) it must include only illicit flows. It is therefore a more conservative estimate than the World Bank Residual model.
I am a proponent of the new method. While it is likely illicit financial flows are underestimated, from an academic perspective it is useful to take a cautious approach. Also, the fact that many developing countries (e.g. China) have loosened capital controls in recent years would encourage the World Bank Residual Model to capture more licit funds. While capital flight in its entirety is useful to understand, GFI’s objective is to estimate illicit funds only. This adjustment will ensure it does just that.
The second part of Kar’s model, which estimates money lost through trade mispricing, is unchanged and a much more significant component of the total. GFI calculates these figures by comparing world trade data. Trade data is bilateral, which means that we can compare what County A says it exported to the world against what the world says it imported from Country A. If there is a discrepancy it could indicate an illicit outflow (or inflow). I say could because these statistics are by no means perfect. Some of the observed differences are just the result of statistical errors. But some of the difference is not statistical noise; it’s illicit transfers of wealth. And that’s what GFI tries to measure.
I do have one observation regarding the reporting of these numbers. Given that the top ten countries on GFI’s IFF list account for a pretty massive portion of the worldwide total, I think we need a sophisticated method to scale illicit financial flows. China alone exported an estimated $420 billion in 2010, a staggering number to be sure, but the nation also has a massive economy. I would argue the fact that Malaysia lost $64.4 billion in 2010 is just as important, if not more so, given the relative size of its economy. The heat map on page 21 nicely illustrates the total magnitude of IFFs, but does not shine as much light on their economic impact.
I know that this exercise wouldn’t be as simple as scaling IFFs to a particular measure, like GDP or population size. There are obvious problems with both. While I don’t think there is any perfect way to do it, and I do think the total numbers are significantly more important, it could be an important and useful exercise.
GFI, Dev Kar, and Sarah Freitas have produced a useful and relevant analysis here. As always, GFI continues to authorize excellent research, which, as I have noted before, does not always conform to the “herd mentality.” It is this kind of thinking that is required to change an unacceptable status quo.