Global Financial Integrity


Magnitudes versus Methodologies?


Global Financial Integrity is pleased to note growing interest in the estimation of illicit financial flows and their effect on emerging market and developing countries. We are writing to offer a series of thoughts surrounding the reality of this concern and its political significance.

Illicit Financial Flows on the Global Agenda

Global Financial Integrity (GFI) has devoted the 10 years of its existence to anchoring onto the political-economy table the issue of illicit financial flows (IFFs) and the oppressive burden these flows place upon economic, political, and security issues in 150 countries and on billions of affected citizens. Together with the work of many other organizations and individuals, we are very gratified to see this fundamental concern now firmly ensconced in the Sustainable Development Goals, the Financing for Development document, and the Addis Tax Initiative.

World leaders have agreed that the order of magnitude of illicit financial flows compels attention and that the international community must urgently take steps to curtail these flows. Given that we are dealing with unrecorded, hidden financial flows, we will never have exact data on the size of this problem. Yet, like corruption, like terrorist financing, like criminality, all of which contain uncertainties as to their definitions and scopes, we have come together to recognize the severity of the problem and have together committed ourselves to its curtailment. This constitutes a major step beyond the Washington Consensus, as we enter into the Sustainable Development Goals with a sense of mutual accord and shared responsibilities. Moving forward, the relationship between the magnitudes of illicit flows and the methodologies surrounding their analysis needs to be put into perspective.


GFI generally focuses primarily on the issue of illicit financial outflows from emerging market and developing countries, essentially because we regard these outflows as the most damaging aspect of the problem. Illicit outflows drain hard currency reserves, heighten inflation, reduce tax collection, limit investment, and impair free trade. In most cases such outflows convey resources permanently into western coffers, creating persistent inequality and delayed development.

Our estimates indicate that in 2013 illicit outflows shifted $1 trillion out of the 150 countries we analyze. And this phenomenon has been growing at some six percent per year. The published sources of data on which we draw also suggest that the commercial sources of IFFs are larger than the corrupt and criminal sources of these flows. The commercial component is very often transmitted through over-invoicing of imports and under-invoicing of exports, a technique that may also be used to move corrupt and criminal funds.

These broad parameters are consistent with what we observe to be the explosion in global money laundering and cross-border crime. Despite decades of anti-money laundering efforts, every expert resource acknowledges that this global scourge, estimated in trillions of dollars, continues to grow. And despite decades of fighting drug trafficking, human trafficking, counterfeiting, and other transnational crimes, these activities, likewise in the trillions of dollars, continue to increase. We are dealing with a reality that brings harm to the lives of billions of people.

With global emphasis now focused on domestic resource mobilization, GFI has also been emphasizing its analysis of illicit financial inflows. Illicit inflows, often occurring through under-invoicing of imports and over-invoicing of exports, reduce the collection of VAT taxes, customs duties, and company income taxes and encourage fraudulent claims for VAT refunds. Governments are leaving billions of dollars on the table every year that should instead be increasing domestic resources.

Perhaps the most important point in the measurement of magnitudes of illicit financial flows is the fact that our available data touch only a part of this problem. Not included in any of the data we compile are the following:

Same invoice faking, done as a matter of agreement between importers and exporters, does not create a gap in comparative trade data.

Misinvoicing of services and intangibles, because it is easy to misinvoice such transactions and virtually no useful data exist on the global magnitude of these categories of commerce.

Cash movements across borders, as occurs in many instances of cross-border criminal activity, because there are no reliable data on this phenomenon.

If we had decent data on these components of IFFs, the evident magnitudes of IFFs would be considerably higher.


Illicit financial flows can be indicated in 1) the errors and omissions component in balance of payments (BOP) statistics, and 2) excessive gaps in comparative export and import figures in bilateral trade statistics. Both of these data sources have been used for more than a half century. GFI did not create the focus on errors and omissions in BOP statistics. Likewise GFI did not invent the analysis of gaps in export and import statistics as indicative of trade misinvoicing. GFI was the first organization to take this gap analysis and apply it to data from all emerging market and developing countries to arrive at a global analysis of trade misinvoicing.

GFI has twice amended its methodology. First, it was pointed out that broad measures of capital flight might include some components of financial transfers that are legitimate. We studied the point and made appropriate adjustments which reduced our BOP calculation of illicit flows to errors and omissions only. Second, it was noted that Hong Kong data is available which reduces evident trade misinvoicing out of China. Again, we studied the data from Hong Kong and factored it into our calculations which modified our China illicit flows estimate. Neither of these adjustments materially affected the global magnitudes of IFFs in our analyses.

More recently it has been suggested that GFI’s analyses may contain other issues, some of which are as follows:

The markup factor from free on board (FOB) to cost, insurance, and freight (CIF). GFI utilizes a 10 percent markup factor as does the IMF. We have run calculations assuming a 5 percent figure and a 15 percent figure and have found that neither of these alternatives materially affects the total flows estimates produced. We understand that the IMF is studying this question further to see if the 10 percent figure can be refined for particular commodity groups and geographical distances, and we look forward to incorporating such refinements into our analyses.

Currency fluctuations. Currencies are traded millions of times daily, and it is impossible to have a perfect measure of fluctuations in values. Again, GFI has carefully studied the issue, and we see no reason to believe that broad measures of currency fluctuations introduce any significant concerns into estimates of illicit financial flows.

Netting inflows and outflows. It has been suggested that illicit financial outflows and illicit financial inflows should be netted. There is no concept of net crime, the net of illicit activities affecting a country. Both illicit outflows and inflows are harmful to resource mobilization in different ways. A recent World Bank paper acknowledged that illicit outflows and inflows should more properly be added together, thus providing a more comprehensive analysis of the impact of this phenomenon. GFI does not net the two and has no plan to do so.

Noise in the data. Every economic study of which we are aware is dependent upon data that may contain flaws. In the IFFs agenda, such flaws may arise from mistakes in entries of trade data, mistakes in classifications of goods traded, mistakes in BOP components, etc. Indeed, the IMF has amended its original 1948 BOP manual five times—in 1951, 1961, 1977, 1993, and 2013. And there still remain significant concerns about GDP calculations. So fundamental as to put into question whether a sophisticated economy such as Japan is in fact growing or declining. Noise within national statistics is a reality regardless of the purposes to which the calculations are being put, yet such statistics are used as the basis for analyses underpinning major global economic decisions every day. GFI works with the same data.

GFI constantly reviews its methodologies and statistics. Starting in 2017 we will publish on our website on a quarterly basis regularly revised IFFs data in order to take into account amendments made by reporting countries, often correcting preceding data provided. Amendments to data and amendments to methodologies will continue as appropriate.

Macro or Micro?

GFI uses macroeconomic data for its annual global analyses and individual country analyses, and we further bolster our individual country analyses with microeconomic commodity-level data as reported by UN Comtrade and by individual nations publishing their trade statistics. Both approaches help create an understanding of the magnitudes of illicit flows and both have their challenges.

The macroeconomic data indicating orders of magnitude of the IFFs problem hurting emerging market and developing countries as a whole and countries individually have served to secure the IFFs issue onto the global agenda. Without this data, IFFs would not be included in the commitments made by nearly every nation of the world to address the very serious harm done by these resources shifting out of and into middle income and poorer countries.

Commodity-level data can be especially useful in making decisions about where to focus oversight and investigative efforts. We find, for example, that natural resource exports from developing countries are frequently under-invoiced, and machinery and equipment imports into these countries are frequently over-invoiced. GFI has for several years provided this kind of detailed information to selected developing countries on a confidential basis.

Recognizing the value of transaction-level commodity data, GFI has developed a comprehensive global trade database, encompassing 80,000 Harmonized System categories of goods as exported and imported by 30 reporting countries. GFI is making this system, called GFTrade, available to emerging market and developing countries, so that Customs departments can see instantly if transactions may be fraudulently valued and need to be investigated further. This is a risk assessment tool, enabling such assessments to be made in real time as transactions are on the docks awaiting import or export. Furthermore, this database enables GFI and partner Customs departments to perform very detailed analyses at the level of individual commodities and even individual transactions to determine the risk of trade misinvoicing.

Magnitudes and Methodologies

All aspects of the illicit flows and trade misinvoicing issues need to be continuously developed. With more comprehensive data, components of the IFFs problem not presently subject to analysis can be included. This will undoubtedly increase the level of illicit flows currently estimated in both directions.

With individual commodity data, we can work quite purposefully with developing countries to curtail misinvoicing in selected categories of imports and exports. GFI does this currently.

Improving methodologies will continue to be a constant effort of GFI and other researchers, as we strive to make our data as accurate as can be achieved. Given the inherent nature of unrecorded, hidden, illicit flows, perfection will be unattainable.

None of the data and methodology issues or choices of what to analyze affects the overwhelming importance of the IFFs reality. Suppose, for example, that in the future with all-encompassing data inclusive of components not now measurable and a perfected methodology widely agreed, global illicit outflows and illicit inflows could be accurately calculated as follows:

In Billions of U.S. Dollars

 Illicit OutflowsIllicit InflowsCombined
Estimate A$500$500$1,000
Estimate B$1,000$1,000$2,000
Estimate C$1,500$1,500$3,000

What policies for curtailing IFFs would be changed based upon the most accurate of these numbers? None. The magnitude of the problem is severe at every level.

Going forward, it is not a question of magnitudes versus methodologies. We must strive to very substantially reduce the magnitude of the IFFs problem. And we must strive to perfect the best and most accurate analytical methodologies available to serve this objective.

Illicit financial flows impair the lives of billions of people across our shared world. Curtailing these flows can help nations reach their Sustainable Development Goals aimed at improving health services, education, water, sanitation, food security, infrastructure, the environment, economic growth, and political stability.

GFI will be pleased to enter into any discussions as may be useful in achieving these ends.