— most IFF estimates are based on data anomalies on Licit financial flows (LFFs)
— The main datasets: the IMF’s Direction of Trade Statistics and Balance of Payments data, United Nations Commodity Trade Statistics Database, Bank for International Settlements datasets, Foreign Affiliates Statistics and Foreign Direct Investments datasets.
— Most datasets cannot be used to measure crime and corruption-related IFFs, and these types of IFFs are mainly measured using data from investigations, suspicious transaction reports, prosecutions, convictions or surveys
methods of estimating iffs
the source funds for IFFs may be perfectly legal, while the avoidance of tax, for example, may be technically legal yet illicit by societal norms.
- The Hot Money Method: estimates commercial-related IFFs through net errors and omissions in payment balances provided in different datasets. These anomalies are regarded as including illicit financial flows (Roy and Khalid 2015: 6).
- The Trade Mispricing Model: assesses trade-related IFFs by looking for disparities arising from over-invoicing of imports and under-invoicing of exports after adjusting for ordinary price differences.
- The World Bank Residual Method: estimates commercial-related IFFs as the difference between the source of funds (external debt and foreign direct investment) and the use of funds (current account deficit and reserves).
- The Dooley Method: relies on the privately held foreign assets reported in the balance
Using datasets of LFF licit financial flows, to estimate IFFs
Four main types of iff generating activities:
1. Tax and commercial IFFs:
egs: tariff, duty and revenue offences, tax evasion, corporate offences and market manipulation.
grey iffs: tax avoidance, transfer mispricing, debt shifting, relocation of intellectual property, tax treaty shopping, tax deferral and changing corporate structures and headquarter locations. When such activities directly or indirectly generate flows crossing country borders, they create IFFs.
2. IFFs from corruption:
egs: from corruption, such as bribery, embezzlement, abuse of functions, trading in influence, illicit enrichment, directly or indirectly, result in cross-border flows, they create IFFs
3. Theft-type activities and financing of crime and terrorism:
eg: financial flows from theft, extortion, illicit enrichment and kidnapping. When the associated financial flows cross country borders, they comprise of IFFs
4. IFFs from illegal markets:
domestic and international trade in illicit goods, such as drugs, firearms or services such as smuggling of migrants. IFFs are produced by the flows linked to the international trade of illicit goods and services, as well as by cross-border flows from operating the illicit income from such activities
After identifying the source activity, the second step is to estimate flows:
Bank for International Settlements (BIS) datasets:
(BIS) collaborates with central banks and other national authorities, to compile statistics on vital information such as financial stability, international monetary spillovers and global liquidity. The BIS produces two datasets of interest here: locational banking statistics (LBS) and consolidated banking statistics (CBS).
The LBS is a bilateral analysis of capital flows between countries, capturing the currency and geographical composition of internationally active banks’ balance sheets. it covers banking activities across borders, focusing on locations of the banking office. It covers the majority of the world’s transnational bank deposits, and all significant banking centres contribute to the dataset
International Monetary Fund’s Direction of Trade Statistics (DOTS) database
is the value of merchandise exports and imports disaggregated according to a country’s primary trading partners” . It depicts trade flows between major areas of the world, including area and world aggregates.
reports on imports are based on a cost, insurance and freight (CIF) basis and exports are based on a free on board (FOB) basis, with the exception of which imports are also available FOB.
Global Financial Integrity (GFI) uses the DOTS database to estimate IFFs for all countries. For instance, it selected DOTS bilateral reports for 148 countries trading with 36 advanced economies between 2006 and 2015
The top 30 countries, ranked by iffs, as a percentage of total trade:
Mozambique (48.1%), Malawi (44.1%), Zambia (43%), Honduras (39.7%), Namibia (38.7%) and Myanmar (30.8%). For the list of top 30 countries ranked by dollar value of illicit inflow, top countries included Vietnam (US$22.5 billion), Thailand (US$20.9 billion), Panama (US$18.3 billion), Kazakhstan (US$16.5 billion), Indonesia (US$15.4 billion), Belarus (US$6.1 billion), Argentina (US$4.8 billion) and Morocco (US$3.9 billion)
International Monetary Fund’s Balance of Payments (BOP)
is “a statistical statement that summarises transactions between residents and non-residents during a [given] period” . It consists of data drawn from the goods and services account, the primary income account, the secondary income account, the capital account and the financial account.
As with the DOTS, the IMF has warned against measuring IFFs using discrepancies in UN Comtrade as the trade invoices submitted by an importer and exporter could match though there are IFFs, and conversely, they might mismatch even when there is no illicit trade
in the Comtrade database for seven country- commodity pairs, including gold exports from South Africa. It calculated that “virtually all gold exported by South Africa leaves the country unreported”, pointing fingers at mining companies of smuggling billions of dollars’ worth of gold (UNCTAD 2016: 28). The South African government objected to the report, and in turn commissioned a report from economics consultancy Eunomix. The consultant concluded that mining companies and public agencies report gold exports, but not in a format compatible with UN Comtrade’s requirements (Mineral Council South Africa 2016). As such, they were able to account for three-quarters of the discrepancy in trade statistics. Hence, the experience of UNCTAD highlighted that one cannot automatically assume that Comtrade discrepancies in trade statistics are indications of IFFs.
Foreign Affiliates Statistics (FATS)
A foreign affiliate is defined as “an enterprise resident in the compiling country over which an institutional unit not resident in the compiling country has control, or an enterprise not resident in the compiling country over which an institutional unit resident in the compiling country has control” (Eurostats 2012: 13). The Foreign Affiliate Statistics (FATS), which are disseminated by organisations such as Eurostats2, the OECD3 and national statistical agencies, describe the overall activity of foreign affiliates divided into inward and outward FATS. Inward FATS show statistics of the activity of foreign affiliates resident
in the compiling economy, whereas outward FATS show statistics of the activity of foreign affiliates abroad controlled by the compiling economy
proven: about 40% of multinational profits (more than US$700 billion in 2017) are shifted to tax havens each year. Such profit shifting decreases corporate income tax revenue by more than US$200 billion, or 10% of global corporate tax receipts
Foreign direct investments datasets FDIDs:
The Global Revenue Dataset (GRD) is a compilation of data from multiple international and country-level sources on government revenues, which allows for more reliable and comparable cross-country tax research (
The GRD Explorer tool allows users to access the GRD, compare countries, regions, and indicators and visualise the data.6 It derives its information from sources such as IMF GFS, IMF Article IV
Corporate Tax Statistics Database
The OCED released in June 2020 the corporate tax database which aggregates information on the global tax and economic activities of nearly 4,000 multinational enterprises headquartered in 26 jurisdictions and operating across more than 100 jurisdictions
Most data is too aggregated to allow detailed investigation of specific base erosion and profit shifting channels (for example, there is no distinction between royalties and interest in related party payments, and no information on intangible assets).
Data based on financial accounting might not accurately represent how items are reported for tax purposes. Differences in accounting rules could affect the comparability of data between jurisdictions.
Several jurisdictions are yet to submit aggregated country-by-country statistics to the OECD for publication, affecting a broader coverage of tax statistics and getting a full picture on tax activities by multinational companies.
In the absence of specific guidance, multi- national companies may have included intra-company dividends in profit figures. This means that profit figures could be double counted, compromising the accuracy of the datasets.
In the case of stateless entities, the inclusion of transparent entities, such as partnerships, may give rise to double counting of revenue and profit, compromising the accuracy of the dataset.
Inclusion of pension funds or university hospitals could distort the relationship between profits and taxes
IMF SURVEYS
(CEPII) used this bilateral dataset to identify grey zones (which are potentially unethical) in global finance. Their finding was that the bulk of international assets in tax havens are “abnormal”, i.e. unexplained by standard gravity factors, and that such abnormal assets are increasing over time while being concentrated in six main jurisdictions: the Cayman Islands, Bermuda, Luxembourg, Hong Kong, Ireland and the Netherlands
Such a discovery is of significance because the agenda on tax avoidance and global finance might interact more than has been expected. If tax avoidance generates disproportionate assets of securities in certain jurisdictions, then the fight against it, especially in the context of shrinking fiscal space due to the COVID-19 crisis, might have unintended consequences on global financial balances
Tax Justice Network’s IFF Vulnerability Tracker:
The Tax Justice Network launched in June 2020 the Illicit Financial Flows Vulnerability Tracker which measures and visualises the vulnerability of a country to various forms of illicit financial flows over different periods of time.7 It is intended to assist countries to identify the trading partners and channels that create the greatest IFFs risks to their economies
FATF lists examples of data items deemed useful to collect and maintain for the purpose of mutual evaluations or other purposes (including but not limited to tracking illicit financial flows) while cautioning that such statistics may vary depending on national circumstances. These national datasets could range from the “number of legal persons and arrangements created and operating in the jurisdiction, broken down by: type of legal person and arrangement” to the number and value of suspicious transaction reports (STR) received
Indirect sources
Due to the challenges of tracking IFFs, indirect mechanisms may be employed to identify the source and/or destination of illicit financial flows. Public disclosure of the beneficial owners of corporate vehicles through the creation of public registers, for example, may be one such mechanism that makes it easier to flag irregularities and improve checks and balances in the global financial system (Roy and Khalid 2015).
Crime and corruption-related IFFs
Crime-based models, which estimate the scale of crime IFFs by limiting investigation and estimation to a crime type on geographic region or case study, calculate corruption-related IFFs….EG: the World Bank study “Ill-gotten Money and the Economy: Experiences from Malawi and Namibia” attempted to estimate the economic magnitude of ill-gotten money generated by different kinds of criminal activities in Malawi and Namibia
challenges of using datasets, to calculate IFFs
A major challenge in estimating IFFs from licit financial movements is the lack of access to data, mainly due to sensitivity and confidentiality issues. According to the report of the second expert meeting on the statistical measurement of illicit financial flows, an exercise to estimate IFFs in South Africa showed that data from many government agencies exists even at the level of the individual transactions but it is almost impossible to gain access due to sensitivity and confidentiality issues. This may also apply to private and public financial institutions who may refuse access to data due to privacy or data protection measures….Moreover, an absence of a universal methodology to monitor IFFs is believed to limit policy interventions and impair opportunities to tackle the challenge.
Estimates may be much higher or lower than the actual IFFs from the country. Ultimately, “most methods produce a constructed estimate, which will deviate from the true estimate by an error term of unknown size and direction”