A claim that “Exporters, importers shift Rs 13.2 trillion overseas via dodgy invoicing” circulates widely. Not a social media meme. An article published by one Kapila Bandara in the Sunday Times. According to the author, “Sri Lanka’s businesses involved in the exports and imports trade have plundered US$ 36.833 billion” during 2009-2017.
The author is not alone. Dhanushka Gihan Pathirana, publishing a newspaper article, states that “Between 2009 and 2018 a staggering $ 41.5 billion was transferred out of Sri Lanka by the corporate elite through trade mis-invoicing”. He is spreading the “news”, as a fact, through other mainstream and social media.
These claims are based on a recent report published by Global Financial Integrity (GFI), a Washington DC based think tank. If this claim is correct, it is a serious issue which needs immediate attention. To put it in the context, by end 2008, Sri Lanka’s total external debt was only US$ 15.1 billion. The amount reached US$ 51.6 billion by end 2017, an increase of US$ 36.5 billion.
If the claims in above articles are true, the increase in Sri Lanka’s foreign debt during 2009-2017 could have been completely avoided if this US$ 36.8 billion “plundered out of the country” could be saved. If the capital flow away from the country during 2018-2021 averaged the same value as during 2009-2017, the country would have been debt free now!
This claim, however, is nothing more than a gross misinterpretation of the GFI report. The report presents estimates for the possible trade mis-invoicing but does not interpret the number as an estimate for possible capital flight out of the county, though moving capital across borders can be one of the several motives behind mis-invoicing.
Trade mis-invoicing is an issue to be concerned, capital flight is another. Though somewhat interconnected, these are two very different issues. Highlighting the possible magnitude of trade mis-invoicing is admirable, both by the original researchers and those who introduced the report to Sri Lankans. However, mis-interpreting the report can be very dangerous, particularly in the current context.
The false alarm, not based on the GFI report or any other facts, can easily be used to organize people against the business elite of the country.
“Bring your money back to the country!”
This demand would make the business elite in the country utterly helpless. Some of them may have savings abroad, earned legally or illegally and moved out legally or illegally. But no person is likely to bring such money back under threat simple because a misinformed crowd, by nature, cannot be satisfied by doing so. They will continue to demand to bring the “remaining money” which nowhere exists. The end result could be nothing more than mass violence against the business community.
I do not accuse that the GFI report is being misinterpreted with this politically motivated ill-intention. The act may simply reflect the inability to understand the report. Yet, those who attempt to instill this idea should be cautious about the possible harm it can cause inadvertently.
The Global Financial Integrity Report
In this report, the authors present estimates for the possible magnitude of trade-related mis-invoicing as the sum of discrepancies between the numbers from each pair of counties using disaggregated trade data available at the UN Comtrade Database. This database is publicly available, and the methodology used for the analysis is clearly explained in the report. Therefore, any third-party researcher has the ability to replicate what GFR has done and verify their numbers. The method is not “highly technical”. Any person with some level of general intelligence can understand the method, no need of any specialized training in economics or international trade.
Based on the report, “the sum of the value gaps identified in trade between 134 developing countries and all of their global trading partners in 2018” is US$ 1.6 trillion, China is responsible for a fifth of the pie. The five countries with highest “value gaps” are China, Poland, India, Russia, and Malaysia.
China – US$ 305.0 billion
Poland – US$ 62.3 billion
India – US$ 38.9 billion
Russia – US$ 32.6 billion
Malaysia – US$ 30.7 billion
The differences between the numbers reported by Sri Lanka and its trading partners, defined as “value gaps” in the report, are presented in the report for each of the 9 years from 2009 to 2017. These numbers are correctly quoted in Kapila Bandara article.
When someone imports or exports a good, details about the transaction are recorded at the customs of the respective country. A summary of these transaction details is collected by the Word Trade Organization (WTO) from the member countries. The UN Comtrade Database has these numbers. The GFI research team has compared the numbers reported by each pair of countries during each year involving each category of items to identify any differences, the “value gaps”. Here is an example form the GFI website.
“For example, if Ecuador reported exporting US$400 million in bananas to the United States in 2016, but the US reported having imported only US$375 million in bananas from Ecuador in that year, this would reflect a mismatch, or value gap, of US$25 million in the reported trade of this product between the two trading partners for that year.”
The report clearly recognizes that “it is difficult to know which side of the transaction mispriced the shipment”. Therefore, the mismatch between the numbers from Sri Lanka and its trading partners during 2009-2017 of US$ 36.833 billion does not imply that the entire value gap is due to mis-invoicing by Sri Lankans. The wrong assumption implies that each and every trader partner has been completely honest. The mismatch, if the estimate is correct, reflects the aggregate sum of mis-invoicing by Sri Lankans and their trading partners. Another related limitation of this methodology is that it does not identify the cases of under invoicing and over invoicing separately; that decomposition is impossible.
The main reason for mis-invoicing is to evade taxes. The GFI report also clearly recognizes this. Here, are some hypothetical examples related to the Sri Lankan context:
(a) Under invoicing when importing, i.e., show a lower value than the actual value to pay less taxes. Sri Lanka customs charge taxes based on the estimated value of the item to prevent this issue.
(b) Importing under a different category to pay a lower tax rate. For example, bisecting a used car, importing the two pieces as spare parts, and fixing those together after importing and selling to local customers.
(c) Over invoicing when exporting- a BOI business may benefit by showing that they exported a higher quantity of their products than the actual quantity so that they can sell more than the allowed quota of 10% to the local market while enjoying tax-free benefits.
In those cases, the intention is tax evasion though capital flight can be a side effect. In the present context, trade mis-invoicing could be a popular pathway for moving capital away from the country, but I doubt whether it was so during 2009-2017.
Trade mis-invoicing and capital flight
Under what circumstances capital flight is possible through trade mis-invoicing and in which direction?
(a) Over invoicing exports – capital moves in
(b) Under invoicing exports – capital moves out
(c) Over invoicing imports – capital moves out
(d) Under invoicing imports – capital moves in
The estimates of discrepancies or “value gaps” in GFI report potentially includes all four types of these cases, even if we assume that all trade partners of Sri Lanka have been perfectly honest and only Sri Lankans have practiced mis-invoicing. Therefore, if someone claims that the estimate represents capital moving out from Sri Lanka, the second underlying assumption is imports are always overpriced and exports are always underpriced. If it has happened the other way round, the estimated number may imply a capital inflow of 36.833 billion during 2009-2017 and a potential foreign debt burden of US$ 88.4 billion if not for this capital flow!
It is weird to assume that the discrepancy is entirely due to unidirectional capital flows. This is most likely due to flows in both directions which may not necessarily sum zero. However, any resultant estimate of capital flight is unlikely to be very high. We do not have any information in the GFI report to verify this, but one can retrieve original raw data to investigate this, if needed.
Additionally, while under invoicing of exports also allows to evade taxes, over invoicing of imports is highly costly in terms of spurious tax payments. An importer must be in a highly desperate situation if the person wants to pay taxes to the government for a non-existent fake import. The premium for sending foreign exchange out via informal channels is usually cheaper than paying taxes on imaginary imports.
This argument does not preclude capital flight through under invoicing of exports. But the question is why. Foreign exchange earned by an importer is legitimate private capital. Though there are restrictions now, it is possible to move such capital out under normal circumstances. What is the purpose of working hard to save most of the earnings outside the country and never use that money or the returns? Exporters often leave their earnings abroad temporary but only to bring back later when the exchange rate turns favorable. This does not lead to permanent capital flight.
These last arguments are based on logical reasoning, not on empirics. However, they can also be supported using data. For example, in 2020, Sri Lanka’s exports to Uganda was US$ 7,671,775 according to Sri Lankan data but imports to Uganda from Sri Lanka have only been US$ 6,546,918 based on Ugandan data. This may show evidence of over invoicing by Sri Lankan exporters, under invoicing by Ugandan importers or a mix of both. There’s no easy way to find out what really has happened. The discrepancy could also be due to other reasons than mis-invoicing. For example, the shipment may have left Sri Lanka in December 2020 to reach Uganda in January 2021.
The GFI report disaggregates these data, and the exercise helps to identify more discrepancy. For example, Sri Lanka has exported “Rubber and articles thereof” worth US$ 401,667 to Uganda in 2020 while Uganda has received imports worth US$ 523,416 from Sri Lanka under the same category. This could possibly mean under invoicing by Sri Lankan exporters or over invoicing by Ugandan importers. If this means any capital flight the flow should be from Sri Lanka to Uganda and the value gap is US$ 121,749. But if we look at the “Beverages, spirits and vinegar” category, Sri Lankan data show an export value of US$ 2,753,685 while the respective import value is only US$ 2,689,121, a value gap of US$ 64,564. However, if this implies a capital flight the direction is from Uganda to Sri Lanka, which offsets a part of the previous flow. The discrepancy shown in the GFI report is the sum of these two numbers, not the difference. That is because the intention of GFI is to estimate the effect of trade-related mis-invoicing, not capital flows.
Trade data for identifying suspicious transactions
Does that mean trade data are not helpful for identifying suspicious capital flows? The writer, by no means, attempts to argue against the possibility of illegal financial flows through mis-invoicing. The argument, simply, is that the interpretation of the identified discrepancy, or the sum of value gaps, as an amount of capital moved away from the country is grossly wrong. Yet, disaggregated trade data are useful to identify suspicious transactions, which allows to initiate further inquiries and investigations. For example, Sri Lanka has exported “Printed books, newspapers, pictures and other products of the printing industry; manuscripts, typescripts and plans” worth US$ 2,180,576 to Uganda in 2020 but the value declared at the other end is a mere US$ 278. This can be a tip of an iceberg!