In a comprehensive study on the level of trade misinvoicing in India in 2016, GFI found that the estimated potential tax revenue losses to the Indian Government that year is US$13.0 billion, equivalent to 5.5 percent of the value of India’s total government revenue collections in 2016. Trade misinvoicing is a method for moving money illicitly across borders which involves the deliberate falsification of the value, volume, and/or type of commodity in an international commercial transaction of goods or services by at least one party to the transaction and constitutes the largest component of illicit financial flows as measured by GFI.
Using a trade gap analysis, GFI was able to break down the estimated potential tax revenue losses to misinvoicing by measuring illicit financial inflows and outflows for both import and export under- and over-invoicing. GFI estimates that the value of the trade gap for misinvoiced goods equals US$74 billion, or 12 percent of the country’s total trade of US$617 billion in 2016.
Other notable findings:
- Of the total estimated potential lost revenue of US$13.0 billion, approximately US$4.0 billion was due to export misinvoicing and approximately US$9.0 billion was due to import misinvoicing.
- The US$9.0 billion in import misinvoicing can be further broken down by uncollected VAT tax (US$3.4 billion), uncollected customs duties (US$2.0 billion), and uncollected corporate income tax (US$3.6 billion).
- In 2016, some of the Indian imports most at risk for high values of import under-invoicing were edible fruits and nuts, sugars, vehicles and cereals.
- In 2016, some of the Indian imports most at risk for high values of import under-invoicing were from imports from USA, Australia, South Africa and Ghana.
- Looking at both high-risk commodities and high-risk trade partners in 2016, GFI found that under-invoiced imports of edible fruits and nuts from Ghana, mineral fuels from Australia and South Africa and electrical machinery from China were highlighted as potential high-level risks for revenue losses.
- In 2016, almost two-thirds of Indian imports that appear to be most at risk for some degree of potential revenues losses are imports from just one country – China, which was by far India’s largest source of imports in 2016.
GFI urges India to adopt a public registry of beneficial ownership information on all legal entities and to consider using GFI’s online tool GFTrade, designed by GFI to build the capacity of customs authorities to better detect misinvoicing as transactions are occurring and take corrective steps in real time. India should also encourage other countries to adopt a beneficial ownership registry, to fully implement FATF’s anti-money laundering recommendations, country-by-country reporting, tax information exchange initiatives and the Addis Tax Initiative.
To undertake a trade gap analysis, GFI uses data provided by the United Nations Comtrade (Comtrade) database, which each year collects reported data from most countries about their annual imports and exports. For this analysis, GFI used the Comtrade data for India in 2016 to cross reference India’s reports on its exports and imports against the corresponding reports submitted by all of India’s trade partners around the world for 2016. In these data sets, we looked for gaps in export and import statistics that are suggestive of trade misinvoicing.
India reported to Comtrade a total value of nearly US$356.7 billion in imports and nearly US$260.3 billion in exports for a total value of trade of US$617 billion in 2016. Drawing on this data, GFI then applied a series of treatments to the Comtrade data in order to undertake our trade gap analysis. These steps are described in detail below and in Section III. B. ‘Statistical treatments of the basic Comtrade data.’ After compiling India’s trade data and that of its trade partners for 2016, we then eliminated three different sets of trade data from consideration. We first eliminated all cases of “orphaned” imports – meaning those records in the database for which India reported a value for imports of a commodity or good from a particular country while that country reported no exports of that good to India in that year. Next, we eliminated all cases of “lost” exports – meaning records of exports reported by India’s trade partners as goods shipped to India in a particular year, but which were not reported as imported by India in that year.
After eliminating all cases of “orphaned” and “lost” records from the Comtrade data for India in 2016, we still needed to identify and eliminate a third category of records called “others”. Among the remaining records of “matched values”, i.e., trades for which both India and its trading partners reported values for that year, “others” are records that do not meet three criteria: 1) non-zero values for the trade must be reported by both the reporting country and its partner; 2) non-zero volumes (quantities) for the trade must be reported by both the reporting country and its partner; and 3) the volumes must be reported in the same physical units of measurement by both the reporting country and its partner. If any of the remaining records of “matched values”, did not comply with all three criteria, these were also eliminated as “others”.
Finally, once all of the cases of “orphaned”, “lost” and “others” records have been eliminated from the Comtrade data for India in 2016, and we have applied other technical treatments to the data (detailed in section III. B), we are then left with the remaining sets of “matched” trades to be used in our trade gap analysis. In our trade gap analysis, we identify any gaps found in the reporting data when the reported values by both partners do not match. For example, if India reported paying US$5 million for alarm clocks imported from China in 2016 but China only reported exporting US$3 million in alarm clocks to India in 2016, this would represent a trade gap of US$2 million. With India as our focus country, this would reflect a case of import over-invoicing by India.
For more information of GFI’s methodology and analysis, including limitations, please download the report and visit page 9.