Illicit Financial Flows through Trade Mis-Invoicing in India: An Empirical Analysis of the Major Commodities Involved in Mis-Invoicing

Abstract

This article identifies, based on reported trade data between India and its 19 trading partners, the major commodities exhibiting mis-invoicing during 2000–2018, the extent of mis-invoicing, major trade partners and the associated ports of trade in India. The computed differences in the trade values are too large to be explained by accounting or classification errors, presenting strong evidence of mis-invoicing; however, only a small percentage of commodities account for bulk of mis-invoicing consistently over the years. There is also evidence of the same commodities exhibiting under-invoicing (UI) in trade with some countries and over-invoicing (OI) in trade with others. The tariff rates seem to influence the type of import mis-invoicing—OI being mainly in commodities with higher tariff and UI in commodities with lower tariff. The article contributes to the existing literature by identifying the specific commodities with their 6-digit HS codes, and commodity groups prone to mis-invoicing, which can provide a robust framework for further investigation of transaction level data that may help pinpoint the parties involved in mis-invoicing and associated illicit flows. The findings of the article provide inputs for policies to mitigate the impact of illicit financial flows through trade mis-invoicing.

JEL Classifications: F13, F14, C32, H26

Keywords

Trade mis-invoicing under-invoicing over-invoicing imports exports

Introduction

Illicit financial flows (IFF) have been receiving considerable attention after being included in the UN’s Sustainable Development Goals 2030. Trade mis-invoicing (TM), which according to The Economic Commission for Africa (2010) is the act of misrepresenting the price or quantity of imports or exports in order to hide or accumulate money in other jurisdictions, is believed to be a major source of cross border IFF because international trade provides large volumes and access to the international financial system, enabling large-scale movement of funds with ease. Illicit outflows can be facilitated by over-invoicing (OI) imports (more money is paid for the imports than their real value) and under-invoicing (UI) exports (export proceeds are less than the value of goods exported). Illicit inflows can be facilitated by UI imports and OI exports. One of the popular methods of computing TM involves comparing the import and export data reported by the trading countries since the import of one will be the export of the other. The computed trade gaps, after making accounting adjustments, are taken as a measure of TM. This approach, however, suffers from a major handicap. While the mis-invoicing takes place at the transaction level, the transaction level data is not available in public domain. The publicly available data is at a higher, aggregated level; for example, national level data aggregated for each commodity or national level aggregated trade data (export and import). Since the UI and OI at the transaction level get cancelled out in aggregation, the estimates using commodity level data will provide more realistic estimates of TM than the aggregated trade (import and export) level data but less accurate estimates than computed from transaction level data. Tiwari (2021) calls this phenomenon as the Aggregation Paradox, and has estimated that in case of India’s trade with 19 major trading partners over 2000–2018, the import UI of $441.77 billion, computed from aggregated national level trade data, is actually aggregation of import UI of $684.54 billion and import OI of $244.77 billion, computed from disaggregated, 6-digit HS code commodity level data. Similarly, the export UI of $29.98 billion, computed from aggregated national level trade data, is actually aggregation of export UI of $307.31 billion and export OI of $277.33 billion, computed from data disaggregated at 6-digit HS code level. The estimates will be much more when transaction level data are used.

The Aggregation Paradox underscores the importance of using the most detailed commodity level data to identify the commodities predominantly involved in TM, given that transaction level data is not available in public domain. In this article, we discuss the results of our analysis of 6-digit HS-code commodity data, sourced from UN Comtrade, and identify the major commodities involved in import and export mis-invoicing and their ports of trade. We also seek answers to questions whether the extent and direction of mis-invoicing in the commodities are consistent over time and across trading partners. These findings can be useful in shaping appropriate policies for mitigating TM and the associated IFF.

Survey of Literature

The literature on TM can be traced back to the 1950s, staring with Morgenstern (1950), who studied the trade gaps in the reported trade statistics between France and England and provided a theoretical framework for TM. Carton and Slim (2018) have categorized the subsequent work into four distinct phases. The first phase (1960s and 1970s) established the theoretical and empirical framework. The notable works in this phase were Bhagwati (1964, 1967), Bhagwati and Hansen (1973), Bhagwati et al. (1974) and Bhagwati (1981). The second phase (1980s and 1990s) contributed to a better understanding of the interlinkages between TM and illicit capital flights (CFs). The third phase started around the 2000s and focused on practices such as trade-based money laundering. The fourth phase, which started about a decade back, has focused on the development financing implications of TM and IFFs. It received impetus after the report of the UNECA (2015) and inclusion in the UN’s Sustainable Development Goals of target 16.4 relating to the IFFs.

The Partner Country Method (PCM) uses the mirror trade statistics to determine the trade gaps. Initially proposed by Morgenstern (1950), the early contributions to this approach consist of Bhagwati (1964) and Morgenstern (1965). The former compared the cost, insurance and freight (CIF) import values and the free on board (FOB) export values for Turkey’s major trading partners, and reported evidence of import UI in products with high tariffs. McDonald (1985) found a positive correlation between mis-invoicing and export taxes. Yeats (1978) reported variations in the extent of TM across product categories.

Marjit et al. (2000) pointed that India’s exports to the USA during 1951–1996 were almost always under-reported and suggested that devaluation reduced the black-market premium on the foreign exchange, which in turn reduces under-reporting.

Fisman and Wei (2001, 2004) examined the trade between China and Hong Kong and found that 1% increase in the tax rate led to 3% increase in the ‘evasion gap’, defined as the difference between the reported exports and imports. Mishra et al. (2007), in a similar study for India, reported a 0.1% increase in evasion for each percentage point increase in tariffs.

Biswas and Marjit (2007), using a three-country model, showed that initially the low or zero tariff is likely to encourage illegal capital outflow and higher tariff is conducive to illegal transactions in foreign exchange. However, eventually, a low tariff regime takes care of illegal capital outflow as well as the black market for foreign exchange.

de Boyrie et al. (2005, 2007) and Zdanowicz (2007) provide a perspective on the movement of illegal money through commodity transactions. Marjit et al. (2008) proposed that highly controlled and regulated environment leads to misinterpretation of official statistics. Berger and Nitsch (2008) found high correlation between the reporting gaps and the level of corruption in both partner countries. Beja (2008) reported TM of US$287.6 billion between China and its trade partners during 2000–2005. Buehn and Eichler (2011) show a positive link between TM and quotas and tariffs.

Carrere and Grigoriou (2014) analysed the HS 6-digit Comtrade data and reported that at the disaggregated level, the trade transactions consist not only of matched transactions (those for which positive trade values have been reported by both sides) but also of orphan (import without corresponding export from the partner) and lost (no reported import but there is reported export from the partner) transactions, and that the percentage of matched transactions decreases with increasing degree of disaggregation, viz. from HS2 to HS4 and HS6 levels.

Kar and Spanjers (2014) of the Global Financial Integrity (GFI) provided an assessment of the magnitude of IFF from developing countries. They combined the two principal approaches to estimating the illicit flows, viz. analysis of (a) balance of payments statistics or other macroeconomic data and (b) international trade statistics, and add the values obtained from both approaches to arrive at the total illicit flows. Jha and Truong (2014) analysed the bilateral UN Comtrade trade data to compute TM for India’s trade with 17 countries for the period 1988–2012 and estimated that the cumulative CF between 1988 and 2012 exceeded $186 billion.

Nitsch (2016) questioned the main assumptions of GFI, viz. the Gross Excluding Reversals (GER) approach, which focuses only on the outflows; a flat 10% deflator for freight and insurance costs; and the assumption that advanced countries’ data was free of errors and mis-invoicing as compared to developing countries. Nitsch suggested an approach consisting of more micro evidence obtained ‘from the field’, restricting the empirical analysis to a few large countries that account for the overwhelming majority of illicit outflows, and analysis of observed differences in matched partner trade statistics at the product level. Hong and Pak (2017) also questioned the assumption that there was no mis-invoicing in partner countries and proposed an alternative method that does not rely on the trade statistics of partner countries.

A UNCTAD study (2016) that analysed the primary commodity exports from Chile, Ivory Coast, Nigeria, South Africa and Zambia, using disaggregated data from the UN Comtrade database, showed substantial TM in all five countries, with notable variation in the patterns across countries, products and trading partners.

Qureshi and Mahmood (2016) estimated the TM in Pakistan and reported that import UI had a positive relationship with customs tariffs and interest rate and reduced with improvements in the current account balance and political stability. Kellenberg and Levinson (2016) showed that in addition to tariffs, the reporting differences also vary systematically with country characteristics like incomes, auditing standards, corruption and trade agreements.

Carton and Slim (2018) applied the PCM to the bilateral annual trade data from UN Comtrade in 35 OECD countries over 2006–2016, and concluded that TM remains a significant factor in explaining the trade discrepancies and that the current understanding of the TM can be improved by more research at the commodity level.

Gara et al. (2018) analysed the mirrored data of Italy’s external trade from 2010 to 2013 at 6-digit level of goods classification and found robust correlations between trade gaps and differential tariff and income tax rates, and trade openness.

Cheung et al. (2020) estimated CF to Germany through TM and showed that economic policy uncertainty, European Commercial Bank (ECB) collateral policy and currency misalignment are more important driving factors than the traditional determinants such as covered interest differentials, which play only a limited role.

GFI (2019) used the Comtrade data and estimated that the revenue loss to India in 2016 due to TM was US$13 billion.

Biswas et al. (2022) reported significant differences in the FDI and trade data for the United States and China during 1983–2017 and attributed the differences to hidden capital flows, based on vector auto regression (VAR) and auto regression distributed lag (ARDL) models analysis. They found that an increase in hidden capital outflows through the trade channel triggers a hidden capital inflow through the FDI channel.

Tiwari (2021), based on an analysis of the bilateral HS-6 digit UN Comtrade data in respect of India’s trade with 19 major trading partners over 2000–2018, has demonstrated that the illicit inflows through TM in case of India have exceeded the outflows in most of the years; that the computed amount of mis-invoicing will vary according to the level of aggregation of the underlying data; that the orphan and lost trade, ignored in the computations of TM, could have a significant impact on the amount of TM; and that there is a statistically significant relationship between the illicit outflows through TM and inflows through FDI.

Objective of the Research, Data and Research Methodology

From a review of the literature, it is quite clear that TM is a significant issue in studying the cross border IFF and that in the absence of transaction level data the partner country trade statistics must be analysed at the most disaggregated commodity level in order to get more realistic estimates than the estimates using aggregated trade data. The objective of this article is to identify in case of India the major commodities, at the most detailed, 6-digit HS code level, that display TM, using the UN Comtrade (2014) data (http://Comtrade.un.org/data/) for the period 2000 to 2018. The 19 countries included in this study are Australia, Belgium, Brazil, China, Hong Kong, France, Germany, Indonesia, Italy, Japan, Malaysia, South Korea, Saudi Arabia, Singapore, South Africa, Switzerland, Thailand, UK and USA. The countries were selected such as to include at least one country from each continent. UAE, a major trading partner, was excluded as there were gaps in the data reported at the HS-6 level.

We use the PCM to estimate trade gaps. PCM compares the values of import and export reported by India with the values of corresponding export or import reported by the partner country. The computed trade gap is taken as an indicator of TM and the associated IFFs. In doing so, we deflate the CIF import values by a factor of 1.1 (IMF, 1993). Although the IMF has revised the factor to 1.06 (Marini et al., 2018), we use 1.1 to keep the estimates conservative.

The data analysed by us using Tableau and MS-Excel contains transactions in 5,008 commodities valuing $5.13 trillion; however, only transactions valuing $4.39 trillion (86%) were analysed as the remaining transactions, though reported by India, did not have corresponding transactions reported by partner countries and therefore were not amenable to the PCM.

We use the notation M_i for import reported by India, E_j for the corresponding export reported by the partner country j, E_i for export reported by India and M_j for the corresponding import reported by the partner country. The mis-invoicing is represented by the following expressions:

Import mis-invoicing: [(M_i/1.1) – E_j]: A negative value indicates UI. A positive value indicates OI.

Export mis-invoicing: [E_i – (M_j/1.1)]: A negative value indicates UI. A positive value indicates OI.

We also use the following two ratios for analysis:

CIFFOB ratio = (M_i/1.1)/E_j

FOBCIF ratio = E_i/(M_j/1.1)

Analysis and Findings

A Small Percentage of Commodities Account for Bulk of Mis-Invoicing

Although the mis-invoicing is seen in most of the commodities, only a few commodities account for the bulk of mis-invoicing. The top 100 commodities, which constituted only 3% to 6% of the number of commodities mis-invoiced, have accounted for 58% to 73% of the amount of mis-invoicing (Table 1).

Table 1.

Percentage of Commodities that Display Mis-Invoicing

		No. of Commodities Mis-Invoiced	Amount of Mis-Invoicing $b	Amount Accounted for by Top 25	Amount Accounted for by Top 50	Amount Accounted for by Top 100
1	Import under-invoice	3,352	686.54	40%	48%	58%
2	Import over-invoice	1,656	244.77	49%	59%	70%
3	Export under-invoice	2,695	307.31	38%	49%	62%
4	Export over-invoice	2,313	277.33	50%	61%	73%

Source: The authors.

The Number of Commodities Displaying Import Under-Invoicing Is More than Double the Number Displaying Import Over-Invoicing

Of the 5,008 commodities, 3,352 (67%) showed import UI totalling $686.54 billion (27% of total import of $2.53 trillion). The remaining 1,656 commodities (33%) showed import OI of $244.77 billion (9.7% of total import of $2.53 trillion) (Table 1). The number of commodities indicating UI (3,352) is more than double the number of commodities indicating OI (1,656), and the amount of UI is almost three times the amount of OI, which would seem natural in view of the propensity to evade customs duty through under-valuation. The sum of the import OI of (+) $244.77 billion and UI of (–) $686.54 billion was net import UI of $441.77 billion. The succeeding paragraphs discuss the granular results of commodity level data analysis, which presents a more realistic picture of the nature and extent of mis-invoicing.

Only a Few Commodity Groups Contribute to Bulk of Import Under-Invoicing

The Government of India classifies the commodities under 99 chapters, each representing a group of commodities. The individual commodities are assigned 6/8-digit codes, corresponding to the international Harmonized System (HS) of commodity classification. The first two digits of a commodity code represent the chapter number.

Our analysis indicated that 79 of the top 100 commodities, which account for 58% (around $400 billion) of the import UI of $686.54 billion, belong to only 11 (of the 99) chapters of commodity classification. Table 2 shows that the commodities falling under Chapters 71, 27, 84 and 85 account for the maximum amount of import UI. The highest number of commodities, viz. 22 and 19 falls under Chapters 84 and 85, respectively.

Table 2.

Chapter-wise Distribution of Top 100 Commodities Showing Import UI

Chapter	Chapter Name	No. of Commodities	Amount of UI $ b
71	Natural or cultured pearls, precious or semi-precious stones, precious metals, metals clad with precious metal and articles thereof; imitation jewellery; coin	8	110.44
27	Mineral fuels, mineral oils and products of their distillation; bituminous substances; mineral waxes	4	83.17
84	Nuclear reactors, boilers, machinery and mechanical appliances; parts thereof	22	48.71
85	Electrical machinery and equipment and parts thereof; sound recorders and reproducers, television image and sound recorders and reproducers, and parts and accessories of such articles	19	46.25
88	Aircraft, spacecraft and parts thereof	1	17.49
29	Organic chemicals	8	12.86
87	Vehicles other than railway or tramway rolling-stock, and parts and accessories thereof	4	9.54
72	Iron and steel	6	9.29
39	Plastics and articles thereof	4	8.88
15	Animal or vegetable fats and oil and their cleavage products; prepared edible fats; animal or vegetable waxes	2	7.47
999999	Commodities not specified according to kind	1	5.9
	Total	79	360

Source: The authors.

Figure 1 plots the top 20 commodities and the extent of their mis-invoicing. Bituminous coal, diamond, gold, fixed wing aircraft and silver in unwrought form are some of the most important commodities that show import UI.

The overall CIFFOB ratio for these commodities ranges from as low as 0.07 (Bituminous coal) to 0.96 (Gold in unwrought forms on monetary). Bituminous coal (code 270112) shows the lowest CIFFOB ratio (0.07) and highest amount of UI ($75.34 billion), more than even its reported import value of $5.95 billion and accounts for 10.97% of total import UI of $686.54 billion. The extremely low CIFFOB ratio of 0.07 means that the import value recorded by India is only 7% of the export value recorded by the partners. Gold in unwrought form (code 710812), which accounted for the highest amount of imports ($166.77 billion) is only at 10th place (Figure 1) in terms of the amount of UI ($5.57 billion or 0.81% of total import UI of $686.54 billion).

Figure 1.

Source: The authors.

Import Over-Invoicing Is also Dominated by a Small Percentage of Commodities Falling Under 10 Chapters

To recall from Table 1, only 6% (the top 100) commodities account for 70% ($171.46 billion) of the import OI of $244.77 billion. Table 3 shows that 73 of the top 100 commodities that account for $146 billion of import OI fall under only 10 (of the 99) chapters of the commodity classification. The maximum number of commodities fall under Chapters 84 and 85; however, the maximum amount of OI is accounted for by the commodities falling under Chapter 27 (mineral fuels), 88 (aircraft) and 85 (electrical machinery, etc.).

Table 3.

Chapter-wise Distribution of Top 100 Commodities Showing Import OI

Chapter	Chapter Name	No. of Commodities	Amount of Over-Invoicing $ b
27	Mineral fuels, mineral oils and products of their distillation; bituminous substances; mineral waxes	5	47
88	Aircraft, spacecraft and parts thereof	3	20
85	Electrical machinery and equipment and parts thereof; sound recorders and reproducers, television image and sound recorders and reproducers, and parts and accessories of such articles	14	16
84	Nuclear reactors, boilers, machinery and mechanical appliances; parts thereof	18	13
29	Organic chemicals	3	13
72	Iron and steel	8	10
38	Miscellaneous chemical products	6	8
39	Plastics and articles thereof	5	7
89	Ships, boats and floating structures	4	7
73	Articles of iron or steel	7	5
	Total	73	146

Source: The authors.

Figure 2 shows the top 20 commodities (1.21% of total over-invoiced commodities) that accounted for 45.7% ($111.87 billion) of import OI. These comprise coal except anthracite (HS 270119) that accounts for the maximum 17% ($40.6 billion) of the total import OI of $244.77 billion, followed by fixed wing aircraft (HS 880230) that accounts for 5% and organic compounds (HS 294200). The CIFFOB ratio ranges from 1.32 (HS 270400) to 177.11 (HS 271390). The values of the CIFFOB ratios are significant. For example, the CIFFOB ratio for HS 270119 (coal) indicates that the import recorded by India was 2.36 times the export recorded by the partner countries. Almost all the top 20 commodities display unusually high CIFFOB ratios, topped by HS 271390 (residues of petroleum oil) that displays a ratio of 177.11 and HS 294200 (organic compounds) that shows a ratio of 34.73. Such high ratios cannot be attributed to accounting or classification errors and need detailed investigation.

Figure 2.

Source: The authors.

Import Over-Invoicing Is Predominantly in Goods with Low Tariff While Import Under-Invoicing Is Predominantly in Goods with Higher Tariffs

Mishra et al. (2007), in a study for India, reported a 0.1% increase in evasion for each percentage point increase in tariffs. Intuitively, the tariff rates would influence the mis-invoicing decision. For example, UI, which is primarily to evade tariff, would more likely be in commodities with high tariff. Similarly, OI, which is primarily for facilitating illicit capital outflows, would tend to take place in commodities with low tariff so as to pay the minimum possible tariff on the over-invoiced commodities.

The top 100 commodities showing import OI fall under 29 chapters of the HS classification. The tariff rates for these commodities (taken from https://data.wto.org/) over the period 2000–2018, when plotted against the average tariff rates for the chapters under which these commodities fall (Figure 3), indicate that 46.47% to 74.7% commodities had tariff rates less than or equal to the chapter average. While this corroborates the assumption that import OI is more likely in commodities that have relatively lower rates of customs duty, it also implies that the import duty is not the sole driver of import OI.

Figure 3.

Source: The authors.

Similarly, the tariff rates for the 52.5%–78.5% of the top 100 commodities showing UI were greater than or equal to the chapter average tariff rates (Figure 4) corroborating the assumption that import UI is likely in commodities that have relatively higher rates of customs duty.

Figure 4.

Source: The authors.

More Commodities Show Export Under-Invoicing than Export Over-Invoicing

As shown in Table 1 the number of commodities showing Export UI (2,695, i.e., 54% of the total 5,008 exported commodities) is more than the number of commodities showing export OI (2,313, i.e., 46%). The main countries to which the exports were under-invoiced are China, USA and South Korea. The prominent countries the exports to which were over-invoiced are Saudi Arabia, Hong Kong, Singapore and South Africa.

Only a Few Commodity Groups Dominate Export Under-Invoicing

Table 1 showed that the top 100 commodities accounted for 62% of the export UI of $307.31 billion. Table 4 shows that 68 of these commodities that accounted for over $157 billion of export UI fall under only 13 of the 99 chapters or commodity groups, the prominent among them being Chapter 26 (ores, slags and ash), 29 (organic chemicals), 63 (other textile articles), 61 (articles of apparel), 64 (footwear) and 30 (pharmaceuticals).

Table 4.

Chapter-wise Distribution of the Top 100 Commodities Showing Export UI

Chapter	Chapter Name	No. of Items	Amount of UI $ b
26	Ores, slag and ash	3	32.71
29	Organic chemicals	17	29.87
63	Other made-up textile articles; sets; worn clothing and worn textile articles; rags	7	17.5
61	Articles of apparel and clothing accessories knitted or crocheted	5	11.94
64	Footwear, gaiters and the like; parts of such articles	2	10.51
30	Pharmaceutical products	2	10.48
84	Nuclear reactors, boilers, machinery and mechanical appliances; parts thereof	8	8.16
62	Articles of apparel and clothing accessories, not knitted or crocheted	7	7.4
68	Articles of stone, plaster, cement, asbestos, mica or similar materials	3	6.88
71	Natural or cultured pearls, precious or semi-precious stones, precious metals, metals clad with precious metal and articles thereof; imitation jewellery; coin	4	6.26
85	Electrical machinery and equipment and parts thereof; sound recorders and reproducers, television image and sound recorders and reproducers, and parts and accessories of such articles	3	6.11
41	Raw hides and skins (other than fur skins) and leather	2	5.39
73	Articles of iron or steel	5	4.53
	Total	68	157.74

Source: The authors.

Figure 5.

Source: The authors.

Figure 5 displays the top 20 commodities showing export UI. It is seen that

Commodity code 260111-iron ore is the most predominant commodity being under-invoiced. It has a FOBCIF ratio of 0.57, indicating that the FOB export value recorded by India is only 57% of the (deflated) imported value recorded by the partners.

The FOBCIF ratios range between 0.01 (HS100620) and 0.99 (HS 710239). Rice, husked (HS100620) has the lowest ratios (0.01), followed by heterocyclic compounds (HS293410) with a ratio of 0.03 and table linens (HS 630251) with a ratio of 0.08.

These unusually low FOBCIF ratios point to the stark differences between the recorded trade values by India and its trading partners.

Export Over-Invoicing Is Also Dominated by Only a Few Commodity Groups

Recalling Table 1, the top 100 of the 2,313 commodities that displayed export OI accounted for 73% of the export OI of $277.33. In total, 48 of these commodities fall under only 13 of the 99 chapters of commodity classification (Table 5). The bulk of the OI is seen in Chapters 27 (mineral fuels), 29 (organic chemicals), 87 (vehicles and accessories) and 73 (articles of iron and steel).

Table 5.

Chapter-wise Distribution of Commodities Showing Export Over-Invoicing

Chapter	Chapter Name	No. of items	Amount $ billion
27	Mineral fuels, mineral oils and products of their distillation; bituminous substances; mineral waxes	2	30.58
29	Organic chemicals	4	15.93
87	Vehicles other than railway or tramway rolling-stock, and parts and accessories thereof	5	12.29
73	Articles of iron or steel	7	11.66
88	Aircraft, spacecraft and parts thereof	2	9.57
71	Natural or cultured pearls, precious or semi-precious stones, precious metals, metals clad with precious metal and articles thereof; imitation jewellery; coin	2	9.07
64	Footwear, gaiters and the like; parts of such articles	3	8.64
38	Miscellaneous chemical products	3	8.28
61	Articles of apparel and clothing accessories knitted or crocheted	6	7.30
68	Articles of stone, plaster, cement, asbestos, mica or similar materials	1	5.95
84	Nuclear reactors, boilers, machinery and mechanical appliances; parts thereof	5	4.87
62	Articles of apparel and clothing accessories, not knitted or crocheted	5	3.39
39	Plastics and articles thereof	3	2.89
	Total	48	130.42

Source: The authors.

The top 20 commodities showing export OI are shown in Figure 6.

Commodity code 271000 (petroleum oils, oils obtained from bituminous minerals, preparations thereof)-petroleum oils and oils obta is the most prominent commodity showing export OI, with a FOBCIF ratio of 1.14. The FOBCIF ratios range between 1.11 (HS 713319-jewellery) and 19.02 (HS 294200-organic compounds). The other commodities that show very high ratios are footwear-HS 640351 (14.32) and granite slabs-HS 680223 (10.2). These FOBCIF ratios are significant and indicate that the export values of these commodities recorded by India are several times more than the import values recorded by the importing trading partners.

Figure 6.

Source: The authors.

The Trade Mis-Invoicing in Top 20 Commodities Has Been Broadly Consistent Over the Years and Across the Trading Partners, with a Few Exceptions

Figure 7 plots the CIFFOB and FOBCIF ratios of the top 20 commodities that show mis-invoicing during the 19 years from 2000 to 2018 (380 possible ratios). The two upper charts in Figure 7 pertain to imported commodities and the two charts in the lower part of Figure 7 pertain to exported commodities. The ratios for the under-invoiced commodities plot predominantly below the constant line 1 (88% for imports and 93% for exports) and the ratios for the over-invoiced commodities plot predominantly over the (88% for both imports and exports) constant line 1, indicating that these commodities show mis-invoicing consistently over the period 2000–2018. A similar position obtains across trading partners showing that the mis-invoicing in a commodity has been fairly consistent across trading partners, though exceptions were noticed in a small percentage of cases, indicating evidence that the same commodity may have been both under-invoiced and over-invoiced in trade with different trade partners.

Figure 7.

Source: The authors.

Majority of the Mis-Invoiced Commodities Fall Under Only a Few Chapters of Commodity Classification

We recall from Table 1 that the top 100 commodities in each group represented only 3%–6% of the number of mis-invoiced commodities in that group, but accounted for 58%–73% of the amount of mis-invoicing. We also recall from Tables 2–5 that a majority of these commodities belonged to only a few commodity groups. For example, majority of the commodities showing import UI fall under 11 (of the 99) chapters of commodity classification (Table 2); commodities showing import OI fall mainly under 10 chapters (Table 3); commodities showing export UI fall mainly under 13 chapters (Table 4); and commodities showing export OI fall mainly under 13 chapters (Table 5). Table 6 summarizes the position. Table 7 shows that more than two-thirds of these 400 commodities (100 from each group of mis-invoicing) fall under 22 chapters of commodity classification.

Table 6.

Summary of Indicated Occurrence of Misinvoicing

		Import UI	Import OI	Export UI	Export OI
1	Total no. of commodities involved in mis-invoicing	3352	1656	2695	2313
2	Total amount of mis-invoicing ($ billion)	686.54	244.77	307.31	277.33
3	Percentage of top 100 mis-invoiced commodities	2.98%	6.04%	3.71%	4.32%
4	Percentage of mis-invoicing accounted for by Top 100 commodities	58%	70%	62%	73%
5	No. of chapters in which the majority of top 100 commodities fall	11	10	13	13

Source: The authors.

Table 7.

Chapters under Which Majority of Mis-Invoiced Commodities Fall

Chapter No.	Chapter Name	Import UI	Import OI	Export UI	Export OI
15	Animal or vegetable fats and oil and their cleavage products; prepared edible fats; animal or vegetable waxes	X
26	Ores, slag and ash			X
27	Mineral fuels, mineral oils and products of their distillation; bituminous substances; mineral waxes	X	X		X
29	Organic chemicals	X	X	X	X
30	Pharmaceutical products			X
38	Miscellaneous chemical products		X		X
39	Plastics and articles thereof	X	X		X
41	Raw hides and skins (other than fur skins) and leather			X
61	Articles of apparel and clothing accessories knitted or crocheted			X	X
62	Articles of apparel and clothing accessories, not knitted or crocheted			X	X
63	Other made-up textile articles; sets; worn clothing and worn textile articles; rags			X
64	Footwear, gaiters and the like; parts of such articles			X	X
68	Articles of stone, plaster, cement, asbestos, mica or similar materials			X	X
71	Natural or cultured pearls, precious or semi-precious stones, precious metals, metals clad with precious metal and articles thereof; imitation jewellery; coin	X		X	X
72	Iron and steel	X	X
73	Articles of iron or steel		X	X	X
84	Nuclear reactors, boilers, machinery and mechanical appliances; parts thereof	X	X	X	X
85	Electrical machinery and equipment and parts thereof; sound recorders and reproducers, television image and sound recorders and reproducers, and parts and accessories of such articles	X	X	X
87	Vehicles other than railway or tramway rolling-stock, and parts and accessories thereof	X			X
88	Aircraft, spacecraft and parts thereof	X	X		X
89	Ships, boats and floating structures		X
999999	Commodities not specified according to kind	X

Source: The authors.

Table 7 shows that while 14 commodities are involved in import mis-invoicing, 18 are involved in export mis-invoicing and some involved in both. There are four groups of commodities—those that display only one, two, three or all four kinds of mis-invoicing:

Commodities falling under Chapter 29 (organic chemicals) and 84 (nuclear reactors, boilers, machinery and mechanical appliances) display all the four kinds of mis-invoicing, viz. UI and OI in both imports and exports;

The groups of commodities that exhibit three kinds of mis-invoicing are as follows: Mineral fuels, oils and products of their distillation; bituminous substances; mineral waxes (Chapter 27); plastics and articles thereof (Chapter 39); pearls, precious or semi-precious stones, precious metals, imitation jewellery, etc. (Chapter 71); articles of iron or steel (Chapter 73); electrical machinery and equipment, etc. (Chapter 85); and aircraft, spacecraft and parts thereof (Chapter 88);

The group of commodities displaying only two types of mis-invoicing include the following: Miscellaneous chemical products (Chapter 38); articles of apparel, etc. (Chapter 61 and 62); footwear, etc. (Chapter 64); articles of stone, plaster, cement, etc. (Chapter 68); iron and steel (Chapter 72) and vehicles (Chapter 87); and

The remaining groups display only one type of mis-invoicing and include animals or vegetable fats and oils (Chapter 15); ores, slag and ash (Chapter 26); pharmaceutical products (Chapter 30); raw hides and skins (Chapter 41); other made-up textile articles (Chapter 63); and ships, boats and floating structures (Chapter 89).

Major Indian Ports Through Which the Mis-Invoiced Commodities Are Traded Can Be Identified

The data regarding the commodities and their ports of trade, which are published by the DGCIS, was not available for the full period 2000–2018 nor was it available HS code wise. DGCIS groups the commodities Principal Commodities (PC) wise, by grouping HS codes into 169 PC groups. Using the data for one year, viz. 2018 and a correspondence table between the PCs and the underlying HS codes available from the DGCIS site, we plotted the major ports through which the mis-invoiced commodities were traded in 2018. The results are shown in Figure 8, which shows that only a few ports handled most of the commodities mis-invoiced in 2018.

Figure 8.

Source: The authors.

Using the Trade Data Analysis a Risk Matrix Can Be Developed for Further Investigation

Given that the majority of TM is displayed by only a few commodity groups, this information can be combined with information about ports of trade, the country of trade and year of trade to develop a risk matrix for follow-up and further investigation. This can provide a robust framework for selecting, from among the millions of trade transactions, high risk transactions for conducting targeted transaction level investigation.

Conceptualizing a General Theory of Mis-Invoicing

The finding that the same commodity may be involved in one, two, three or all four kinds of import and export mis-invoicing perhaps indicates that the individual motive for mis-invoicing may be a more important factor than the general trade policy for that commodity. We discussed in Section 3.5 that while the tariff rates may play a role, they are not the sole determinant, and certainly not a deterrent, for import OI. While the primary motive for import UI is to evade customs duty, the motive for import OI is to illicitly transfer funds overseas. Similarly, while the motive for export UI is illicit outflow, the motive for export OI would be to claim export incentives, in addition to bringing back the funds illicitly transferred out. Regression analysis by Tiwari (2021) for each of the four types of mis-invoicing, computed from the 6-digit commodity level data, shows statistically significant relationships as follows:

Import UI shows positive relationship with export UI and exchange rate and negative relationship with tariff rate.

Import OI shows negative relationship with exchange rate and tariff, and positive relationship with export OI and FDI inflows.

Export UI shows negative relationship with exchange rate, and positive relationship with FDI inflows, import UI and export growth.

Export OI shows positive relationship with import OI.

This leads us to the following question: Can these different inflows (FDI, import UI and export OI) and outflows (import OI and export UI) be woven into a general theory of mis-invoicing and IFF? The Government of India White Paper on Black Money (Ministry of Finance, 2012) refers to round tripping of illicit funds and the role of FDI in round tripping. According to IMF (Damgaard et al., 2018), about 40% of the world FDI is without any underlying economic activity. Abotsi (2018) has reported that FDI net inflows have a positive and significant influence on IFF, while Ndikumana and Sarr (2019) have, based on econometric analysis of the data of 30 African countries from 1970 to 2015, reported positive and robust relationship between FDI and capital outflows. Crylow (2020), based on analysis of Zimbabwe’s trade from 2000 to 2016, reported that export misreporting increased with an increase in net FDI as a percentage of GDP. Biswas et al. (2022) have also reported hidden capital flows through trade from China to United States and through FDI from the United States to China.

Import UI, for which the primary motive is evading customs duty, would need to be financed abroad. It would follow from the findings of Tiwari (2021), discussed above, that this could be achieved through export UI. Similarly, illicit outflows through import OI could return through export OI and FDI inflows, with all their attendant fiscal and monetary benefits. The illicit outflows through export UI, apart from financing the import UI, could also be round tripped through FDI inflows and export OI.

The findings of this article serve a useful purpose in identifying the specific commodities that are majorly involved in illicit outflows and inflows. These comprise commodities with large volumes of trade (e.g., natural resources) or commodities that are high value items (e.g., precious metals and precious stones etc.), hence will be a natural choice as vehicles for illicit flows. Their identification is an important step in addressing the problem of IFF through TM. Tiwari (2021) has demonstrated that aggregation of data from transaction to commodity level and from commodity to trade level, produces successively lower estimates of mis-invoicing as the UI and OI at the lower levels cancel each other out. Since TM takes place at the transaction level, its true extent can be understood only by analysing transaction level data. Since the transaction level data are not in public domain, the estimates made in this article using the 6-digit HS code commodity level data are the next best option. It is in this context that this article is an improvement over other exercises that estimate TM using the country level aggregated import and export data.

Summary and Conclusions

This article demonstrates that while mis-invoicing is seen in a large number of commodities imported and exported by India, only a small percentage of commodities falling under a few chapters of commodity classification account for the bulk of mis-invoicing in imports and exports. The article identifies the top 20 main commodities, at HS-6 code level, involved in UI and OI of import and export and demonstrates that mis-invoicing in these commodities has been fairly consistent over time (2000–2018) and across countries, with a few exceptions in which the same commodity has shown both UI and OI. The article identifies the major ports associated with trade in mis-invoiced commodities and suggests that the details of the identified commodities can be combined with ports of trade to develop a risk matrix for further investigation. The article also suggests a general framework for explaining the TM and the associated illicit financial inflows and outflows in which the FDI inflows play a major part. This is in tune with the existing literature and the authors’ own findings on the determinants of TM. Finally, the article underlines the importance of undertaking TM studies at the disaggregated commodity level data, as the aggregated trade level data would not provide correct estimates due to cancelling out of UI and OI during the process of aggregation (Aggregation Paradox).

The article’s unique contribution to the existing literature lies in identifying the HS-6 codes of the commodities that account for most of the mis-invoicing in India’s exports and imports, the extent of mis-invoicing, the countries with which they were traded, and the associated ports of trade in India. The findings of the article can provide a robust framework for further investigation by the authorities who have the details of transaction level trade. The conceptual framework suggested in this article for determinants of illicit financial outflows and inflows associated with trade and the round tripping of funds involving also the FDI inflows could form a concrete basis for policy formulation to effectively address the issue of TM and the associated IFF.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

References

Abotsi

A. K.

(2018). Influence of governance indicators on illicit financial outflow from developing countries. Contemporary Economics, 12(2), 139–152. https://ssrn.com/abstract=3211093

Beja

(2008). Estimating trade mis-invoicing from China: 2000–2005. China and World Economy, 16(2), 82–92.

Berger

, & Nitsch

(2008). Gotcha! A profile of Smuggling in international trade [CESifo Working Paper No. 2475]. CESifo.

Bhagwati

J. N.

(1964). On the under-invoicing of imports. Bulletin of the Oxford University, Institute of Economics and Statistics, 27(4), 389–397.

Bhagwati

J. N.

(1967). Fiscal policies, faking trade declarations, and the balance of payments. Oxford Bulletin of Economics and Statistics, 29(1), 61–77.

Bhagwati

J. N.

(1981). Alternative theories of illegal trade: Economic consequences and statistical detection. Weltwirtschaftliches Archiv, 117, 409–427.

Bhagwati

J. N.

, Krueger

, & Wibulswasdi

(1974). Capital flight from LDCs: A statistical analysis. In Bhagwati

J. N.

(Ed.), Illegal transactions in international trade: Theory and measurement (pp. 148–154). North-Holland Publications.

Bhagwati

J. N.

, & Hansen

(1973). A theoretical analysis of smuggling. Quarterly Journal of Economics, 87(2), 172–87.

Biswas

A. K.

, & Marjit

(2007). Preferential trade and mis-invoicing: Some analytical implications. International Review of Economics & Finance, 16(1), 130–138.

10.

Biswas

A. K.

, von Hagen

, & Sarkar

(2022). FDI Mismatch, trade mis-reporting, and hidden capital movements: The USA-China case. Journal of International Money and Finance, 120, 102534. 10.1016/j.jimonfin.2021.102534

11.

Buehn

, & Eichler

(2011). Trade misinvoicing: The dark side of world trade. World Economy, 34(8), 1263–1287. Retrieved, 2 June 2020, from https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1467-9701.2011.01375.x

12.

Carrere

, & Grigoriou

(2014). Can mirror data help to capture informal international trade? Policy issues in international trade and commodities [Research Study Series No. 65]. UNCTAD. Retrieved, 2 June 2020, from https://unctad.org/en/PublicationsLibrary/itcdtab65_en.pdf

13.

Carton

, & Slim

(2018). Trade misinvoicing in OECD countries: What can we learn from bilateral trade intensity indices? [MPRA Paper No. 87695]. MPRA. https://mpra.ub.uni-muenchen.de/87695/

14.

Cheung

Y.-W.

, Steinkamp

, & Westermann

(2020). Capital flight to Germany: Two alternative measures. Journal of International Money and Finance, 102, 102095. 10.1016/j.jimonfin.2019.102095

15.

Cobham

, & Jansky

(2020). Estimating illicit financial flows: A critical guide to the data, methodologies, and findings. Oxford University Press.

16.

Crylow

(2020). Trade misreporting: A case of Zimbabwe foreign trade. Journal of Development Economics and Management Research Studies, A Peer Reviewed Open Access International Journal, 6(6), 1–16.

17.

Damgaard

, Elkjaer

, & Johannesen

(2018). Piercing the veil. IMF Finance & Development, 55(2). https://www.imf.org/external/pubs/ft/fandd/2018/06/inside-the-world-of-global-tax-havens-and-offshore-banking/damgaard.htm

18.

de Boyrie

, Nelson

, & Pak

(2007). Capital movement through trade misinvoicing: The case of Africa. Journal of Financial Crime, 14(4), 474–489.

19.

de Boyrie

, Pak

, & Zdanowicz

(2005). Estimating the magnitude of capital flight due to abnormal pricing in international trade: The Russia-USA case. Accounting Forum, 29(3), 249–270.

20.

Fisman

, & Wei

S.-J.

(2004). Tax rates and tax evasion: Evidence from ‘missing imports’ in China. Journal of Political Economy, 112(2), 471–496.

21.

Gara

, Giammatteo

, & Tosti

(2018). Magic mirror in my hand … How trade mirror statistics can help us detect illegal financial flows. The World Economy, 42(11), 3120–3147. 10.1111/twec.12840

22.

Global Financial Integrity. (2019). India: Potential revenue losses associated with trade misinvoicing. Retrieved, 4 April 2020, from https://gfintegrity.org/wp-content/uploads/2019/06/GFI-India-2019.pdf

23.

Hong

, & Pak

(2017). Estimating trade misinvoicing from bilateral trade statistics: The devil is in the details. The International Trade Journal, 31(1), 3–28.

24.

IMF. (1993). A guide to direction of trade statistics. Retrieved, 15 March 2021, from 10.5089/9781451948431.071

25.

Jha

, & Truong

D. N.

(2014). Trade misinvoicing and macroeconomic outcomes in India [ASARC Working Papers]. The Australian National University, Australia South Asia Research Centre. 10.2139/ssrn.2422098

26.

Kar

(2010). The drivers and dynamics of illicit financial flows from India: 1948–2008. Global Financial Integrity. https://www.gfintegrity.org/wp-content/uploads/2014/02/GFI_India.pdf

27.

Kar

, & Spanjers

(2014). Illicit financial flows from developing countries: 2003–2012. Global Financial Integrity. https://www.gfintegrity.org/wp-content/uploads/2014/12/Illicit-Financial-Flows-from-Developing-Countries-2003-2012.pdf

28.

Kellenberg

, & Levinson

(2016). Misreporting trade: Tariff evasion, corruption, and auditing standards [Working Paper No. 22593]. National Bureau of Economic Research. https://www.nber.org/papers/w22593

29.

Marini

, Dipplesman

, & Stanger

(2018). New estimates for direction of trade statistics [IMF Working Papers, WP/18/16, 37 p.]. IMF. https://ssrn.com/abstract=3124401

30.

Marjit

, Biswas

A. K.

, & Beladi

(2008). Minimize regulations to regulate—Extending the Lucas critique. Economic Modelling, Elsevier, 25(4), 623–627.

31.

Marjit

, Dasgupta

, & Mitra

(2000). Currency devaluation and exports: Separating actual from statistical. Economic & Political Weekly, 35(18), 1553–1558.

32.

McDonald

D. C.

(1985). Trade data discrepancies and the incentive to smuggle: An empirical analysis. International Monetary Fund Staff Papers, 32(4), 668–692.

33.

Ministry of Finance. (2012). Black Money. White Paper. Department of Revenue, Central Board of Direct Taxes. Retrieved, 21 December 2021, from https://dor.gov.in/sites/default/files/FinalBlackMoney.pdf

34.

Mishra

, Subramanian

, & Topalova

(2007, March). Policies, enforcement, and customs evasion: Evidence from India [IMF Working Paper]. IMF. 10.5089/9781451866247.001

35.

Morgenstern

(1950). On the accuracy of economic observations: Foreign trade statistics. In Bhagwati

& Chipman

(Eds), International economics: Theory and measurement (Ch. 7). North Holland Publishing Company.

36.

Morgenstern

(1965). On the accuracy of economic observations (2nd ed.). Princeton University Press.

37.

Ndikumana

, & Sarr

(2019). Capital flight, foreign direct investment and natural resources in Africa [Working Paper No. 2019–12]. Department of Economics, University of Massachusetts.

38.

Nitsch

(2016). Trillion-dollar estimate: Illicit financial flows from developing countries [Darmstadt Discussion Papers in Economics 227]. Department of Law and Economics, Technische Universität Darmstadt, p. 15. https://www.econstor.eu/handle/10419/141281

39.

Qureshi

, & Mahmood

(2016). The magnitude of trade misinvoicing and resulting revenue loss in Pakistan. The Lahore Journal of Economics, 21(2), 1–30. 10.35536/lje.2016.v21.i2.a1.

40.

Tiwari

(2021). An empirical investigation into financial crimes and illicit financial flows in India [Unpublished doctoral thesis]. Department of Business Administration, Aligarh Muslim University.

41.

UN Comtrade. (2014). Commodity trade statistics database. http://comtrade.un.org/

42.

UNCTAD. (2016). Trade misinvoicing in primary commodities in developing countries: The cases of Chile, Côte d’Ivoire, Nigeria, South Africa and Zambia (UNCTAD/SUC/2016/2, 1st version). Retrieved, 16 December 2016, from https://unctad.org/webflyer/trade-misinvoicing-primary-commodities-developing-countries-cases-chile-cote-divoire

43.

UNECA. (2015). Illicit financial flows, report of the high-level panel on illicit financial flows from Africa (p. 122). Conference of Ministers of Finance, Planning and Economic Development, AU/ECA. https://repository.uneca.org/handle/10855/22695

44.

UNODC & OECD. (2016). Coherent policies for combatting illicit financial flows [Issue Brief Series]. https://www.un.org/esa/ffd/wp-content/uploads/2016/01/Coherent-policies-for-combatting-Illicit-Financial-Flows_UNODC-OECD_IATF-Issue-Brief.pdf

45.

World Customs Organisation. (n.d.). Study report, 2018: Illicit financial flows via trade mis-invoicing. http://www.wcoomd.org/-/media/wco/public/global/pdf/media/newsroom/reports/2018/wco-study-report-on-iffs_tm.pdf?la=fr

46.

Yeats

A. J.

(1978). On the accuracy of partner country trade statistics. Oxford Bulletin of Economics and Statistics, 40(4), 341–361.

47.

Zdanowicz

(2007). Trade-based money laundering and terrorist financing. Review of Law and Economics, 5(2), 854–878.