Tax Evasion from Cross-Border Fraud: Does Digitalization Make a Difference?1
Author: Emmanouil Kitsios, João Tovar Jalles, and Ms. Genevieve Verdier

Contributor Notes

Author’s E-Mail Address: EKitsios@imf.org; JJalles@iseg.ulisboa.pt; GVerdier@imf.org
Publication Date:
13 Nov 2020
eISBN:
9781513561189
Language:
English
Keywords:
panel data; fiscal policy; carousel fraud; e-government; technology; taxation; WP; tariff revenue; digitalization index; tax authority; transaction data; government policy; trade fraud; tariff rate; revenue gain; exporter characteristic; revenue collection; trade tax
Download PDF (1.5 MB)

How can governments reduce the prevalence of cross-border tax fraud? This paper argues that the use of digital technologies offers an opportunity to reduce fraud and increase government revenue. Using data on intra-EU and world trade transactions, we present evidence that (i) cross-border trade tax fraud is non-trivial and prevalent in many countries; (ii) such fraud can be alleviated by the use of digital technologies at the border; and (iii) potential revenue gains of digitalization from reducing trade fraud could be substantial. Halving the distance to the digitalization frontier could raise revenues by over 1.5 percent of GDP in low-income developing countries.

Abstract

How can governments reduce the prevalence of cross-border tax fraud? This paper argues that the use of digital technologies offers an opportunity to reduce fraud and increase government revenue. Using data on intra-EU and world trade transactions, we present evidence that (i) cross-border trade tax fraud is non-trivial and prevalent in many countries; (ii) such fraud can be alleviated by the use of digital technologies at the border; and (iii) potential revenue gains of digitalization from reducing trade fraud could be substantial. Halving the distance to the digitalization frontier could raise revenues by over 1.5 percent of GDP in low-income developing countries.

I. Introduction

Examples of tax evasion at the border are not hard to come by. In late 2018, a single individual was found guilty of evading Chinese tariff payments of $54 million on 1.3 million tons of oil products imported to China.2 A World Bank report estimates that well-connected firms in Tunisia evaded $1.2 billion in tariffs between 2002 and 2009 by undervaluing imports (Rijkers, Baghdadi and Raballand, 2015). Such tariff evasion could significantly erode revenue mobilization efforts, particularly in low-income countries. Even in more advanced economies, such evasion is costly. Cross-border trade fraud to evade customs duties, VAT, and excises have important public revenue implications both for developed and developing economies. For example, Missing Trader Intra Community (MTIC) fraud (also known as carousel fraud) exploits the zero-rating of export and deferral of tax on intra-European Union (EU) imports that allows trading across Member State borders to be VAT free (Keen and Smith, 2007). This type of fraud has been estimated to incur between EUR 45 billion to 60 billion tax losses to the EU annually.3

How can governments reduce the prevalence of cross-border fraud? In this paper, we argue that the use of digital technologies offers an opportunity to reduce fraud and increase revenue. Digitalization—the integration in everyday life of digital technologies that facilitate the availability and processing of more reliable, timely, and accurate information—presents an important opportunity for fiscal policy since both expenditure and tax policies depend crucially on information about economic actors. However, relevant and reliable information is not always available or easy to use, constraining the design, implementation, and evaluation of tax and spending policies. Digitalization can improve tax compliance by enhancing operational efficiency and the quality of information in trade transactions. Digital information facilitates the collection of authentic, accurate and complete information about traded goods, enhancing the ability of border agents to collect the appropriate level of trade taxes.

This paper presents evidence that (i) cross-border trade fraud is non-trivial—exporters and importers consistently report different values for goods traded, a crude estimate suggests it could represent up to 6.6 percent of GDP in low-income countries;4 (ii) such tax evasion can be alleviated by the use of digital technologies; and (iii) potential revenue gains could be substantial.5 As in others in the literature (e.g. Kellenberg and Levinson, 2019), we exploit variations in bilateral trade transactions, using data on 28 intra-EU and 85 cross-country trade transactions over the period 2003–16. Tax evasion is measured using discrepancies in trade statistics between origin and destination countries. However, we first focus on intra-EU trade reporting gaps to stress that trade misreporting may occur even within customs unions, where misreporting incentives lie on incentives to evade VAT and excises rather than customs duties. The literature has paid less attention to the implications of trade fraud on VAT revenue even though the latter accounts for a large portion of the estimated VAT gaps in the EU (for an exception see Gradeva, 2014).6 Also, previous studies typically relied on disaggregated industry-by-industry measures of trade misreporting.7 Similar to Kellenberg and Levinson (2019), we examine trade misreporting at the country level to focus on policy-relevant measures other than tariffs that do not differ by sector of activity.

The remaining of the paper is organized as follows. Section II documents how digitalization has changed the conduct of government policy and more specifically how digital tools can improve the collection of information on traded goods at the border. Section III describes the empirical methodology and Section IV presents the results. Section V concludes.

II. Government Digitalization and Tax Compliance

Digital technologies have transformed government operations over the recent years. Vast improvements in the collection, processing, tracking, and dissemination of information have helped enhance public service delivery. Governments are increasingly turning digital, as indicated by the widespread use of national websites and automated financial management systems (Figure 1). Governments in large advanced economies have performed better on average in digital adoption, but many small or developing countries have taken the lead regionally, including Estonia in Europe, Chile in Latin America, Singapore in Asia, and Rwanda and South Africa in sub-Saharan Africa (Figure 2).

Figure 1.
Figure 1.

Number of Countries with Selected Digital Services

Citation: IMF Working Papers 2020, 245; 10.5089/9781513561189.001.A001

Sources: United Nations e-Government Survey 2016 and World Bank 2016.
Figure 2.
Figure 2.

Digital Adoption Index for Governments Across Regions

(latest available year)

Citation: IMF Working Papers 2020, 245; 10.5089/9781513561189.001.A001

Source: World Bank, 2016.Note: Data labels in the figure use International Organization for Standardization (ISO) country abbreviations. The World Bank’s Digital Adoption Index measures the global spread of digital technologies for 171 countries. The government cluster is the average of three indices: core administrative systems, online public services and digital identification. The countries listed are the top- and bottom-ranking countries in each region. AP=Asia and Pacific; CIS=Commonwealth of Independent States; EUR=Europe; LAC=Latin America and the Caribbean; MENA=Middle East and North Africa; NA=North America; SSA=sub-Saharan Africa.

Digitalization allows tax authorities to offer electronic tax filing, pre-populate tax returns, and verify customs and business activity. These could improve tax compliance and enforcement by reconciling payment differences, monitoring real-time revenue collection, performing audits, and using big data to assess taxpayer risks. They also reduce the time burden associated with administration. The World Bank (2016) estimates that electronic filing and payments have on average reduced the time for taxpayers and tax authorities by 25 percent in the five years after the digital system was introduced. At the same time, digitalization can help link information across systems, for example, information from electronic transactions can be linked to value-added taxes (VAT). Some countries have made substantial efforts to digitalize their tax administration. In South Africa, the use of electronic tax submissions, customs declarations, and payments has risen from below 20 percent to close to 100 percent over the past decade, following efforts to modernize and automate administrative processes in tax administration. In Estonia, tax administrators have used big data to identify high-risk and anomalous behavior of taxpayers to improve compliance.8

Digitalization can improve tax compliance by enhancing operational efficiency and the quality of information in trade transactions—particularly within custom unions where border control is lacking. Information is crucial for collecting taxes and duties at the border, in particular information about the product classification, volume and value of goods traded. This information is typically provided by importers and exporters who have an incentive to misreport transactions to evade duties or taxes. To verify information provided by importers and exporters, custom officers need access to third-party information. Direct access to accurate third-party information is facilitated by digitalization. Digital information is more resilient against manipulation than paper documents, facilitating the submission of authentic documents. Blockchain technology could also help secure the authenticity of submitted information, as all transactions are recorded, including the initial submission and all subsequent modifications.9 Digitalization can also help secure the accuracy of reporting at the border. The analysis of historical customs transaction data can enable tax administration to discriminate more effectively between high and low-risk declarations and to mobilize its resources to prevent evasion more efficiently. However, while digitalization can significantly reduce problems related to authenticity and accuracy, obstacles remain when it comes to completeness of information, particularly when the trade payment includes credit where the financial flows linked to the transaction do not sum up to the value of the goods. Nevertheless, the use of digitalization tools could help the revenue mobilization efforts of countries as trade taxes still represent a non-trivial share of revenues, particularly in developing economies where they constitute close to 10 percent of total revenues on average (Figure 3).

Figure 3.
Figure 3.

Taxes on International Trade, 2015

(Percent of total revenue)

Citation: IMF Working Papers 2020, 245; 10.5089/9781513561189.001.A001

Source: IMF, World Economic Outlook database.Note: AEs=advanced economies; EMEs=emerging market economies; LIDCs=low-income developing countries.

III. Methodology and Data

We estimate the following trade gravity model, which builds on the work of Kellenberg and Levinson (2019):

Vxmtm−Vxmtx(Vxmtm+Vxmtx)/2=β1Zxmtm+β2Zxmtx+β3Zxmtσ+at+axm+ϵxmt(1)

where Vxmtm is the annual total trade shipped from exporting country x to importing country m as reported by the importer; Vxmtx is the same value as reported by the exporter. The dependent variable is defined as the difference between these two values and proxies trade misreporting.10 This difference is subsequently normalized by the average reported trade flow to form the so-called “trade gap”.11 In general, the trade gap between two countries tends to increase with the distance between the two trading partners, since the value reported by exporters is free-onboard (FOB) while the value reported by importers includes cost, insurance and freight (CIF). Thus, the set of independent variables considered includes a matrix of bilateral proxies for CIF costs Zxmσ (e.g. distance, common borders and languages as in typical gravity-type models), as well as dummies to capture year specific (at) and country-pair specific fixed-effects (axm).

To assess which underlying factors—including the potential role played by digitalization— determine the size of the trade gap, a gravity model approach is employed. Recognizing that the trade gap could be driven by both importer and exporter characteristics, matrices of observable country characteristics (ZxmtmandZxmtx for importers and exporters, respectively) such as VAT rates and weighted average tariff rates, are included that may be related to incentives to misreport trade flows.12 In addition, typical trade gravity models include variables such as GDP and GDP per capita to proxy for the size and development level respectively, of each partner, while inflation and exchange rates are also included here as they may affect the value of the transacted goods while in transit. Controlling for whether a country participates in regional trade agreements, or whether it is a GATT or WTO member, is useful in proxying for unobserved customs collaboration. Finally, country-pair specific time-invariant characteristics—such as distance between two countries—are absorbed by the country-pair fixed effects axm.

The main regressor of interest is digitalization as proxied by the UN’s Online Service Index. This variable assesses the scope and quality of public sector online services, including online services for tax submission and registration of businesses. The index is normalized between 0 and 1 and it is available since 2003. It is significantly correlated to other digitalization indices available and has broader sample coverage across countries and over time compared to WB’s Digital Adoption Index and WEF’s Government Success in ICT Promotion13 (Table 1).

Table 1.

Pairwise Correlations of Digitalization Indices

article image
Note: *, **, and *** indicate the statistical significance level of 10%, 5%, and 1%, respectively. ICT denotes information and communication technology.

Bilateral trade data are obtained from IMF’s Direction of Trade Statistics (DOTS) which reports the values of imports and exports in U.S. dollars. The macro-variables were obtained from the World Economic Outlook, the World Development Indicators and IMF’s Tax Database. CEPII’s Gravity Dataset was used for trade agreement participation and distance. Governance indicators on the control of corruption, the implementation of the rule of law, and effective governance, were retrieved from the World Governance Indicators (WGI) database. Controlling for such indices prevents confounding the estimate on digitalization with the effect of broader governance factors. The table in Appendix 1 includes information on the variables and data sources used.

IV. Empirical Results

Trade fraud leading to tax evasion can be proxied using discrepancies in trade statistics at the origin and destination countries. Existing studies in this area typically follow the approach suggested by Fisman and Wei (2004), identifying evasion based on a correlation between tax or tariff rates and reporting discrepancies between importers and exporters (see also Javorcik and Narciso, 2008; Mishra, Subramanian, and Topalova, 2008; Ferrantino, Liu, and Wang, 2012; Kellenberg and Levinson, 2019). In practice, the value reported by importers includes CIF and—in principle—should exceed the value reported by exporters that is FOB. When negative, this trade gap—the difference between these two reported values—provides a crude indication of trade fraud, when unexplained by other factors such as valuation changes. The median trade gap ratio across countries are significantly different from zero, ranging between -2.4 percent of GDP for advanced economies (AEs) and -6.6 percent of GDP for low-income developing countries (LIDCs) (Figure 4).

Figure 4.
Figure 4.

Trade Gap Ratios, 2016

(Difference between importer and exporter reported values in percent of GDP)

Citation: IMF Working Papers 2020, 245; 10.5089/9781513561189.001.A001

Sources: IMF, Direction of Trade Statistics (DOTS); IMF, World Economic Outlook; and IMF staff calculations.Note: The chart presents negative trade gaps as indicative proxies of trade misreporting. AEs=advanced economies; EMEs=emerging market economies; LIDCs=low-income developing countries.

To the extent that digitalization reduces trade misreporting, it may help improve revenue collection. Tables 2 and 3 show the results of estimating the gravity equation (1). Table 2 refers to the sample of 28 European Union countries over the period 2003–16. A distinct advantage of using the EU sub-sample is to stress that trade misreporting may occur even within custom unions, where misreporting incentives lie on incentives to evade VAT and excises rather than custom duties.14 Column 1 estimates the gravity equation (1) via OLS, and point estimates suggest a positive, yet statistically insignificant, association between improved digitalization indices and the trade reporting gap, suggesting a lower incidence of trade fraud when governments make progress in digitalization.15

Table 2.

Trade Gap Regressions Using Intra-EU Trade Data

article image
Note: Robust standard errors in parentheses, *, **, *** denote statistical significance at the 10, 5 and 1 percent levels, respectively. Controls include country fixed effects, and year fixed effects. ‘Im.’ refers to importer and ‘Ex.’ refers to exporter. ‘Ex.Tariff rate’ drops out due to perfect collinearity with ‘Im.Tariff rate’. Columns 2 and 3 report the first-stage results from using as IVs the importers’ and exporters’ logarithm of patents over R&D, respectively. Columns 4 and 5 report the second-stage (TSLS-2) results before and after censoring the dependent variable at 1st and 99th percentiles.
Table 3.

Trade Gap Regressions Using All Partners Trade Data

article image
Note: Robust standard errors in parentheses, *, **, *** denote statistical significance at the 10, 5 and 1 percent levels, respectively. Controls include country fixed effects, and year fixed effects. ‘Im.’ refers to importer and ‘Ex.’ refers to exporter. Columns TSLS-1a and TSLS-1b report the first-stage results from using as IVs the importers’ and exporters’ logarithm of patents over R&D, respectively. Columns 4 and 5 report the second-stage (TSLS-2) results before and after censoring the dependent variable at 1st and 99th percentiles.

Columns 2, 3, and 4 report the results from implementing two-stage least squares (TSLS) to address potential problems related to omitted variable bias and reverse causality. Such concerns could arise if, e.g., a higher incidence of import misreporting mobilized public authorities of the importing country to foster digitalization efforts so as to reduce tax evasion. This could bias downward the estimated effect of digitalization, given that the policy decision to improve digitalization is negatively correlated with the trade gap and positively correlated with the digitalization index.

We instrument the digitalization index using a measure of R&D efficiency—the ratio of patents to R&D intensity (R&D expenditure in percent of GDP). This instrumental variable is positively and strongly correlated to the endogenous digitalization index both for importers and exporters (columns 2 and 3). The exclusion restriction relies on the assumption that the trade gap itself is not correlated with differences in the instruments once macro-variables are explicitly controlled for. Table 2 reports the first-stage Kleibergen-Paap F-statistics, which exceed the Stock and Yogo (2005) critical values for weak instrument diagnostics. The coefficient estimate for importer’s digitalization is higher in magnitude than the OLS estimate, pointing to downward bias of the later due to endogeneity. The negative coefficient on the importer’s VAT rate is in line with the assumption that the incentive to underreport imports rises with the VAT rate.

Column 5 shows the second-stage results after censoring the dependent variable symmetrically at the 1st and 99th percentiles to address the concern that outliers could be driving the results. The results in columns 4 and 5 indicate that destination countries with more digitalized governments tend to have a larger reported value of imports relative to the exports the countries of origin are reporting.16 This relationship remains significant after controlling for other key determinants, including tariffs and tax rates, the level of development and governance17 Table 3 broadens the sample to include all trading partners available in the DOTS database. The resulting estimates confirm the previous EU subsample conclusion that importer’s digitalization index is positively associated with the reporting of imports. The estimation includes an index to control for corruption that is found significant in the broader sample.

A back-of-the-envelope calculation of the potential revenue gains from reducing trade fraud exploits specification (1) holding other factors constant and using column 6’s estimated coefficient on the digitalization index. Denote VTotalm=Σx(Vxmm)andVTotalx=Σx(Vxmx). Assuming that the importer’s digitalization advancements increase reported imports VTotalm without affecting VTotalx, one can proxy the potential revenue gain from the corresponding increase in reported imports relative to exports as follows:

Re⁡venueGaint=τrate.Δ(VTotalm−VTotalx)(2)

where τrate refers to the tax rate of interest (i.e., VAT or tariff rate).

Equation (2) can be written in terms of the change in the digitalization index of the importer, Δzm, and its estimated impact βdigitm:

Re⁡venueGaint=τrate⋅12(VTotalm+VTotalx)βdigitm⋅Δzm(3)

Rearranging equation (2) to obtain equation (3) assumes that, except for the digitalization index, the remaining set of determinants and imports in the denominator of the trade gap are held constant. Holding constant imports in the denominator effectively biases our estimate downward, allowing for a conservative estimate of the gains from reaching the digitalization frontier.

Reducing the distance to the digitalization frontier for each importer by 50 percent implies increasing zm by Δzm = 0.5 * (1 – zm), as the maximum value the digitalization index can attain is one. The revenue gains are found by applying the latest country-specific VAT and weighted tariff rates, along with the average trade flow (VTotalm−VTotalx) reported in 2016 to equation (3) and assuming βdigitm.

Halving the distance to the digitalization frontier could raise the median VAT revenue by 1.1 percent of GDP for low-income developing countries, 0.7 percent of GDP for emerging market economies and advanced economies, and 0.4 percent for the EU (Table 4). Similarly, median tariff revenue could increase by 0.4 percent of GDP for low-income developing countries, 0.2 percent of GDP for emerging market economies, and 0.04 percent of GDP for advanced economies. These results are only indicative of potential revenue gains because reducing the distance to the digitalization frontier is likely to require significant fiscal resources and the removal of institutional barriers.

Table 4.

Median Revenue Gains per Country Group from Closing Half the Distance to the Digitalization Frontier, 2016

(Percent of GDP)

article image
Note: Latest available VAT rates were used to compute the revenue gains. EU-28 = European Union group of 28 countries (Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Italy, Ireland, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Poland, Portugal, Romania, Slovak Republic, Slovenia, Sweden, Spain, United Kingdom); VAT = value-added tax.

V. Conclusion

Our paper is the first—to our knowledge—to document a lower incidence of trade fraud when governments enhance information collection and processing through digitalization. The results point to significant potential revenue gains of digitalization from reducing trade fraud.

Future work should focus on further analyzing the transmission mechanism—how improvements in digitalization translate into better enforcement at the border. While the estimates of this paper provide a broad range for the revenue potential from digitalization, they do not provide a granular analysis of what specific digital tools—data matching, access to third-party information, etc.—are the most useful for reducing evasion.

Beyond narrow estimates, the challenges of digitalizing revenue administration should not be underplayed. Digitalization itself can offer new fraud opportunities. Individuals and firms can take advantage of new technology to hide sensitive information and evade taxes. For example, new risks emerged with the digitalization of Estonia’s tax administration: when registering and filing taxes online, fraudsters created large number of ghost entities to generate multiple small credit claims that fell below the threshold for audit (IMF, 2018). Governments must also have adequate administrative and institutional capacity and mobilize resources to take advantage of digital dividends (see IMF 2018 for a discussion of these issues).

Even if digitalization broadens options for governments to better design and implement policies, the viability these policies ultimately depend on political resolve. The challenge is to adopt digital tools to enhance government policies, while mitigating the risks associated with digitalization.

Appendix 1. Data Sources

article image
Note: CEPII = Centre d’Etades Prospectives et d’Informations Internationales; GATT/WTO = General Agreement in Tariffs and Trade/World Trade Organization; ICT = information and communication technology; R&D = research and development; UN = United Nations; VAT = value-added tax; WB = World Bank; WEF = World Economic Forum.

References

  • European Commission, 2017, “Proposal for a Council Directive Amending Directive 2006/112/EC as Regards Harmonising and Simplifying Certain Rules in the Value-Added Tax System and Introducing the Definitive System for the Taxation of Trade between Member States,” SWD (2017) 325 Final (Brussels: European Commission).

    • Search Google Scholar
    • Export Citation

  • Ferrantino, Michael J., Xuepeng Liu, and Zhi Wang, 2012, “Evasion Behaviors of Exporters and Importers: Evidence from the US–China Trade Data Discrepancy,” Journal of International Economics, Vol. 86, No. 1, pp. 141–57.

    • Search Google Scholar
    • Export Citation

  • Fisman, Raymond, and Shang-Jin Wei, 2004, “Tax Rates and Tax Evasion: Evidence from ‘Missing Imports’ in China,” Journal of Political Economy, Vol. 112, No. 2, pp. 471–500.

    • Search Google Scholar
    • Export Citation

  • Gradeva, Katerina, 2014 (Unpublished report) “VAT Fraud in Intra-EU Trade,” (Frankfurt: Goethe University, Frankfurt).

  • International Monetary Fund, 2017, “Capitalizing on Good Times – Chapter 2: Digital Government,” April 2018 Fiscal Monitor (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Javorcik, Beata S., and Gaia Narciso, 2008, “Differentiated Products and Evasion of Import Tariffs.” Journal of International Economics, Vol. 76, No. 2, pp. 208–22.

    • Search Google Scholar
    • Export Citation

  • Jean, Sebastien, and Cristina Mitaritonna, 2010, “Determinants and Pervasiveness of the Evasion of Customs Duties,” AgFoodTrade Working Paper 2010–05 (Paris: CEPII Research Center).

    • Search Google Scholar
    • Export Citation

  • Keen, Michael and Stephen Smith, 2007, “VAT Fraud and Evasion: What Do We Know, and What Can be Done?,” IMF Working Paper 07/31 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Kellenberg, Derek, and Arik Levinson, 2019, “Misreporting Trade: Tariff Evasion, Corruption, and Auditing Standards,” Review of International Economics, Vol. 27, No. 1, pp. 106–29.

    • Search Google Scholar
    • Export Citation

  • Krishna, A., M. Fleming, and S. Assefa, 2017, “Instilling Digital Trust Blockchain and Cognitive Computing for Government,” in Digital Revolutions in Public Finance, ed. by Sanjeev, Gupta, Michael Keen, Alpa Shah, and Geneviève Verdier (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Mishra, Prach, Arvind Subramanian, and Petia Topalova. 2008. “Tariffs, Enforcement, and Customs Evasion: Evidence from India,” Journal of Public Economics, Vol. 92, No. 10-11, pp. 1907–25.

    • Search Google Scholar
    • Export Citation

  • Rijkers, Bob, Leila Baghdadi, and Gaël Raballand, 2015, “Political Connections and Tariff Evasion: Evidence from Tunisia,” World Bank Policy Research Working Paper No. 7336 (Washington: The World Bank Group).

    • Search Google Scholar
    • Export Citation

  • Stock, James H. and Motohiro Yogo, 2005, “Testing for Weak Instruments in Linear IV Regression,” in Identification and Inference for Econometric Models: Essays in Honor of Thomas Rothenberg, ed. by Donald W. K. Andrews and James H. Stock (New York: Cambridge University Press).

    • Search Google Scholar
    • Export Citation

  • World Bank, 2016, “World Development Report 2016: Digital Dividends,” A World Bank Group Flagship Report (Washington: The World Bank Group).

    • Search Google Scholar
    • Export Citation
1

The authors are grateful to Francesca Caselli, Xavier Debrun, Cem Dener, Vitor Gaspar, Yuko Hashimoto, Michael Keen, Hiau Looi Kee, Toni Matsudeira, Janos Nagy, Catherine Pattillo , Adrian Peralta-Alva, Matthew Salomon, Abdelhak Senhadji, Philippe Wingender, and participants of the April 2018 Fiscal Monitor Workshop for useful comments and suggestions. Thanks also go to Kyungla Chae for excellent research assistance. The views expressed in this paper only reflect those of the authors.

4

Note that some of this difference may represent measurement errors.

5

The paper only explores the impact of digitalization in reducing trade fraud, but other efforts can also be effective to address tax fraud more broadly. For example, the usefulness of an effective implementation of anti-money laundering and combating the financing of terrorism measures in line with the Financial Action Task Force standards should not be underestimated in supporting efforts to combat tax evasion, other types of tax crimes and trade fraud. See IMF (2018) for a discussion of these issues related to illicit financial flows.

6

The share of the MTIC fraud in the VAT gap has been estimated to average 24 percent, with the remainder of the VAT gap attributed to losses of revenue arising from other factors such as domestic fraud and evasion (see p.20 in European Commission, 2017).

8

See IMF (2018) for additional country case studies.

9

Blockchain is a list of secure, immutable records or blocks of electronic transactions stored cryptographically. The use of blockchain in customs administration remains limited so far. Dubai Customs is exploring the use of blockchain for the import and re-export process of goods (Krishna, Fleming, and Assefa, 2017).

10

This is a crude measure of misreporting as it may include measurement errors as well as capacity constraints— e.g. greater measurement errors in countries with lower tax administration capacity. To some extent, we control for this possibility by including income levels and other institutional variables on the right-hand side.

11

The trade gap as defined can have a maximum value of 2 and a minimum value of -2. The estimation below is robust to the exclusion of such extreme values.

12

While we control for tax rates, we do not directly control for differences in the broad tax regime which may also affect incentives to misreport.

13

The index has been combined with human capital and telecommunication technology indicators to form alternative composite digitalization indices, such as the United Nation’s e-government index and the World Bank’s Digital Adoption Index.

14

Missing trader fraud is not specific to the EU. However, the European Commission has recognized this problem to be an important one and has incorporated estimates of VAT fraud in its VAT gap analysis.

15

The standard errors reported are robust to allow for different variance across country pairs. Results are robust to clustering standard errors at the country pair level to account for bilateral trade correlation across time.

16

The underreporting of imports can occur both when the gap is positive and when the gap is negative. The main channel at work is that improved digitalization of the importing country is positively correlated with the recording of imports, and therefore with the revenue resulting from imported goods.

17

Results are robust to the inclusion of alternative governance quality indicators, such as the rule of law or government effectiveness indices provided by the World Governance Indicators (WGI) database.

Same Series

Other IMF Content

Other Publishers

Cover IMF Working Papers
Tax Evasion from Cross-Border Fraud: Does Digitalization Make a Difference?
Author: Emmanouil Kitsios, João Tovar Jalles, and Ms. Genevieve Verdier
Volume/Issue:
Volume 2020: Issue 245
Publisher:
International Monetary Fund
ISBN:
9781513561189
ISSN:
1018-5941
Pages:
16
DOI:
https://doi.org/10.5089/9781513561189.001
  • View in gallery

    Number of Countries with Selected Digital Services

  • View in gallery

    Digital Adoption Index for Governments Across Regions

    (latest available year)

  • View in gallery

    Taxes on International Trade, 2015

    (Percent of total revenue)

  • View in gallery

    Trade Gap Ratios, 2016

    (Difference between importer and exporter reported values in percent of GDP)

Number of Countries with Selected Digital Services

Digital Adoption Index for Governments Across Regions

(latest available year)

Taxes on International Trade, 2015

(Percent of total revenue)

Trade Gap Ratios, 2016

(Difference between importer and exporter reported values in percent of GDP)