Skip to main content
Log in

Multinational corporations and tax havens: evidence from country-by-country reporting

  • Published:
International Tax and Public Finance Aims and scope Submit manuscript

A Correction to this article was published on 16 March 2021

This article has been updated

Abstract

A growing body of economics literature shows that multinational corporations (MNCs) shift their profits to tax havens. We contribute to this evidence by comparing a range of available data sets focusing on US MNCs, including country-by-country reporting data, a full sample of which has been released in December 2019 for the first time. With each of the data sets, we analyse the effective tax rates that US MNCs face in each country and the amount of profits they report. Using country-by-country reporting data, we have been able to establish that lower effective corporate tax rates are associated with higher levels of reported profits when compared with different indicators of real economic activity. This corresponds to the notion that MNCs often shift profits to countries with low effective tax rates—without also shifting substantive economic activity. Consequently, we identify the most important tax havens for US MNCs as countries with both low effective tax rates and high profits misaligned with economic activity.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1

source overlaps. Source: Authors. Note: This is an illustrative diagram only. For representation purposes, the distributions shown are less skewed than the empirical distributions

Fig. 2

Source: Authors on the basis of various data sets. Notes: IRS data on companies with positives profit (CBCR*), IRS data on all companies (CBCR), IRS data on all CFCs with positive profits (CFC*), BEA data (BEA), BEA data without income cost adjustment (BEA 2), Eurostat data and Orbis data on companies with positive profits (Orbis*). The correlation is visualized for the five indicators analysed: a employees, b turnover, c tangible assets, (d) profits, and e taxes accrued

Fig. 3

Source: Authors on the basis of various data sets. Notes: a Correlation between IRS data on companies with positives profits (CBCR*), IRS data on all companies (CBCR), IRS data on all CFCs with positive profits (CFC*), BEA data (BEA), BEA data without income cost adjustment (BEA 2), Orbis data on companies with positive profits (Orbis*), tax returns data from IRS (Dowd et al.) and statutory corporate income tax (CIT) rates. b ETR by region. The tax rate is unweighted for all but Dowd et al. and CIT, since data on profits and taxes by region is unavailable for those data sets

Fig. 4

Source: Authors on the basis of the IRS CBCR, BEA and Orbis data. Notes: a Negative misalignment. b Positive misalignment. The misalignment profits are calculated using the CCCTB formula, where the location of assets, revenue, employees and wages determines the expected location of the profits. The databases used were: IRS data on companies with positive profits (CBCR*), IRS data on all companies (CBCR), IRS data on all CFCs with positive profits (CFC*), BEA data (BEA), BEA data without income cost adjustment (BEA 2), Orbis data on companies with positive profits (Orbis*). The location of the accumulated earnings (divided by 10) is marked in brown (CBCR* (acc. Earn.)

Fig. 5

Source: Authors on the basis of the IRS CBCR data. Notes: a Relationship between the ETR and profit misalignment (measured as profits in excess of expected profits). Note that countries with the highest positive misalignment tend to have lower ETRs. b Relationship between ETR and the probability of positive misalignment. c Relationship between CIT and the probability of positive misalignment for countries with at least $100 million absolute misalignment. Line indicates the modelled logistic regression and shaded area indicates 95% confidence intervals calculated from 1000 bootstrap samples. Countries in b, c have been slightly moved vertically to improve legibility. The confidence intervals for the coefficient of ETR/CIT are b [− 0.67, − 0.13] and c [− 0.19, − 0.006]

Similar content being viewed by others

Change history

Notes

  1. An illuminating case study is Starbucks, which was selected in 2013 by the European Commission as one of the first state aid cases resulting from an advanced pricing agreement; subsequently, in 2019 the General Court of the European Union’s Court of Justice found that Starbucks’ transfer pricing analysis was reasonable based on Dutch law and OECD Transfer Pricing Guidelines in place at the time of the negotiation of the advanced pricing agreement (Byrnes 2019).

  2. Our dependent variable (misalignment) contains extreme values, both positive and negative. We use the cubic root of the misalignment to ensure that the residuals are approximately normally distributed, while allowing both positive and negative values.

References

Download references

Acknowledgements

We are grateful for great comments from Kimberly Clausing, Tim Dowd, Miroslav Palanský, Alexandra Rusu, Caroline Schimanski and Francis Weyzig. Javier Garcia-Bernardo has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement number 638946). Petr Janský acknowledges support from the Czech Science Foundation (P403/18-21011S). To ensure transparency and replicability, and in line with open science practices, our entire database and code can be found here: https://osf.io/ew67b/.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Petr Janský.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

The original online version of this article was revised: The co-author name “Javier Garcia-Bernando” should read “Javier Garcia-Bernardo”.

Appendix

Appendix

List of countries by region

EU28 = Bosnia and Herzegovina (BA), Cyprus (CY), Romania (RO), Croatia (HR), Finland (FI), Luxembourg (LU), Spain (ES), Austria (AT), Latvia (LV), Slovak Rep. (SK), Poland (PL), Italy (IT), Norway (NO), United Kingdom (GB), Netherlands (NL), France (FR), Hungary (HU), Germany (DE), Portugal (PT), Greece (GR), Bulgaria (BG), Belgium (BE), Estonia (EE), Denmark (DK), Malta (MT), Sweden (SE), Slovenia (SI), Ireland (IE), Lithuania (LT).

Africa = Sierra Leone (SL), Uganda (UG), Sudan (SD), Lesotho (LS), Eritrea (ER), South Sudan (SS), Réunion (RE), Togo (TG), Saint Helena, Ascension and Tristan da Cunha (SH), Seychelles (SC), Gabon (GA), Morocco (MA), Egypt, Arab Rep. (EG), Sao Tome and Principe (ST), Mauritania (MR), Ghana (GH), Malawi (MW), Equatorial Guinea (GQ), Madagascar (MG), Djibouti (DJ), Botswana (BW), Angola (AO), Guinea-Bissau (GW), Mauritius (MU), Congo (Dem. Rep.) (CD), Zimbabwe (ZW), Guinea (GN), Cameroon (CM), Burundi (BI), Ethiopia (ET), Niger (NE), Mali (ML), Rwanda (RW), Benin (BJ), Comoros (KM), Cabo Verde (CV), Tanzania, United Republic of (TZ), Senegal (SN), Kenya (KE), Algeria (DZ), Central African Republic (CF), Ivory Coast (CI), Mayotte (YT), Zambia (ZM), The Gambia (GM), Liberia (LR), Somalia (SO), Nigeria (NG), Libya (LY), Mozambique (MZ), Eswatini (SZ), Burkina Faso (BF), South Africa (ZA), Tunisia (TN), Chad (TD), Congo, Rep. (CG).

Southeast_Asia = Malaysia (MY), Singapore (SG), Laos (LA), Cambodia (KH), Timor-Leste (TL), Viet Nam (VN), Philippines (PH), Indonesia (ID), Myanmar (MM), Thailand (TH), Brunei Darussalam (BN).

Northeast_Asia = Russia (RU), Taiwan (TW), China (CN), Japan (JP), Korea, Democratic People's Rep. of (KP), South Korea (KR), Mongolia (MN).

Middle_East = Lebanon (LB), Jordan (JO), Oman (OM), Kuwait (KW), Ukraine (UA), Israel (IL), Iran (IR), Bahrain, Kingdom of (BH), West Bank (PS), Syrian Arab Republic (SY), Turkey (TR), Yemen, Republic of (YE), Iraq (IQ), Qatar (QA), Saudi Arabia (SA).

OFCs = Cyprus (CY), St. Vincent and the Grenadines (VC), Togo (TG), Seychelles (SC), Anguilla (AI), Samoa (WS), Liechtenstein (LI), Belize (BZ), Isle of Man (IM), UK Caribbean (Montserrat (MS), Cayman Islands (KY), Turks and Caicos Islands (TC), Virgin Islands, British (VG)), Mauritius (MU), Panama (PA), St. Kitts and Nevis (KN), Guyana (GY), Cayman Islands (KY), Bermuda (BM), Netherlands Islands, Caribbean (CW), Guernsey (GG), Gibraltar (GI), Marshall Islands, Republic of (MH), Jersey (JE), Barbados (BB), Virgin Islands, British (VG), Liberia (LR), Bahamas, The (BS),

Source: Authors.

See Figs. 6, 7, 8, 9, 10,  11,  12, 13 and  14, Tables 5 and 6.

Fig. 6
figure 6

Source: Authors on the basis of the IRS CBCR data

Comparison between the 2016 and 2017 CBCR data releases.

Fig. 7
figure 7

Source: Authors on the basis of the IRS CBCR data. Notes: The absolute values of taxes paid minus accrued are displayed. The left side (blue) shows negative values of taxes paid minus taxes accrued. The right side (red) shows positive values

Tax accrued versus tax paid.

Fig. 8
figure 8

Source: Authors on the basis of the IRS CBCR data

Effective tax rate using tax accrued versus tax paid.

Fig. 9
figure 9

Source: Authors based on GDP data (x axis) and the Gallup World Poll survey on median income per capita (y axis), collected from World Population Review (2020)

Comparison between wages estimated using GDP per capita and median income per capita.

Fig. 10
figure 10

Source: Authors on the basis of various data sets. Notes: IRS data on companies with positives profit (CBCR*), IRS data on all companies (CBCR), IRS data on all CFCs with positive profits (CFC*), BEA data (BEA), BEA data without income cost adjustment (BEA 2), Eurostat data and Orbis data on companies with positive profits (Orbis*). The correlation is visualized for the five indicators analysed: a employees, b turnover, c tangible assets, d profits, and e taxes accrued. For each of the coloured squares, representing the correlation between CBCR* and other databases, a scatter plot showing the outliers is shown

Correlation between the data sets.

Fig. 11
figure 11

Source: Authors on the basis of various data sets. Notes: aggregated values of IRS data on companies with positives profits (CBCR*) compared with the following data sets: IRS data on all companies (CBCR), IRS data on all CFCs with positive profits (CFC*), BEA data (BEA), BEA data without income cost adjustment (BEA 2), Eurostat data and Orbis data on companies with positive profits (Orbis*) samples. Coverages below one (shaded in green) indicate higher values in the CBCR* data set. The list of countries by region is found in Appendix. Note that the BEA sample is more comparable with CBCR (not CBCR*), since both samples include all companies, including those with negative profits. Orbis uses fixed assets instead of tangible assets, which leads to the overestimation of assets in (a)

Differences between samples by region.

Fig. 12
figure 12

Source: Authors on the basis of various data sets. Notes: a Correlation between IRS data on companies with positives profits (CBCR*), IRS data on all companies (CBCR), IRS data on all CFCs with positive profits (CFC*), BEA data (BEA), BEA data without income cost adjustment (BEA 2), Orbis data on companies with positive profits (Orbis*), tax returns data from IRS (Dowd et al.) and statutory corporate income tax (CIT) rates. b ETR by region. The tax rate is unweighted for all but Dowd et al. and CIT, since data on profits and taxes by region is unavailable for those data sets. cg Correlation between CBCR* and c CIT rates, d CBCR, e BEA, f Orbis*, and g Dowd et al. Countries where the estimations differ by more than 50% are annotated

Estimates of effective tax rates.

Fig. 13
figure 13

Source: Authors on the basis of the IRS CBCR, BEA and Orbis data

Top 10 countries with the largest negative (a) and positive (b) misaligned profits with the US included (in contrast with Fig. 4).

Fig. 14
figure 14

Source: Authors on the basis of the IRS CBCR data

Correlation between ETR and misalignment profits using accumulated earnings.

Table 5 Indicative comparison of published IRS and expected OECD tables.
Table 6 Overview of the available data sources relevant for US MNCs.

.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Garcia-Bernardo, J., Janský, P. & Tørsløv, T. Multinational corporations and tax havens: evidence from country-by-country reporting. Int Tax Public Finance 28, 1519–1561 (2021). https://doi.org/10.1007/s10797-020-09639-w

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10797-020-09639-w

Keywords

JEL Classification

Navigation