Profit Shifting of Multinational Corporations Worldwide Javier Garcia-Bernardo, Petr Jansk´ y Institute of Economic Studies, Faculty of Social Sciences, Charles University, Prague, Czechia The European Tax Observatory, the European Commission’s Joint Research Centre and DG TAXUD, 28 September, 2021
53
Embed
Profit Shifting of Multinational Corporations Worldwide
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Profit Shifting of MultinationalCorporations Worldwide
Javier Garcia-Bernardo, Petr Jansky
Institute of Economic Studies, Faculty of Social Sciences, CharlesUniversity, Prague, Czechia
The European Tax Observatory, the European Commission’s JointResearch Centre and DG TAXUD, 28 September, 2021
Introduction Data Methodology Results: OECD data Conclusion
The effects of profit shifting of multinational corporations(MNCs)
Uneven level playing fieldLower government revenuesGlobalisation perceived as inequitable
2
Introduction Data Methodology Results: OECD data Conclusion
Overview
Data: Country-by-country reporting (CBCR) by MNCs for manycountriesMethodology: A logarithmic function to model the extremelynon-linear relationship between profits and tax rates
1 Scale2 Tax Havens3 Headquarters4 Low-income countries
3
Introduction Data Methodology Results: OECD data Conclusion
Overview
Data: Country-by-country reporting (CBCR) by MNCs for manycountriesMethodology: A logarithmic function to model the extremelynon-linear relationship between profits and tax rates
1 Scale2 Tax Havens3 Headquarters4 Low-income countries
3
Introduction Data Methodology Results: OECD data Conclusion
Contributions to the existing literature (and policy debates)
Methodology: Hines and Rice (1994), Dowd et al. (2017)Data: Clausing (2020), Garcia-Bernardo, Jansky, and Tørsløv(2021), Fuest, Hugger, et al. (forthcoming), Garcia-Bernardo,Jansky, and Zucman (2021)
1 Scale: Crivelli et al. (2016), Alvarez-Martınez et al. (2021), Tørsløvet al. (2020), Bilicka (2019), Dharmapala and Riedel (2013)
2 Tax havens: Zucman (2015), Guvenen et al. (2021)3 Headquarters: Dischinger et al. (2014), Wright and Zucman (2018)4 Low-income countries: Fuest, Hebous, et al. (2011), Jansky and
Palansky (2019), Johannesen et al. (2020)
4
Introduction Data Methodology Results: OECD data Conclusion
The country-by-country reporting data
Aggregated large MNCs’ profits and taxes in around 190 countriesProfit-making affiliates for effective tax rates (ETRs) and bothprofit- and loss-making affiliates for real operations of MNCsThe 2017 US CBCR dataThe 2016 OECD CBCR data with data imputations to furtherimprove coverageThe data are a major step forward, albeit imperfect
5
Introduction Data Methodology Results: OECD data Conclusion
Country availability
Austra
liaAustria
Belgium
Bermuda
Canada
Chile
China
Denmark
Finland
France
Indonesia
India
Ireland
Italy
Japan
South Korea
Luxembourg
Mexico
Netherlands
Norway
Poland
Singapore
Slovenia
Sweden
United States
South Africa
Reporting country
NetherlandsBermuda
Puerto RicoLuxembourg
IrelandCayman Islands
Hong KongSwitzerland
British Virgin IslandsSingapore
United Kingdom
Partner country
6
Introduction Data Methodology Results: OECD data Conclusion
Methodology
Tax semi-elasticity model: linear, quadratic and logarithmic(Also: reallocation of the shifted profit and misalignment model)
7
Introduction Data Methodology Results: OECD data Conclusion
Tax semi-elasticity
The most common model (Hines and Rice, 1994)log (πi )︸ ︷︷ ︸
Profits booked
= β0 + β1 log (Ki )︸ ︷︷ ︸Capital
+β2 log (Li )︸ ︷︷ ︸Labor
+ β3(τi )︸ ︷︷ ︸Tax rate
+ βχχ︸︷︷︸Controls
+ε,
For simplicitylog (πi )︸ ︷︷ ︸
Profits booked
∝ β3(τi )︸ ︷︷ ︸Tax rate
Improvement (Dowd et al., 2017; Hines and Rice, 1994)log (πi )︸ ︷︷ ︸
Profits booked
∝ β3(τi )︸ ︷︷ ︸Tax rate
+ β4(τi )2︸ ︷︷ ︸Tax rate squared
Empirical observation: The model still does not fit the data very well
8
Introduction Data Methodology Results: OECD data Conclusion
Tax semi-elasticity
The most common model (Hines and Rice, 1994)log (πi )︸ ︷︷ ︸
Profits booked
= β0 + β1 log (Ki )︸ ︷︷ ︸Capital
+β2 log (Li )︸ ︷︷ ︸Labor
+ β3(τi )︸ ︷︷ ︸Tax rate
+ βχχ︸︷︷︸Controls
+ε,
For simplicitylog (πi )︸ ︷︷ ︸
Profits booked
∝ β3(τi )︸ ︷︷ ︸Tax rate
Improvement (Dowd et al., 2017; Hines and Rice, 1994)log (πi )︸ ︷︷ ︸
Profits booked
∝ β3(τi )︸ ︷︷ ︸Tax rate
+ β4(τi )2︸ ︷︷ ︸Tax rate squared
Empirical observation: The model still does not fit the data very well
8
Introduction Data Methodology Results: OECD data Conclusion
Tax semi-elasticity
The most common model (Hines and Rice, 1994)log (πi )︸ ︷︷ ︸
Profits booked
= β0 + β1 log (Ki )︸ ︷︷ ︸Capital
+β2 log (Li )︸ ︷︷ ︸Labor
+ β3(τi )︸ ︷︷ ︸Tax rate
+ βχχ︸︷︷︸Controls
+ε,
For simplicitylog (πi )︸ ︷︷ ︸
Profits booked
∝ β3(τi )︸ ︷︷ ︸Tax rate
Improvement (Dowd et al., 2017; Hines and Rice, 1994)log (πi )︸ ︷︷ ︸
Profits booked
∝ β3(τi )︸ ︷︷ ︸Tax rate
+ β4(τi )2︸ ︷︷ ︸Tax rate squared
Empirical observation: The model still does not fit the data very well
8
Introduction Data Methodology Results: OECD data Conclusion
9
Introduction Data Methodology Results: OECD data Conclusion
Our model: Logarithmic semi-elasticity
log (πi )︸ ︷︷ ︸Profits booked
∝ β3(τi )︸ ︷︷ ︸Tax rate
+ β4 log (t + τi )︸ ︷︷ ︸Logarithmic tax rate
10
Introduction Data Methodology Results: OECD data Conclusion
Results for ETR 0.1% (Jersey)
0 5 10 15 20 25 30ETR
0
50
100
150
200
250
300In
crea
se in
pro
fits
(1 =
ETR
25%
)
Incrase: 295.1 times
Incrase: 12.4 timesIncrase: 2.7 times
LogarithmicQuadraticLinear
11
Introduction Data Methodology Results: OECD data Conclusion
The scale of estimated revenue losses (billion USD)
Study Profitshifting
Revenueloss
Data(type)
Individualcoun-tries
Countries(num-ber)
Year(data)
Cobham and Jansky (2018) - 90 Revenue Yes 102 2013IMF’s Crivelli et al. (2016) - 123 Revenue No 173 2013Jansky and Palansky (2019) 420 125 FDI Yes 79 2016IMF (2014) - 180 Revenue Yes 46 2012UNCTAD’s Bolwijn et al. (2018) 330-450 200 FDI No 72 2012Tørsløv et al. (2020) 616-646 230 FDI Yes 48 2015OECD’s Johansson et al. (2017) - 100-240 Orbis No 46 2010Clausing (2016) 1076 279 FDI Yes 25 2012This paper 965-994 186-307 CBCR Yes 192 2016
12
Introduction Data Methodology Results: OECD data Conclusion
Introduction Data Methodology Results: OECD data Conclusion
Tax revenue loss as a percentage of total revenue
10 5 0 5 10Tax Revenue Loss (% Total Tax Revenue)
Misalignment
Low income
Lower middle income
Upper middle income
High incomeLosses Gains
-5.0%
-5.3%
10 5 0 5 10Tax Revenue Loss (% Total Tax Revenue)
Logarithmic model
Losses Gains
-3.0%
10 5 0 5 10Tax Revenue Loss (% Total Tax Revenue)
Misalignment
AfricaAsia
EuropeLatin America
Caribean/American isl.Oceania
Northern AmericaLosses Gains
-6.8%
-4.3%9.6%
10 5 0 5 10Tax Revenue Loss (% Total Tax Revenue)
Logarithmic model
Losses Gains
5.4%
14
Introduction Data Methodology Results: OECD data Conclusion
Concluding remarks
Bigger than previously estimatedLow effective tax ratesLow-income countries more hardly hitUS multinational corporations are special
15
Introduction Data Methodology Results: OECD data Conclusion
Implications for a global corporate tax reform
Postponements costly for low-income countries in particularUnanimous support unlikely if only because of the major playersThe importance of tax havens with low effective tax ratesThe importance of the global minimum tax rate
16
Introduction Data Methodology Results: OECD data Conclusion
Profit Shifting of MultinationalCorporations Worldwide
Javier Garcia-Bernardo, Petr Jansky
Institute of Economic Studies, Faculty of Social Sciences, CharlesUniversity, Prague, Czechia
The European Tax Observatory, the European Commission’s JointResearch Centre and DG TAXUD, 28 September, 2021
References Appendix
References I
Alvarez-Martınez, M. T., Barrios, S., d’Andria, D., Gesualdo, M.,Nicodeme, G., and Pycroft, J. (2021). “How Large Is the CorporateTax Base Erosion and Profit Shifting? A General EquilibriumApproach”. Economic Systems Research.
Bilicka, K. A. (Aug. 2019). “Comparing UK Tax Returns of ForeignMultinationals to Matched Domestic Firms”. en. American EconomicReview, 109(8).
Bolwijn, R., Casella, B., and Rigo, D. (2018). “Establishing the Baseline:Estimating the Fiscal Contribution of Multinational Enterprises”.Transnational Corporations, 25(3).
Clausing, K. (2016). “The Effect of Profit Shifting on the Corporate TaxBase in the United States and Beyond”. Available at SSRN 2685442.
– (Nov. 2020). “Five Lessons on Profit Shifting from the US Country byCountry Data”. Tax Notes Internationa and Tax Notes Federal.
1
References Appendix
References IICobham, A. and Jansky, P. (2018). “Global Distribution of Revenue Loss
from Corporate Tax Avoidance: Re-Estimation and Country Results”.en. Journal of International Development, 30(2).
Crivelli, E., de Mooij, R., and Keen, M. (2016). “Base Erosion, ProfitShifting and Developing Countries”. FinanzArchiv: Public FinanceAnalysis, 72(3).
Dharmapala, D. and Riedel, N. (2013). “Earnings Shocks andTax-Motivated Income-Shifting: Evidence from EuropeanMultinationals”. Journal of Public Economics, 97.
Dischinger, M., Knoll, B., and Riedel, N. (Apr. 2014). “The Role ofHeadquarters in Multinational Profit Shifting Strategies”. en.International Tax and Public Finance, 21(2).
Dowd, T., Landefeld, P., and Moore, A. (Apr. 2017). “Profit Shifting ofU.S. Multinationals”. en. Journal of Public Economics, 148.
2
References Appendix
References IIIFuest, C., Hebous, S., and Riedel, N. (2011). “International Debt
Shifting and Multinational Firms in Developing Economies”.Economics Letters, 113(2).
Fuest, C., Hugger, F., and Neumeier, F. (forthcoming). “Corporate ProfitShifting and the Role of Tax Havens: Evidence from German CbCReporting Data”. CESifo Working Paper.
Garcia-Bernardo, J., Jansky, P., and Tørsløv, T. (2021). “MultinationalCorporations and Tax Havens: Evidence from Country-by-CountryReporting”. International Tax and Public Finance.
Garcia-Bernardo, J., Jansky, P., and Zucman, G. (2021). “Did the TaxCuts and Jobs Act Reduce Profit Shifting by US MultinationalCompanies?”
3
References Appendix
References IVGuvenen, F., Mataloni Raymond J, J., Rassier, D. G., and Ruhl, K. J.
(2021). Offshore Profit Shifting and Aggregate Measurement: Balanceof Payments, Foreign Investment, Productivity, and the Labor Share.Working Paper 23324. National Bureau of Economic Research.
Hines, J. R. and Rice, E. M. (1994). “Fiscal Paradise: Foreign Tax Havensand American Business”. The Quarterly Journal of Economics, 109(1).
Jansky, P. and Palansky, M. (2019). “Estimating the Scale of ProfitShifting and Tax Revenue Losses Related to Foreign DirectInvestment”. International Tax and Public Finance, 26(5).
Johannesen, N., Tørsløv, T., and Wier, L. (Oct. 2020). “Are LessDeveloped Countries More Exposed to Multinational Tax Avoidance?Method and Evidence from Micro-Data”. The World Bank EconomicReview, 34(3).
4
References Appendix
References VJohansson, A., Skeie, O. B., Sorbe, S., and Menon, C. (2017). “Tax
Planning by Multinational Firms: Firm-Level Evidence from aCross-Country Database”. OECD Economics Department WorkingPapers, 2017(1355).
Tørsløv, T., Wier, L., and Zucman, G. (2020). “The Missing Profits ofNations”. National Bureau of Economic Research Working Paper,2018, revised April 2020(24071).
Wright, T. and Zucman, G. (2018). “The Exorbitant Tax Privilege”.National Bureau of Economic Research Working Paper, 24983.
Zucman, G. (2015). The Hidden Wealth of Nations: The Scourge of TaxHavens. Chicago, IL: University of Chicago Press.
5
References Appendix
Robustness checks and sensitivity analyses (1)
1 A variety of methodological approaches, semi-elasticity andmisalignment
2 The robustness of the 25 per cent ETR threshold3 A comparison of our results to those of Tørsløv et al. (2020)4 A comparison the tax revenue loss with a variety of benchmarks5 Limiting the sample to those countries that report information on at
least eight offshore centres6 The sensitivity of our results to the offset in the logarithmic model7 A comparison of the logarithmic specification with other
specifications that can accommodate extreme non-linearities,including 1/(τ + ETR)1, 1/(τ + ETR)2, 1/(τ + ETR)3 andcoth(τ + ETR))
6
References Appendix
Robustness checks and sensitivity analyses (1)
1 A variety of methodological approaches, semi-elasticity andmisalignment
2 The robustness of the 25 per cent ETR threshold
3 A comparison of our results to those of Tørsløv et al. (2020)4 A comparison the tax revenue loss with a variety of benchmarks5 Limiting the sample to those countries that report information on at
least eight offshore centres6 The sensitivity of our results to the offset in the logarithmic model7 A comparison of the logarithmic specification with other
specifications that can accommodate extreme non-linearities,including 1/(τ + ETR)1, 1/(τ + ETR)2, 1/(τ + ETR)3 andcoth(τ + ETR))
6
References Appendix
Robustness checks and sensitivity analyses (1)
1 A variety of methodological approaches, semi-elasticity andmisalignment
2 The robustness of the 25 per cent ETR threshold3 A comparison of our results to those of Tørsløv et al. (2020)
4 A comparison the tax revenue loss with a variety of benchmarks5 Limiting the sample to those countries that report information on at
least eight offshore centres6 The sensitivity of our results to the offset in the logarithmic model7 A comparison of the logarithmic specification with other
specifications that can accommodate extreme non-linearities,including 1/(τ + ETR)1, 1/(τ + ETR)2, 1/(τ + ETR)3 andcoth(τ + ETR))
6
References Appendix
Robustness checks and sensitivity analyses (1)
1 A variety of methodological approaches, semi-elasticity andmisalignment
2 The robustness of the 25 per cent ETR threshold3 A comparison of our results to those of Tørsløv et al. (2020)4 A comparison the tax revenue loss with a variety of benchmarks
5 Limiting the sample to those countries that report information on atleast eight offshore centres
6 The sensitivity of our results to the offset in the logarithmic model7 A comparison of the logarithmic specification with other
specifications that can accommodate extreme non-linearities,including 1/(τ + ETR)1, 1/(τ + ETR)2, 1/(τ + ETR)3 andcoth(τ + ETR))
6
References Appendix
Robustness checks and sensitivity analyses (1)
1 A variety of methodological approaches, semi-elasticity andmisalignment
2 The robustness of the 25 per cent ETR threshold3 A comparison of our results to those of Tørsløv et al. (2020)4 A comparison the tax revenue loss with a variety of benchmarks5 Limiting the sample to those countries that report information on at
least eight offshore centres
6 The sensitivity of our results to the offset in the logarithmic model7 A comparison of the logarithmic specification with other
specifications that can accommodate extreme non-linearities,including 1/(τ + ETR)1, 1/(τ + ETR)2, 1/(τ + ETR)3 andcoth(τ + ETR))
6
References Appendix
Robustness checks and sensitivity analyses (1)
1 A variety of methodological approaches, semi-elasticity andmisalignment
2 The robustness of the 25 per cent ETR threshold3 A comparison of our results to those of Tørsløv et al. (2020)4 A comparison the tax revenue loss with a variety of benchmarks5 Limiting the sample to those countries that report information on at
least eight offshore centres6 The sensitivity of our results to the offset in the logarithmic model
7 A comparison of the logarithmic specification with otherspecifications that can accommodate extreme non-linearities,including 1/(τ + ETR)1, 1/(τ + ETR)2, 1/(τ + ETR)3 andcoth(τ + ETR))
6
References Appendix
Robustness checks and sensitivity analyses (1)
1 A variety of methodological approaches, semi-elasticity andmisalignment
2 The robustness of the 25 per cent ETR threshold3 A comparison of our results to those of Tørsløv et al. (2020)4 A comparison the tax revenue loss with a variety of benchmarks5 Limiting the sample to those countries that report information on at
least eight offshore centres6 The sensitivity of our results to the offset in the logarithmic model7 A comparison of the logarithmic specification with other
specifications that can accommodate extreme non-linearities,including 1/(τ + ETR)1, 1/(τ + ETR)2, 1/(τ + ETR)3 andcoth(τ + ETR))
6
References Appendix
Robustness checks and sensitivity analyses (1)
1 A variety of methodological approaches, semi-elasticity andmisalignment
2 The robustness of the 25 per cent ETR threshold3 A comparison of our results to those of Tørsløv et al. (2020)4 A comparison the tax revenue loss with a variety of benchmarks5 Limiting the sample to those countries that report information on at
least eight offshore centres6 The sensitivity of our results to the offset in the logarithmic model7 A comparison of the logarithmic specification with other
specifications that can accommodate extreme non-linearities,including 1/(τ + ETR)1, 1/(τ + ETR)2, 1/(τ + ETR)3 andcoth(τ + ETR))
6
References Appendix
Robustness checks and sensitivity analyses (2)
8 A different redistribution formula
9 We estimate missing data using 1,000 bootstrapped data samples(using a median, showing confidence intervals)
10 A comparison of the location of employees and revenue according toour missing data model with the information in the original data aswell as GDP
11 A comparison of our missing data imputation method with othermodels
12 A robustness test in which the data of China is not adjusted
7
References Appendix
Robustness checks and sensitivity analyses (2)
8 A different redistribution formula9 We estimate missing data using 1,000 bootstrapped data samples
(using a median, showing confidence intervals)
10 A comparison of the location of employees and revenue according toour missing data model with the information in the original data aswell as GDP
11 A comparison of our missing data imputation method with othermodels
12 A robustness test in which the data of China is not adjusted
7
References Appendix
Robustness checks and sensitivity analyses (2)
8 A different redistribution formula9 We estimate missing data using 1,000 bootstrapped data samples
(using a median, showing confidence intervals)10 A comparison of the location of employees and revenue according to
our missing data model with the information in the original data aswell as GDP
11 A comparison of our missing data imputation method with othermodels
12 A robustness test in which the data of China is not adjusted
7
References Appendix
Robustness checks and sensitivity analyses (2)
8 A different redistribution formula9 We estimate missing data using 1,000 bootstrapped data samples
(using a median, showing confidence intervals)10 A comparison of the location of employees and revenue according to
our missing data model with the information in the original data aswell as GDP
11 A comparison of our missing data imputation method with othermodels
12 A robustness test in which the data of China is not adjusted
7
References Appendix
Robustness checks and sensitivity analyses (2)
8 A different redistribution formula9 We estimate missing data using 1,000 bootstrapped data samples
(using a median, showing confidence intervals)10 A comparison of the location of employees and revenue according to
our missing data model with the information in the original data aswell as GDP
11 A comparison of our missing data imputation method with othermodels
12 A robustness test in which the data of China is not adjusted
7
References Appendix
-100 100 200 300USD bn
0
50
100N
umbe
r of s
imul
atio
ns
185.0B
Median 236.5B
318.0B
USA
-100 100 200 300USD bn
0
20
40
60
-3.5B
Median 101.2B
192.9B
DEU
-100 100 200 300USD bn
0
200
400
82.1BMedian 90.5B
101.4B
FRA
-100 100 200 300USD bn
0
100
200
300
45.6B
Median 64.6B
80.0B
BRA
-100 100 200 300USD bn
0
250
500
750
45.6BMedian 49.9B
55.8B
ITA
-100 100 200 300USD bn
0
200
400
600
Num
ber o
f sim
ulat
ions
34.3BMedian 38.1B
42.9B
MEX
-100 100 200 300USD bn
0
20
40
60
-81.8B
Median 37.5B
123.6B
GBR
-100 100 200 300USD bn
0
200
400
600
27.7BMedian 33.2B
39.9B
IND
-100 100 200 300USD bn
0
200
400
24.4BMedian 30.6B
36.5B
COL
-100 100 200 300USD bn
0
200
400
600
19.2BMedian 21.5B
24.4B
ZAF
-100 100 200 300USD bn
0
200
400
Num
ber o
f sim
ulat
ions
17.8BMedian 27.2B
40.6B
BRB
-100 100 200 300USD bn
0
100
200
300
15.2B
Median 28.6B
50.7B
GIB
-100 100 200 300USD bn
0
200
400
34.6BMedian 40.4B
58.1B
PRI
-100 100 200 300USD bn
0
200
400
36.8BMedian 49.0B
61.8B
CHE
-100 100 200 300USD bn
0
100
200
300
46.3B
Median 58.3B
79.8B
VGB
-100 100 200 300USD bn
0
200
400
Num
ber o
f sim
ulat
ions
52.1BMedian 59.8B
79.2B
BMU
-100 100 200 300USD bn
0
200
400
74.7BMedian 85.6B
99.2B
HKG
-100 100 200 300USD bn
0
50
100
150
42.4B
Median 89.6B
122.1B
CHN
-100 100 200 300USD bn
0
100
200
300
116.9B
Median 133.8B
154.2B
NLD
-100 100 200 300USD bn
0
25
50
75
83.7B
Median 140.4B
276.5B
CYM
Figure: Distribution of the scale of profit shifted estimated by the misalignmentmodel at the country level. The largest origins (top two rows, in blue) anddestinations (bottom two rows, in red) are shown. The variance observed iscreated by the bootstrapping process detailed in Section ??. Non reportingcountries (Germany (DEU), the United Kingdom (GBR), Cayman Islands(CYM) have higher uncertainty than reporting countries such as France (FRA),Italy (ITA) or Bermuda (BMU). The 5% percentile, the median, and the 95%percentile are annotated.