Revista de economía mundial 66, 2024, 67-86

ISSN: 1576-0162

DOI: http://dx.doi.org/10.33776/rem.vi66.7816

IdentIfyIng ProfIt-ShIftIng to tax havenS

USIng CoUntry-By-CoUntry rePort StatIStICS

IdentIfIcacIón del traslado de benefIcIos a paraísos

fIscales utIlIzando las estadístIcas país por país

Ángela Castillo-Murciego

Universidad de Zaragoza

acasmur@unizar.es

Julio López-Laborda

Universidad de Zaragoza

julopez@unizar.es

Recibido: junio 2023; aceptado: enero 2024

aBStraCt

Proﬁt-Shifting is an aggressive tax planning activity that erodes public revenues

worldwide. Using statistics based on recent Country-by-Country Reports, this

article identiﬁes both descriptively and econometrically the existence of such

activity by large Spanish multinationals in EU tax havens. Moreover, the paper

shows that the application of the minimum Effective Tax Rate of 15% to proﬁts

generated in other EU countries by these large multinationals, within the Two-

Pillar solution of the OECD/G20 Inclusive Framework, could increase Spanish

tax revenue by around 2- 17% of annual corporate income tax.

Keywords: Proﬁt-Shifting, Country-by-Country Reports, EU Tax Havens.

reSUmen

El traslado de beneﬁcios es una actividad de planiﬁcación ﬁscal agresiva que

merma los ingresos públicos a nivel mundial. Utilizando las estadísticas basadas

en los recientes informes país por país, este artículo identiﬁca la existencia de

dicha actividad llevada a cabo por las grandes multinacionales españolas en

los paraísos ﬁscales de la UE, de forma tanto descriptiva como econométrica.

Además, el artículo muestra que la aplicación de un tipo impositivo efectivo

mínimo del 15% sobre los beneﬁcios generados en otros países de la UE por estas

grandes multinacionales, en el contexto de la solución basada en los dos pilares

proveniente del marco inclusivo OCDE/G20, podría incrementar los ingresos

ﬁscales de España en torno al 2-17% del Impuesto sobre Sociedades anual.

Palabras clave: Traslado de Beneﬁcios, Declaraciones país por país,

paraísos ﬁscales de la UE.

JEL Classiﬁcation / Clasiﬁcación JEL: F23, H25, H26, H87.

Revista de economía mundial 66, 2024, 67-86

1. IntrodUCtIon

Proﬁt-Shifting (P-S) is an aggressive tax planning activity mainly conducted

by multinational enterprises (MNEs), consisting of using intra-group transactions

to artiﬁcially transfer proﬁts from high-taxed to low-taxed jurisdictions through

strategies that manipulate prices or payment ﬂows at market conditions. For

instance, the transfer pricing strategy manipulates intra-group prices, the thin-

capitalisation strategy manipulates debt interest payments, and the strategic

location of intellectual property manipulates royalty payments.

P-S originates in the mismatch between an evolving economy and a

hundred-year-old international tax system. While this system continues

based on separate accounting1 and the arm’s-length principle for intragroup

transactions,2 MNEs create value globally. Moreover, while the system still

requires the physical presence of a company in a jurisdiction for it to tax the

beneﬁts generated in its territory, the economy is increasingly digital. All of

this facilitates aggressive tax planning activity by MNEs, which take advantage

of the tax loopholes arising from the variety of tax systems worldwide. For

governments, the new characteristics of the economy make it difﬁcult to

identify where MNEs generate proﬁts, but they also try to take advantage of the

mismatch between the new economic setting and the outdated International

Tax System and to attract real economic activity and book proﬁts to their

jurisdiction through tax incentives and low effective tax rates.

P-S activity has become a red-hot topic in recent years because of the

rapidly evolving environment and the weakness of public revenues to ﬁnance

collective needs worldwide. Among other negative consequences, P-S reduces

fairness and equity worldwide, at both the macro- and micro-level. From a

macro perspective, proﬁts are not reported and taxed in the jurisdictions where

they are generated. From a micro perspective, the limited tax burden paid

by some companies might undermine compliance by all taxpayers (Fuest and

Neumeier, 2023). Moreover, since proﬁts are now taxed at lower tax rates, this

aggressive activity causes tax revenue losses all over the world, which otherwise

1 The Separate Accounting System consists of the allocation of taxing rights between the residence

and the source countries by Double Taxation Treaties or subsidiarily, the domestic tax law, so as to

each one can apply its own tax rules.

2 The arm’s-length principle states that the price agreed in an intragroup transaction must be the

same as the price agreed in a similar transaction between two unrelated parties, i.e., the price agreed

at market conditions.

70 Ángela Castillo-Murciego · Julio López-Laborda

could be used to improve or expand the provision of public services. Tørsløv,

Wier and Zucman (2022) estimated that P-S reduced corporate income tax

revenue “by 18% in the European Union, 14% in the US, 5% in other OECD

countries, and 5% in developing countries” (p.26) for 2015. More recently,

Wier and Zucman (2022) have found that the global corporate tax loss derived

from P-S has gone from less than 0.1% of corporate tax revenue in the 1970s

to 10% in 2019. A similar ﬁgure was obtained by Alstadsæter et al. (2023) for

2022, which additionally ﬁnds that US MNEs are responsible for about 40%

of global P-S.

Speciﬁcally, there is evidence that MNEs artiﬁcially shift their proﬁts to

tax haven territories (Castillo-Murciego and López-Laborda, 2023; Tørsløv,

Wier and Zucman, 2022), which are typically identiﬁed by their favourable tax

conditions for foreign investors, among other characteristics (Dharmapala and

Hines, 2009; Dharmapala, 2020 and 2023); i.e., MNEs book proﬁts in tax

havens beyond what can be explained by their real activity in these territories

(Castillo-Murciego and López-Laborda, 2023; Tørsløv, Wier and Zucman,

2022). Speciﬁcally, Alstadsæter et al. (2023) claims that about 35% of proﬁts

generated outside the home country of MNEs were shifted to tax havens in

2022. In this vein, the aim of this article is to identify P-S by large Spanish

MNEs to EU countries considered as tax havens in some (admittedly unofﬁcial)

tax haven lists3 (hereafter, EU tax havens), with the novelty of using recent

statistics based on Country-by-Country Reports (CbCRs). According to Fuest,

Hugger and Neumeier (2022), the EU tax haven countries are Cyprus, Ireland,

Luxembourg, Malta, and The Netherlands.

CbCR is one of the four minimum standards agreed by the members of the

Inclusive Framework within the Base Erosion and Proﬁt Shifting (BEPS) project,

which was launched by OECD and G20 countries in 2013 to tackle P-S. By

2015, 15 actions were developed and one of them was action 13, related

to transfer pricing documentation and CbCRs. These reports try to contribute

reducing P-S by increasing ﬁnancial transparency (Johannesen and Larsen,

2016), and are mandatory for MNEs with consolidated turnover of at least

€750 million.4 The main advantage of the information provided by CbCRs,

compared to the Bureau van Dijk’s Orbis database -the most popular database

about companies’ ﬁnancial information- is the country coverage, especially

regarding tax haven territories.

This case study for Spain represents, then, a ﬁrst step to leveraging

the new data set coming from CbCRs. Because of their country coverage,

CbCRs can be very useful for improving the estimates of P-S, but also the

effectiveness of the actions taken to prevent this activity, particularly the new

Two-Pillar plan agreed by the OECD. As is well known, as of October 2021, a

historical agreement was reached by 135 countries and jurisdictions to reform

international income tax rules. This agreement is based on a Two-Pillar plan

3 See, for instance, Menkhoff and Miethe (2019).

4 See Hoopes, Robinson and Slemrod (2023) for a review of corporate tax disclosure and its effects.

IdentIfyIng ProfIt-ShIftIng to tax havenS USIng CoUntry-By-CoUntry rePort StatIStICS

revISta de eConomía mUndIal 66, 2024, 67-86

(OECD, 2020a and 2020b). Pillar One introduces some innovative elements to

the current International Tax System concerning proﬁt allocation rules to deal

with the digitalisation issue. Speciﬁcally, Pillar One allocates taxing rights to

market jurisdictions regardless of a physical presence in them, affecting some

of the largest MNEs. Pillar Two (also known as the “Global Anti-Base Erosion”

or GloBE rules) seeks to put an end to P-S (and reduce tax competition) by

establishing a minimum Effective Corporate Income Tax rate of 15%, which

large MNEs ultimately have to pay to their country of residence on a Country-

by-Country basis.

Moreover, the results obtained in this paper may be extended to other

countries, especially those also applying a Territorial Tax System for taxing

foreign proﬁts when repatriated,5 which are most of the OECD countries.

These countries allow MNEs to appropriate the tax savings derived from P-S

at the end of the day.

The structure of the rest of the article is as follows. Section 2 brieﬂy

discusses the scant literature based on CbCRs to identify P-S and evaluate

proposals against this aggressive tax planning activity; section 3 describes the

new dataset and its advantages and shows the descriptive and econometric

results regarding the existence of P-S to EU tax haven countries; section 4

examines the consequences in terms of tax revenue from the application of the

Minimum Effective Tax Rate, within the Two-Pillar solution from the OECD/G20

Inclusive Framework; and section 5 concludes.

2. LIteratUre revIew

The literature on P-S is plentiful.6 But only a few recent papers have used

data taken from CbCRs to examine this aggressive tax planning activity and the

possible effects of implementing solutions that seek to prevent it.

Within the international empirical economic literature, Fuest, Hugger and

Neumeier (2022) was the ﬁrst paper to use individual information of German

MNEs’ operations aggregated at tax jurisdiction level for the period 2016-

2017. With a sample of 333 German MNEs and their operations worldwide, it

found a tax revenue loss of €1.6 billion per year derived from P-S to tax havens,

mainly European ones. To reach this result, it ﬁrst estimates a regression

model based on multinational and country level characteristics to examine the

relationship between the global distribution of proﬁts, and corporate income

taxes and tax havens. Then, based on these estimates, it calculates the amount

of shifted proﬁts and the corresponding tax revenue loss.

5 The Territorial Tax System consists of exempting multinationals from the residence tax when proﬁts

are repatriated to the home country and then allowing them to take advantage of the lower tax they

paid abroad. This is the opposite to the practice of countries applying a Worldwide Tax System, which

tax foreign proﬁts when repatriated after giving a tax credit for taxes paid abroad, usually with an

upper ceiling.

6 See Beer, De Mooij and Liu (2020) for a recent review.

72 Ángela Castillo-Murciego · Julio López-Laborda

Bratta, Santomartino and Acciari (2021) uses ﬁrm-level CbCRs of

worldwide MNEs having at least one subsidiary in Italy for 2017. It

discovered the existence of nonlinear responses to taxation and found that

linear speciﬁcations underestimate P-S in countries with corporate tax rate

differentials very different from the average, while they over-estimate P-S in

countries with corporate tax rate differentials similar to the average. It also

found a global tax revenue loss of €245 billion in 2017, the distribution

of artiﬁcially shifted proﬁts being highly concentrated in a few countries:

investment hubs. Based on these results, it concludes that policies establishing

a minimum level of taxation may be very effective in reducing P-S.

Using macro data based on CbCRs for 2017, García-Bernardo, Janský and

Tørsløv (2021), identiﬁes the most important tax havens for US MNEs, i.e.,

those countries with low effective tax rates and high proﬁts not aligned with

economic activity. These comprise Bermuda, Ireland, and The Netherlands.

Also based on aggregate CbCR data from the OECD for year 2017, and

using a new methodology that takes into account the non-linear relationship

between the location of proﬁts and tax rates, García-Bernardo and Janský

(2023) ﬁnds that 70-83% of artiﬁcially shifted proﬁt is deviated to countries

with an effective tax rate (ETR) below 10%.

The revenue effects of the application of the consensus solution involving

Pillar One and Pillar Two have been estimated by various authors and

institutions, such as the OECD (2020c) and the IMF (2022). Using CbCRs’ OECD

statistics for year 2017, Barake et al. (2021a) have estimated an increase of

18% of corporate income tax revenue for Spain derived from the application

of a 15% minimum corporate effective tax rate. This result corresponds to

the most optimistic scenario, in which carve-outs7 for economic substance

envisaged in the OECD agreement are not applied and it is calculated by

computing the “tax deﬁcit”. The tax deﬁcit refers to the difference between

what a company currently pays in corporate income taxes and what it would

have to pay if it were subject to the minimum Effective Corporate Income Tax

rate in each country where it operates, i.e., on a Country-by-Country basis.8

3. data and emPIrICaL anaLySeS

To identify P-S to EU tax havens we use two kinds of analyses: descriptive

and econometric. Both are based on Spanish Country-by-Country statistics

provided by the Spanish Tax Agency for 2018 to 2020. Spanish CbCRs are

required since 2016, but they are not available to the public. The most

disaggregated information disclosed by the Spanish Ministry of Finance is

7 The carve-out for economic substance is one of the potential loopholes that have been added

to the agreement and can greatly weaken the estimated potential additional corporate income tax

revenues (Barake et al., 2021b; Alstadsæter et al., 2023).

8 Similarly, the Spanish Tax Agency has calculated the maximum tax revenue for Spain from the

application of Pillar Two at a minimum effective tax rate of 15% at 2,300 million euros in 2018, i.e.,

0.2% of GDP and 9% of corporate tax revenue (Comité de Personas Expertas, 2022).

IdentIfyIng ProfIt-ShIftIng to tax havenS USIng CoUntry-By-CoUntry rePort StatIStICS

revISta de eConomía mUndIal 66, 2024, 67-86

country-level information on the global activities of the Spanish qualifying

MNEs for 2018 to 2020. Moreover, this information is limited to EU members

(for other countries, information is available at continent level), which greatly

restricts our analysis. Also, García-Bernardo, Janský and Tørsløv (2021) uses

Country-by-Country statistics based on macro data.

Our database thus comprises EU country-level ﬁnancial information on

the activities of Spanish MNEs that are required to submit the CbCR (“Model

231”), i.e., those with consolidated turnover of at least €750 million. They

constitute an annual number of around 120 Spanish qualifying MNEs.9

A summary of the Country-by-Country statistics published by the Spanish

Ministry of Finance is in Table A1 in the Appendices.10 Speciﬁcally, there

is information about the total values of the following variables: number of

entities of Spanish MNEs, unrelated revenues, related party revenues, total

revenues, proﬁt before income tax, corporate income tax (either accrued

or paid)11, positive corporate income tax (either accrued or paid)12, stated

capital, accumulated earnings, tangible ﬁxed assets other than cash, and

number of employees.

Statistics on CbCRs by jurisdiction at the headquarter level for Spain and

most of the world countries are also published for the period 2016 to 2018

by the OECD (only 2017 and 2018 for Spain). For this alternative data source

on CbC statistics, the available ﬁnancial variables are the same as those shown

on the Spanish webpage except for the variable positive corporate income

tax (either accrued or paid), which is a variable created and calculated by the

Spanish Tax Agency. Additionally, the OECD webpage provides data on the

kind of business activity of the MNEs in the corresponding tax jurisdiction.

As claimed above, the main strength of CbCRs is the country coverage and

the information about revenues from intra-ﬁrm transactions they provide. In

contrast, the popular Bureau van Dijk’s Orbis database presents a great deal

of missing data for tax haven jurisdictions because it relies on information in

public business registries, which do not exist in some countries or present

incomplete information. According to Tørsløv, Wier and Zucman (2022),

only 17% of the global proﬁts of MNEs could be traced in Bureau van Dijk’s

Orbis database in 2012. Fuest, Hugger and Neumeier (2022) found that the

Orbis database only covers 35% of German MNEs’ subsidiaries located in

non-havens, assuming data based on CbCRs is complete. This ﬁgure is 26%

for subsidiaries located in European tax havens (information on activities in

Cyprus, Liechtenstein, Malta, and Switzerland is completely missing) and zero

9 Speciﬁcally, 122 in 2018, 124 in 2019, and 126 in 2020.

10 https://sede.agenciatributaria.gob.es/Sede/estadisticas/estadisticas-impuesto/declaracion-pais-

pais-multinacionales-matriz-espanola.html.

11 Accrued corporate income tax encompasses only operations belonging to the corresponding

period; paid corporate income tax is the income paid on a cash basis.

12 Positive corporate income tax is a variable calculated by the Spanish Tax Administration from

corporate income tax (either accrued or paid, as appropriate). The amount of the tax is added from

those jurisdictions whose net corporate tax has a positive value.

74 Ángela Castillo-Murciego · Julio López-Laborda

for subsidiaries located in tax havens outside Europe. Bratta, Santomartino

and Acciari (2021) compared CbCR with Orbis data for the largest MNEs

having at least one subsidiary in Italy and veriﬁed the low coverage of US

MNEs, of non-US MNEs located in the United States, and of MNEs operating

in low-tax and investment-hub countries. García-Bernardo, Janský and Tørsløv

(2021) inferred a higher quality of country-level information about MNCs’

tax payments and proﬁts, as well as better country coverage, comparing

the new data based on Internal Revenue Service CbCRs with a set of data

sources with information about US MNEs, both public (BEA’s and Eurostat’s

foreign afﬁliate statistics and controlled foreign corporation data provided by

the IRS) and private (company balance sheet information, Bureau van Dijk’s

Orbis, and Standard & Poor’s Compustat).

Additionally, in a recent study, Aliprandi and von Zedlitz (2023) found that

in general terms the ﬁgures shown in CbCRs present a high level of reliability

in comparison to respective consolidated ﬁnancial data, though there is room

for improvement to make these reports completely useful for policymakers

and the general public.

3.1. deSCrIPtIve reSULtS

Figures 1 to 4 show the results for the descriptive analysis, which seem to

be consistent with P-S to EU tax haven countries.

In average terms, Figure 1 and Figure 2 show a highly disproportionate

ratio of Pre-Tax Proﬁts to indicators of real economic activity for tax haven

countries in comparison to non-havens, these ratios being especially high for

years 2018 and 2019.13 Thus, it seems that reported proﬁts are not aligned

with real economic activity in these territories, which could be an indication

of the presence of P-S.

Figure 3 shows a higher percentage of related-party revenues in tax

havens than in non-havens. This is also in line with aggressive tax planning

activity, since P-S strategies use intra-group transactions to manipulate

market conditions and report higher (lower) revenues and lower (higher)

expenses in low (high) tax countries. Lastly, Figure 4 brings to light a lower

corporate effective tax rate in tax havens than in non-havens, which is in line

with P-S toward the former territories. There is an exception for the effective

tax rate calculated from the paid tax for 2020. This peculiarity, as well as

those observed in the previous ﬁgures (albeit minor) make the year 2020 an

outlier, which could be related to the coronavirus pandemic.

13 For German MNEs the ratio of returns to tangible assets is approximately 39% in European tax

havens and proﬁt per employee is about 134 thousand euros (Fuest, Hugger and Neumeier, 2022).

IdentIfyIng ProfIt-ShIftIng to tax havenS USIng CoUntry-By-CoUntry rePort StatIStICS

revISta de eConomía mUndIal 66, 2024, 67-86

fIgUre 1. average Pre-tax ProfItS to tangIBLe fIxed aSSetS (%) of SPanISh mneS In tax haven

and non-haven eU CoUntrIeS (exCLUdIng SPaIn), 2018-2020

Source: By the authors, data from the Spanish Ministry of Finance.

fIgUre 2. average Pre-tax ProfItS to nUmBer of emPLoyeeS of SPanISh mneS In tax haven and

non-haven eU CoUntrIeS (exCLUdIng SPaIn), 2018-2020

Source: By the authors, data from the Spanish Ministry of Finance.

fIgUre 3. average reLated Party revenUeS to totaL revenUeS (%) of SPanISh mneS In tax haven

and non-haven eU CoUntrIeS (exCLUdIng SPaIn), 2018-2020

Source: By the authors, data from the Spanish Ministry of Finance.

Note: Related revenues are estimated by the ratio between revenues with related entities and total

revenues.

76 Ángela Castillo-Murciego · Julio López-Laborda

3.2. eConometrIC reSULtS

Two kinds of empirical approaches, direct and indirect, have been used in

the empirical literature to identify the existence of P-S. The direct approaches

consist of identifying particular P-S strategies, while indirect methods are

based on the expected results of P-S. Our econometric analysis to identify

P-S follows the Hines and Rice (1994) indirect approach, which bases the

existence of P-S on ﬁnding a negative relationship between corporate income

taxes and reported proﬁts in a territory, after controlling for the real economic

activity in the corresponding territory. Using this approach and the Bureau van

Dijk’s Amadeus database, Castillo-Murciego and López-Laborda (2017) found

evidence consistent with P-S between Spain and other countries carried out by

Spanish MNEs, and EU and OECD foreign MNEs owning Spanish afﬁliates during

2005-2014.

Particularly, in this article we estimate the following econometric speciﬁcation,

which uses the same dependent and control variables as Fuest, Hugger and

Neumeier (2022), but at macro level:

(1)

The dependent variable (Yit) is alternatively deﬁned as the log of Pre-Tax

Proﬁts or the log of Related Revenues of qualifying Spanish MNEs in EU country

i and year t. The main independent variable is the binary variable tax haven (thit),

which takes the value of one when the country i is a tax haven, and the value of

zero otherwise. We consider that this variable is more indicative of the incentive

fIgUre 4. average effeCtIve CorPorate InCome tax rate (%) of SPanISh mneS In tax haven and

non-haven eU CoUntrIeS (exCLUdIng SPaIn), 2018-2020

Source: By the authors, data from the Spanish Ministry of Finance.

Notes: (1) Effective Corporate Income Tax Rate is estimated by the ratio between positive Corporate

Income Tax (either, accrued or paid) and Pre-Tax Proﬁts. (2) Estonia is excluded from these estimates

for 2018 and Slovenia for 2020 regarding estimates based on paid taxes, because their ETR are

outliers (i.e., the ETR is higher than 1,000% in both cases).

IdentIfyIng ProfIt-ShIftIng to tax havenS USIng CoUntry-By-CoUntry rePort StatIStICS

revISta de eConomía mUndIal 66, 2024, 67-86

to shift proﬁts than indicators of the tax rate, because it focuses on a group

of territories that, in addition to favourable tax conditions for foreign investors,

have other positive features, such as strong governance and institutions; i.e., they

have robust political and legal systems and low levels of corruption (Dharmapala,

2023).14 We expect to ﬁnd a positive coefﬁcient for this variable regarding

the two dependent variables. As to Pre-Tax Proﬁts, tax havens offer the most

favourable taxes for foreign investors and other non-tax incentives to artiﬁcially

shift proﬁts to, and then, to report disproportionately high pre-tax proﬁts there.

With regard to Related Revenues, MNEs use strategies such as the location of

Intellectual Property or ﬁnancial services in tax havens to carry out P-S, which

lead to disproportionately high related revenues in these territories.

The control variables are ﬁrm-level and country-level characteristics.

Regarding ﬁrm-level variables, while Fuest, Hugger and Neumeier (2022) had

individual data for MNEs, though aggregated at country level, we have macro

data. Thus, our ﬁrm-level characteristics refer to the characteristics of all ﬁrms

operating in the corresponding country. These ﬁrm-level characteristics are the

log of the number of employees of Spanish qualifying MNEs in country i and

year t (lg(employeesit)), the log of tangible ﬁxed assets (lg(assetit)), and the log of

unrelated revenues (lg(unrelatedrevit)). Country-level characteristics are the log of

the GDP per capita of country i and year t (lg(gdppcit)), the log of the population

(lg(popit)), and the Corruption Perceptions Index (corruptionit).15

The descriptive statistics of the model variables are in Table A2 in the

Appendices. Table 1 displays the estimated coefﬁcients for the model variables

using a Pooled Linear Regression Model.16

Results seem to be consistent with P-S to EU tax haven countries. The

estimated coefﬁcient for the tax haven variable is positive and statistically

signiﬁcant for the two models incorporating alternative dependent variables.

Thus, Pre-Tax Proﬁts and Related Revenues seem to be higher in tax haven

than in non-haven EU countries, after controlling for ﬁrm and country-level

characteristics inﬂuencing them. Speciﬁcally, reported proﬁts by large Spanish

MNEs in EU tax havens exceed reported proﬁts in EU non-tax havens by a

notably high percentage: 154%. Similarly, intra-ﬁrm revenues in EU tax havens

exceed intra-ﬁrm revenues in EU non-tax havens by 177%.

Regarding control variables, Table 1 shows that the only variables that

seem to be statistically signiﬁcant are unrelated revenues, as an indicator of

14 The main independent variable in the baseline speciﬁcation of Fuest, Hugger and Neumeier

(2022) is the statutory and, alternatively, the effective tax rate. Subsequently, they incorporated a

tax haven dummy variable into their model, and found that the tax semi-elasticity lost a great deal of

its importance in favour of the tax haven variable.

15 GDP per capita based on purchasing power parity, and population, are taken from the World

Development Indicators database: https://databank.worldbank.org/source/world-development-

indicators. Corruption Perceptions Index is taken from Transparency International: https://

www.transparency.org/en/cpi/2021?gclid=CjwKCAjw6MKXBhA5EiwANWLODHA9HDPs-

7WychBgLQoTcaIVH78EW4QU_by8teBpNxdiWM_OpBym7RoCaPYQAvD_BwE.

16 Results are not estimated using country Fixed Effects because in such a case it would not be

possible to identify the effect of the tax haven variable.

78 Ángela Castillo-Murciego · Julio López-Laborda

the importance of the country i as a market country; corruption, for the model

using pre-tax proﬁts as dependent variable; and the number of employees, for

the model using related revenues, all of them having a positive impact on the

dependent variable. These results could further reinforce the importance of

P-S activity in explaining reported proﬁts in a country, in addition to the real

economic and other reasons that affect the overall distribution of pre-tax proﬁts

and related revenues of MNEs. However, we should be cautious in interpreting

the results due to the small number of observations in the sample.

Additionally, we conducted two robustness tests, the results of which are

provided in the Appendices. Table A3 shows the estimated results including

Belgium as an EU tax haven country (Tørsløv, Wier and Zucman, 2022).

Results are similar, but the magnitude of the effect of the tax haven variable on

dependent variables is slightly smaller. Table A4 illustrates the estimates for the

period 2017-2020, where Country-by-Country statistics for year 2017 comes

from the OECD database. Fundamental results are substantially maintained.

4. SImULatIon of the effeCtS of oeCd PILLar two ImPLementatIon

Using the same methodology as Barake et al. (2021a), and the statistics

about CbCRs from the Spanish Ministry of Finance for 2018 to 2020, Table 2

shows the maximum increase in corporate income tax revenue for Spain from

the establishment of the tax rate of 15% to proﬁts generated in EU countries by

taBLe 1. reSULtS

Variable (1) Log (Pre-Tax Proﬁts) (2) Log (Related Revenues)

thit

1.5473***

(0.4275)

1.7751***

(0.4317)

lg(employeesit)-0.3103

(0.2665)

0.5796*

(0.3166)

lg(assetit)0.1523

(0.2045)

0.1249

(0.2487)

lg(unrelatedrevit)1.0158***

(0.2740)

0.0724

(0.3745)

lg(gdppcit)-0.8995

(0.5937)

1.0717

(0.6605)

lg(popit)0.2004

(0.1870)

0.3245

(0.2204)

corruptionit

0.0252**

(0.0097)

0.0078

(0.0137)

N 79 77

R20.9118 0.9003

Source: By the authors.

Notes: (1) ***, ** and * denote statistical signiﬁcance at 1%, 5% and 10% levels, respectively.

Robust Standard Errors are in brackets.

IdentIfyIng ProfIt-ShIftIng to tax havenS USIng CoUntry-By-CoUntry rePort StatIStICS

revISta de eConomía mUndIal 66, 2024, 67-86

large MNEs on a Country-by-Country basis.17 In total, results seem to indicate

that the Spanish government could have collected annual additional revenues

of around 2 to 17% of positive corporate taxes (depending on the year and on

the accrued or paid taxes)18 coming from the EU countries.

17 Council Directive (EU) 2022/2523, of 14 December 2022 (https://eur-lex.europa.eu/legal-

content/EN/TXT/PDF/?uri=CELEX:32022L2523) implementing GloBE in the European Union

establishes that the minimum level of taxation will be applied for both large MNEs and large domestic

groups starting in 2024. Considering also the effect that the application of Pillar Two could have on

subsidiaries of Spanish-parent MNEs that pay corporate income tax in Spain, the Spanish Tax Agency

has estimated a maximum potential collection of 0.45% of GDP, based on 2018 data (Comité de

Personas Expertas, 2022).

18 For EU countries (Spain excluded), positive accrued corporate taxes were 3,074,590 thousand

euros for 2018, 2,811,273 thousand euros for 2019 and 1,233,037 thousand euros for 2020; and

positive paid corporate taxes were 2,959,632 thousand euros for 2018, 2,617,523 thousand euros

for 2019 and 1,637,136 thousand euros for 2020.

taBLe 2: maxImUm addItIonaL revenUeS for SPaIn from the aPPLICatIon of the mInImUm effeCtIve

CorPorate InCome tax: 2018-2020

Country

Additional CIT 2018, in

million € and as a % of EU

CIT (in brackets)

Additional CIT 2019, in

million € and as a % of EU

CIT (in brackets)

Additional CIT 2020, in

million € and as a % of EU

CIT (in brackets)

Accrued CIT Paid CIT Accrued CIT Paid CIT Accrued CIT Paid CIT

Belgium      6.9944

(0.4272)

Bulgaria 0.5210

(0.0169)

0.4010

(0.0135)

1.3467

(0.0479)

1.4557

(0.0556)

0.7096

(0.0575)

0.3586

(0.0219)

Croatia   0.1343

(0.0048) 

Cyprus  1.9296

(0.0048) 

Den-

mark

11.0042

(0.3579)

7.3242

(0.2475) 

Estonia   0.0098

(0.0003)

0.0088

(0.0003)

0.0913

(0.0074)

0.0983

(0.0060)

Ger-

many 131.6176

(5.0283)  

Hungary 2.1969

(0.0715)

1.6459

(0.0556)

2.6775

(0.0952)

2.6635

(0.1018)

1.1621

(0.0942) 

Ireland 79.1202

(2.5734)

28.9422

(0.9779)

82.5594

(2.9367)

79.8844

(3.0519)

7.9947

(0.6484)

1.6227

(0.0991)

Latvia 0.1552

(0.0050)

0.0632

(0.0021) 0.1556

(0.0059)  

Lithu-

ania

0.8983

(0.0292)

0.7103

(0.0240)

0.2456

(0.0087)

1.2296

(0.0470)

0.5057

(0.0410)

0.5567

(0.0340)

(Continúa)

80 Ángela Castillo-Murciego · Julio López-Laborda

In each one of the three years considered, and irrespective of taking

accrued or paid taxes for the calculation of the effective tax rate, Ireland is

always in the top three as a source of additional tax revenue. Table 2 also

shows that there are some EU countries not identiﬁed as tax havens that might

also generate extra revenues for Spain from the application of the minimum

tax rate. Indeed, only 7 out of 27 EU countries (including the United Kingdom

until 2020) apply effective corporate tax rates higher than 15% in any of the

three years. According to these results, it appears that low effective corporate

tax rates are not only related to tax havens, i.e., additional tax revenues could

come from either tax havens or non-haven countries. However, while estimated

additional revenues from non-havens come, in general terms, from real proﬁts,

those from tax havens are mainly based on book proﬁts, i.e., tax havens are

territories with low effective tax rates and high proﬁts not aligned with real

economic activity (García-Bernardo, Janský and Tørsløv, 2021).

Country

Additional CIT 2018, in

million € and as a % of EU

CIT (in brackets)

Additional CIT 2019, in

million € and as a % of EU

CIT (in brackets)

Additional CIT 2020, in

million € and as a % of EU

CIT (in brackets)

Luxem-

bourg

27.6598

(0.8996)

43.0848

(1.4557) 28.5679

(1.0914) 18.9101

(1.1551)

Malta 15.9709

(0.5194)

22.4789

(0.7595)

13.8234

(0.4917) 10.0550

(0.8155) 

Nether-

lands 151.0129

(5.3717)

44.9039

(1.7155)

18.0435

(1.4633) 

Portugal 34.1497

(1.1107)

150.5797

(5.0878) 137.4518

(5.2512)  

Roma-

nia 1.4226

(0.0506)

5.2016

(0.1987)

2.1031

(0.1706) 

Slovakia 2.2643

(0.0736) 2.8954

(0.1030)

1.6544

(0.0632)

2.2135

(0.1795) 

Slovenia 0.5486

(0.0178)

0.2126

(0.0072)

0.1239

(0.0044)

0.0409

(0.0016)  

Sweden 19.0404

(0.6193) 

United

King-

dom

13.4040

(0.4529) 

TOTAL 193.5292

(6.2945)

270.7762

(9.1490)

256.2513

(9.1151)

434.8354

(16.6125)

42.8783

(3.4775)

28.5407

(1.7433)

Source: By the authors. Notes: (1) Additional (Accrued or Paid) Corporate Income Tax is estimated

by making the product between the Tax Deﬁcit (Accrued or Paid) and Pre-tax proﬁts. The Tax Deﬁcit

(Accrued or Paid) is the difference between 15% and (Accrued or Paid) ETRs (see note 1 of Figure

4). (2) Estonia is excluded from these estimates for 2018 and Slovenia for 2020 regarding estimates

based on paid taxes, because their ETR are outliers (i.e., the ETR is higher than 1,000% in both

cases).

IdentIfyIng ProfIt-ShIftIng to tax havenS USIng CoUntry-By-CoUntry rePort StatIStICS

revISta de eConomía mUndIal 66, 2024, 67-86

5. ConCLUdIng remarkS

This article has obtained descriptive and econometric results consistent

with Proﬁt-Shifting (P-S) by large Spanish MNEs to EU tax-haven countries,

using the recent statistics based on Country-by-Country Reports (CbCRs).

Pre-tax proﬁts and related revenues reported in EU tax havens seem to

be disproportionately high in comparison to economic activity in these

jurisdictions. While the use of statistics on CbCRs is the main strength of the

present article and represents a ﬁrst step in the exploitation of this new data

on Spanish MNEs, it also constitutes its main limitation due to the scarcity of

information available to the public on these reports to date.

The main advantage of these statistics is the country coverage. While the

Bureau van Dijk’s Orbis database is missing a great deal of unconsolidated

data regarding some jurisdictions, CbCRs contain country-level information

on MNCs’ tax payments and ﬁnances. However, the available statistics are

currently very limited for Spain. The most disaggregated statistics published

by the Spanish Ministry of Finance on these reports are total ﬁnancial values

by jurisdiction, which does not allow us to consider multinational-level

characteristics. Moreover, country-level data is only available for EU countries,

which further limits the number of observations in the sample.

Therefore, a natural extension of this research should be to replicate the

above analyses later, using more years and individual data on the activities of

Spanish MNEs operating worldwide when available. Indeed, Spain and Italy are

the EU countries with the highest number of CbCRs disclosed by multinationals

(Aliprandi et al., 2022; Aliprandi and Borders, 2023), and some of these

Spanish MNEs provide an excellent level of corporate tax transparency (Godar

et al., 2022).19 Additionally, due to the transposition of Directive 2021/2101,20

large Spanish MNEs will have the obligation to publish their CbCR on the MNE

website for the ﬁscal years starting from June 2024.

Based on the 2018-2020 CbCR information, we have also calculated the

additional tax revenue that Spain would have gained if the minimum effective

tax rate of 15% (without carve-outs) established by the GloBE rules had been

applicable. Considering only proﬁts obtained by Spanish MNEs in EU countries

(Spain excluded), it seems that the establishment of this minimum tax rate

could have originated extra tax revenue for the Spanish public treasury of up to

17% of positive corporate taxes annually. Again, this analysis is also a ﬁrst step

regarding the possible consequences of the application of Pillar Two.

19 See https://taxobservatory.shinyapps.io/company_cbcr_data/.

20 Directive (EU)2021/2101 of the European Parliament and of the Council of 24 November

2021 (https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:32021L2101&from=

EN) requires country-by-country reports of multinational groups operating in the EU with a total

consolidated revenue of at least EUR 750 million to be publicly available starting with ﬁscal year

2025. The Directive has been transposed in Spain by Law 28/2022 of 21 December on the promotion

of the start-up ecosystem.

82 Ángela Castillo-Murciego · Julio López-Laborda

An important issue that could affect the ﬁnal result of the application of

Pillar Two on tax revenue is the behavioural response of MNEs. They may, for

instance, reduce investment and employment in some countries because of

the increase in the effective cost of capital, or relocate real activity to low-tax

territories to take advantage of carve-out rules (Suárez Serrato, 2018; IMF,

2022). This and other behavioural responses of MNEs may change predictions

about the tax revenue consequences of the proposal, and are likely to have

other unintended general economic consequences worldwide.

aCknowLedgementS

The authors gratefully acknowledge funding from the Government of

Aragón, research project S23_23R.

referenCeS

Aliprandi, G., and K. Borders. (2023). “Tax Transparency by Multinationals.

Trends in Country-by-Country Reports Public Disclosure.” EU Tax

Observatory Note.

Aliprandi, G., Borders, K., Gabriel, F., and G. von Zedlitz. (2022). “Public

Country-by-Country Reports: A New Database.” EU Tax Observatory Note.

Aliprandi, G., and G. von Zedlitz. (2023). “Benchmarking Country-by-Country

Reports.” EU Tax Observatory Working Paper Nº. 17.

Alstadsæter, A., Godar, S., Nicolaides, P., and G. Zucman. (2023). Global Tax

Evasion Report 2024. EU Tax Observatory.

Barake, M., Chouc, P.E., Neef, T., and G. Zucman. (2021a). “Revenue Effects

of the Global Minimum Tax: Country-by-Country Estimates.”, EU Tax

Observatory Note Nº. 2.

Barake, M., Neef, T., Chouc, P. E., and G. Zucman. (2021b). “Minimizing the

Minimum Tax? The Critical Effect of Substance Carve-Outs.”, EU Tax

Observatory Note.

Beer, S., De Mooij, R., and L. Liu. (2020). “International Corporate Tax

Avoidance: A Review of the Channels, Magnitudes, and Blind Spots.”

Journal of Economic Surveys, 34(3), 660-688.

Bratta, B., Santomartino, S., and P. Acciari. (2021). “Assessing Proﬁt Shifting

Using Country-by-Country Reports: A Non-linear Response to Tax Rate

Differentials.” DF Working Paper n.11, Dipartimento delle Finanze, Roma.

Castillo-Murciego, A., and J. López-Laborda. (2017). “Are Spanish Companies

Involved in Proﬁt Shifting? Consequences in Terms of Tax Revenues.”

Economics: The Open-Access, Open-Assessment E-Journal, 11(1), 1-47.

Castillo-Murciego, A., and J. López-Laborda. (2023). “Tax-induced Investments

in Tax Havens by Spanish Multinationals.” Applied Economics, 55(53),

6250-6263.

Comité de Personas Expertas. (2022). “Libro Blanco sobre la Reforma

Tributaria.” Madrid. https://www.ief.es/docs/investigacion/comiteexpertos/

LibroBlancoReformaTributaria_2022.pdf

IdentIfyIng ProfIt-ShIftIng to tax havenS USIng CoUntry-By-CoUntry rePort StatIStICS

revISta de eConomía mUndIal 66, 2024, 67-86

Dharmapala, D. (2020). “Do Multinational Firms Use Tax

Havens to the Detriment of Other Countries?” CESifo

Working Papers 8275.

Dharmapala, D. (2023). “Overview of the Characteristics of Tax Havens.” CESifo

Working Papers 10411.

Dharmapala, D., and J. R. Hines. (2009). “Which Countries

Become Tax Havens?” Journal of Public Economics, 93 (9-

10), 1058-1068.

Fuest, C., Hugger, F., and F. Neumeier. (2022). “Corporate Proﬁt Shifting and the

Role of Tax Havens: Evidence from German Country-by-Country Reporting

Data.” Journal of Economic Behavior and Organization, 194, 454-477.

Fuest, C., and F. Neumeier. (2023). “Corporate Taxation.” Annual Review of

Economics, 15, 425-450.

García-Bernardo, J., Janský, P., and T. Tørsløv. (2021). “Multinational

Corporations and Tax Havens: Evidence from Country-by-Country

Reporting.” International Tax and Public Finance, 28(6), 1519-1561.

García-Bernardo, J., and P. Janský. (2023). “Proﬁt Shifting of Multinational

Corporations Worldwide”.

Available at SSRN: https://ssrn.com/abstract=4435224.

Godar, S., Aliprandi, G., Faccio, T., and P. Janský. (2022). “The Long Way to

Tax Transparency: Lessons from the Early Publishers of Country-by-Country

Reports.” EU Tax ObservatoryWorking Paper Nº. 10.

Hines, J.R., and E.M. Rice. (1994). “Fiscal Paradise: Foreign Tax Havens and

American Business.” Quarterly Journal of Economics, 109(1), 149-182.

Hoopes, J.L., Robinson, L., and J. Slemrod. (2023). “Corporate Tax Disclosure.”

NBER Working Paper 31467.

IMF. (2022). “Coordinating Taxation Across Borders.” Fiscal Monitor. Fiscal

Policy from Pandemic to War, Chapter 2. Washington, DC: IMF.

Johannesen, N., and D. T. Larsen. (2016). “The Power of Financial Transparency:

An Event Study of Country-by-Country Reporting Standards.” Economics

Letters, 145, 120-122.

Menkhoff, L., and J. Miethe. (2019). “Tax evasion in new disguise? Examining

tax havens’ international bank deposits.” Journal of Public Economics, 176,

53-78.

OECD. (2020a). “Tax Challenges Arising from Digitalisation – Report on Pillar

One Blueprint: Inclusive Framework on BEPS.” OECD/G20 Base Erosion

and Proﬁt Shifting Project, OECD Publishing, Paris, https://doi.org/10.1787/

beba0634-en.

OECD. (2020b). “Tax Challenges Arising from Digitalisation – Report on Pillar

Two Blueprint: Inclusive Framework on BEPS.” OECD/G20 Base Erosion

and Proﬁt Shifting Project, OECD Publishing, Paris, https://doi.org/10.1787/

abb4c3d1-en.

OECD. (2020c). “Tax Challenges Arising from Digitalisation – Economic

Impact Assessment: Inclusive Framework on BEPS.” OECD/G20 Base

84 Ángela Castillo-Murciego · Julio López-Laborda

Erosion and Proﬁt Shifting Project, OECD Publishing, Paris, https://doi.

org/10.1787/0e3cc2d4-en.

Suárez Serrato, J.C. (2018). “Unintended Consequences of Eliminating Tax

Havens.” NBER Working Paper 24850.

Tørsløv, T.R., Wier, L.S., and G. Zucman. (2022). “The Missing Proﬁts of Nations.”

The Review of Economic Studies, forthcoming.

Wier, L.S., and G. Zucman. (2022). “Global Proﬁt Shifting, 1975-2019.” NBER

Working Paper 30673.

aPPendICeS

taBLe a1. CoUntry-By-CoUntry StatIStICS for SPanISh mneS: 2018-2020

Non-detailed data

-Total value of variables by effective tax rate level (accrued

and paid).

-Ratios about tax contribution, proﬁtability, and economic

substance by effective tax rate level (accrued and paid).

-Territorial distribution of multinationals, entities, and income

by effective tax rate level (accrued and paid): Spain, EU coun-

tries (Spain excluded) and non-EU countries.

-Tax contribution of territories by effective tax rate level (ac-

crued and paid): Spain, EU countries (Spain excluded) and

non-EU countries.

-Proﬁts and economic substance of territories by effective

tax rate level (accrued and paid): Spain, EU countries (Spain

excluded) and non-EU countries.

Detailed data by jurisdiction

-Total value of variables by country (UE-28) or continent.

-Ratios related to proﬁtability and economic substance by

country (UE-28) or continent.

-Average of total value of variables by country (UE-28) or

continent.

-Percentage of variables’ total value by country (UE-28) or

continent.

Source: By the authors, data from the Spanish Ministry of Finance.

IdentIfyIng ProfIt-ShIftIng to tax havenS USIng CoUntry-By-CoUntry rePort StatIStICS

revISta de eConomía mUndIal 66, 2024, 67-86

taBLe a2. deSCrIPtIve StatIStICS: 2018-2020

Variable Obs Mean Std. Dev. Min Max

Pre-tax Proﬁts

(thousand €) 80 509,000 1,110,000 4 7,180,000

revenues (thou-

sand €)

80 997,000 1,660,000 -165,000 5,840,000

Employees 80 12,952.05 24,640.68 4 129,401

Assets (thou-

sand €) 80 2,300,000 5,930,000 30 39,400,000

Unrelated

revenues (thou-

sand €)

80 3,820,000 8,040,000 -3,782,000 45,800,000

GDP pc (thou-

sand €) 80 37.2250 15.7565 18.5159 95.5078

Population

(thousand) 80 16,700 22,800 484,630 83,200

Corruption 80 64.3625 14.24047 42 88

Tax haven 80 0.1875 0.3928 0 1

Source: By the authors.

taBLe a3. reSULtS. extended tax haven LIStS

Variable (1) Log (Pre-Tax Proﬁts) (2) Log (Related Revenues)

thit

1.1500***

(0.3251)

1.2414***

(0.3911)

lg(employeesit)-0.4002

(0.2604)

0.4840

(0.3176)

lg(assetit)0.2912

(0.1964)

0.3057

(0.2359)

lg(unrelatedrevit)0.9659***

(0.2827)

-0.0131

(0.3749)

lg(gdppcit)-0.7458

(0.6058)

1.3186**

(0.6608)

lg(popit)0.1304

(0.1766)

0.2346

(0.2211)

corruptionit

0.0222**

(0.0095)

0.0041

(0.0142)

N 79 77

R20.9094 0.8943

Source: By the authors.

Notes: ***, ** and * denote statistical signiﬁcance at 1%, 5% and 10% levels, respectively. Robust

Standard Errors are in brackets.

86 Ángela Castillo-Murciego · Julio López-Laborda

taBLe a4. reSULtS. PerIod 2017-2020

Variable (1) Log (Pre-Tax Proﬁts) (2) Log (Related Revenues)

thit 1.6284***

(0.3331)

1.8595***

(0.3495)

lg(employeesit)-0.1396

(0.2117)

0.2242

(0.2765)

lg(assetit) -0.0063

(0.1218)

0.0414

(0.1798)

lg(unrelatedrevit) 1.0237***

(0.2148)

0.4705

(0.2853)

lg(gdppcit) -0.7759

(0.4784)

0.7324

(0.6077)

lg(popit) 0.1802

(0.1331)

0.3904**

(0.1853)

corruptionit 0.0260***

(0.0086)

0.0040

(0.0116)

N101 99

R20.9152 0.8899

Source: By the authors.

Notes: ***, ** and * denote statistical signiﬁcance at 1%, 5% and 10% levels, respectively. Robust

Standard Errors are in brackets.