Journal of Public Economics 96 (2012) 946958

Contents lists available at SciVerse ScienceDirect

Journal of Public Economics

journal homepage: www.elsevier.com/locate/jpube

International taxation and multinational rm location decisions

Salvador Barrios a, Harry Huizinga b,, Luc Laeven c, Gatan Nicodme d
a

European Commission, Joint Research Centre, IPTS

Tilburg University and CEPR
c
International Monetary Fund, Tilburg University and CEPR
d
European Commission, ULB, CEPR and CESifo
b

a r t i c l e

i n f o

Article history:
Received 3 November 2008
Received in revised form 18 June 2012
Accepted 18 June 2012
Available online 14 July 2012
JEL classication:
F23
G32
H25
R38

a b s t r a c t
Using a large international rm-level data set, we examine the separate effects of host and additional parent
country taxation on the location decisions of multinational rms. Both types of taxation are estimated to have
a negative impact on the location of new foreign subsidiaries. The impact of parent country taxation is estimated to be sizeable consistent with its international discriminatory nature. Our results show that international double taxation by the parent country despite the general possibility of deferral of taxation until
income repatriation is instrumental in shaping the structure of multinational enterprise.
2012 Elsevier B.V. All rights reserved.

Keywords:
Corporate taxation
Withholding taxes
Dividends
Location decisions
Foreign direct investment
Multinationals

1. Introduction
With globalization and the progressive removal of barriers to trade,
an increasing number of companies develop international activities. To
access foreign markets, rms face a choice between producing goods at
home for exports and producing abroad. A host of tax and non-tax factors affect the decision whether to relocate production abroad. Among
the non-tax factors are the size of a foreign market, its growth prospects, wage and productivity levels abroad, the foreign regulatory
and legal environment, and distance from the parent country (see

The authors thank Jim Hines and Joel Slemrod (the Editors), two anonymous referees,
Wiji Arulampalam, Peter Finnigan, Andreas Hauer, Vanesa Hernandez Guerrero and
seminar participants at the Banco de Espaa, the European Commission, the Solvay Brussels
School of Economics and Management, ZEW Mannheim, the CESifo area conference on
public sector economics in 2009, and the Centre for Business Taxation of Oxford University
Summer Symposium in 2009 for valuable comments. The ndings, interpretations, and
conclusions expressed in this paper are entirely those of the authors. They should not be
attributed to the European Commission or the International Monetary Fund.
Corresponding author at: Department of Economics, Tilburg University, 5000 LE
Tilburg, Netherlands. Tel.: +31 13 4662623.
E-mail address: huizinga@uvt.nl (H. Huizinga).
0047-2727/$ see front matter 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.jpubeco.2012.06.004

Grg and Greenaway (2004), Barrios et al. (2005), and Mayer and
Ottaviano (2007) for recent reviews). The impact of taxation on foreign direct investment (FDI) has been the subject of a sizeable literature, as reviewed by De Mooij and Ederveen (2006) and Devereux
and Mafni (2007).
Studies of the effect of taxation on FDI location decisions generally
examine host country taxation to the exclusion of possible additional
parent country taxation. The contribution of this paper is to jointly consider the impact of host and additional parent country taxation on multinational rm location decisions. As a rst level of taxation, the host
country may impose corporate income taxation on the income of local
foreign subsidiaries. In addition, the host country could levy a nonresident dividend withholding tax on the subsidiary's earnings at the
time they are repatriated to the parent rm. But taxation need not
stop at the host country level. The parent country can further choose
to levy a corporate income tax on the resident multinational's foreign
source income giving rise to international double taxation. We examine
the independent impact of all three levels of taxation on the location decisions of European multinationals over the period 19992003. Specically, we examine how these three levels of taxation affect in which
country a multinational headquartered in a certain country chooses to
locate a new foreign subsidiary.

S. Barrios et al. / Journal of Public Economics 96 (2012) 946958

947

Table 1
Corporate taxation and double tax relief methods for dividends received in European countries in 2003.

Austria
Belgium
Bulgaria
Croatia
Cyprus
Czech Republic
Denmark
Estonia
Finland
France
Germany
Greece
Hungary
Iceland
Ireland
Italy
Latvia
Lithuania
Luxembourg
Malta
Netherlands
Norway
Poland
Portugal
Romania
Russia
Slovak Republic
Slovenia
Spain
Sweden
Switzerland
Turkey
United Kingdom

Corporate tax rate including local

taxes and surcharges (%)

Treatment of foreign dividends

from treaty countries

Treatment of foreign dividends

from non-treaty countries

24
33.99
23.5
20
15
31
30
26
29
35.43
39.59
35
19.64
18
12.5
38.25
19
15
30.38
35
34.5
28
27
33
25
24
25
25
35
28
21.74
33
30

Exemption
Exemption (up to
Indirect credit
Exemption
Exemption
Indirect credit
Exemption
Indirect credit
Exemption
Exemption (up to
Exemption (up to
Indirect credit
Exemption
Exemption
Indirect credit
Exemption (up to
Exemption
Exemption
Exemption
Indirect credit
Exemption
Indirect credit
Indirect credit
Direct credit
Indirect credit
Direct credit
Indirect credit
Exemption
Exemption
Exemption
Exemption
Indirect credit
Indirect credit

Exemption
Exemption (up
Direct credit
Exemption
Exemption
Deduction
Exemption
Indirect credit
Direct credit
Exemption (up
Exemption (up
Indirect credit
Exemption
Exemption
Indirect credit
Exemption (up
Exemption
Exemption
Exemption
Indirect credit
Exemption
Indirect credit
Direct Credit
Direct credit
Indirect credit
No relief
No relief
Exemption
Indirect credit
Exemption
Exemption
Direct credit
Indirect credit

95%)

95%)
95%)

60%)

to 95%)

to 95%)
to 95%)

to 60%)

Notes: Corporate tax rate denotes the statutory corporate tax rate including local taxes and surcharges. Statutory corporate tax rate in Estonia is 0% on
retained earnings but a distribution tax of 26% is applied on distributed prot. Corporate tax rate of France includes a 3% social surcharge and a special
3.3% surcharge for large companies. Corporate tax rate of Germany includes a solidarity surcharge of 5.5%, an average deductible trade tax of 16.14%, and
an exceptional surcharge of 1.5%. Corporate tax rate of Hungary includes a deductible local business tax. Corporate tax rate of Ireland applies to trading
activities. For non-trading activities, the rate is 25%. Corporate tax rate of Luxembourg includes employment surcharges and local taxes. Corporate tax
rate of Switzerland applies to the canton of Zurich and includes cantonal and local taxes in Zurich. Treatment of foreign dividends refers to double tax
relief convention used by parent country. Foreign subsidiaries are assumed to be fully owned. The indirect credit system applies to foreign dividends
from treaty countries in Poland as long as the holding stake is at least 75% for the past 2 years and there exists a bilateral treaty or the EU
parent-subsidiary directive applies. In Portugal, foreign dividends from treaty countries are exempt from corporate taxes if the EU parent-subsidiary directive applies, but the foreign withholding tax is not creditable. In some countries, dividend income is exempt from taxes up to a certain percentage,
indicated between brackets. Source: International Bureau of Fiscal Documentation.

We have collected new detailed information on how parent country

tax systems interact bilaterally with corporate taxation and nonresident withholding taxation in the host country. Specically, we use
information on whether or not countries tax the income of their multinationals on a worldwide basis, and whether foreign tax credits are provided for non-resident withholding taxes only or also for the underlying
host country corporate tax (as, for instance, in the United States). As an
alternative to worldwide taxes, parent countries may partially or fully
exempt foreign source income from taxation. For example, Germany
exempts 95% of the foreign source income of German multinationals
from taxation.
Data on the structures of European multinationals, and in particular
on new foreign subsidiary locations, are obtained from the Amadeus database. This data set allows us to consider multinational companies
resident in a broad set of countries, each potentially having foreign subsidiaries in many other countries. Thus, unlike earlier work this paper
considers multinational rm location choices in a multilateral setting
of N by N countries. In addition to being an innovative approach, this
multi-country framework is necessary to obtain sufcient variation in
additional parent country corporate taxation (not highly correlated
with host country corporate taxation) to be able to separately estimate
its impact on international location decisions.

Our results suggest that host country and additional parent country corporate income taxation both discourage the location of foreign
subsidiaries in a particular country. In our benchmark estimation, the
estimated negative impact of the two types of taxation as derived
from statutory tax information is about of equal size. Our nding
of a signicant role for the additional parent country tax in foreign
subsidiary taxation is new, and perhaps surprising. Many countries
allow parent country taxes on foreign source income to be deferred
until dividend repatriation, reducing the scope for these taxes to affect location decisions. All the same, we nd that foreign subsidiary
location and the resulting international ownership pattern of subsidiaries is sensitive to additional parent country taxation. Multinational rms act on an apparently signicant incentive to bring about
an international ownership pattern of subsidiaries that is internationally tax-efcient.
We perform several robustness checks to better understand the role
of host and parent country taxation in determining foreign subsidiary location. We nd that location decisions of highly protable foreign subsidiaries are less responsive to both host and parent country taxation
than location decisions of less protable foreign subsidiaries, perhaps because high protability reects location- or owner-specic rents that
would be destroyed by an alternative location or national owner. Further,

948

S. Barrios et al. / Journal of Public Economics 96 (2012) 946958

Table 2
Bilateral dividend withholding tax rates in Europe in 2003.
Subsidiary country
Austria
Belgium
Bulgaria
Croatia
Cyprus
Czech Rep.
Denmark
Estonia
Finland
France
Germany
Greece
Hungary
Iceland
Ireland
Italy
Latvia
Lithuania
Luxembourg
Malta
Netherlands
Norway
Poland
Portugal
Romania
Russia
Slovak Rep.
Slovenia
Spain
Sweden
Switzerland
Turkey
United Kingdom

Parent country

Austria

Belgium

Bulgaria

Croatia

Cyprus

Czech Rep.

Denmark

Estonia

Finland

France

Germany

X
0
0
0
0
10
0
0
0
0
0
0
10
15
0
0
10
15
0
0
0
5
10
0
10
5
10
5
0
0
5
16.5
0

0
X
10
10
0
5
0
0
0
0
0
0
10
15
0
0
10
15
0
0
0
15
10
0
5
15
5
5
0
0
10
16.5
0

0
10
X
5
0
10
5
0
10
5
15
0
10
15
0
10
10
15
0
0
5
15
10
15
10
15
10
15
5
0
5
16.5
0

0
10
5
X
0
5
5
0
5
5
15
0
10
15
20
10
5
5
20
0
0
15
5
30
5
5
5
15
15
0
35
11
0

10
10
5
10
X
10
10
0
29
10
10
0
5
15
0
15
10
15
20
0
25
0
10
30
10
0
10
10
15
0
35
16.5
0

10
5
10
5
0
X
15
0
0
10
5
0
5
5
0
15
5
5
0
0
0
5
5
15
10
10
5
5
5
0
5
16.5
0

0
0
5
5
0
15
X
0
0
0
0
0
5
0
0
0
5
5
0
0
0
0
5
0
10
10
15
5
0
0
0
16.5
0

5
25
15
15
0
5
5
X
0
5
5
0
20
5
0
5
5
0
20
0
5
5
5
30
10
15
15
15
15
0
35
16.5
0

0
0
10
5
0
5
0
0
X
0
0
0
5
0
0
0
5
5
0
0
0
0
5
0
5
5
5
5
0
0
5
16.5
0

0
0
5
5
0
10
0
0
0
X
0
0
5
5
0
0
5
5
0
0
0
0
5
0
10
5
10
5
0
0
5
16.5
0

0
0
15
0
0
5
0
0
0
0
X
0
5
5
0
0
5
5
0
0
0
0
5
0
10
5
5
15
0
0
5
16.5
0

Notes: Withholding tax rates apply to dividends paid by fully owned subsidiaries in subsidiary country to parent rm. Bilateral tax treaties are taken into account. Parent-Subsidiary
Directive is binding between EU Member States and provides exemption from withholding tax if equity holding is at least 25%. The reported gures assume an equity holding in the
subsidiary of at least 25%. In Ireland, subsidiaries owned by parent companies resident in EU or treaty countries are exempt from withholding tax provided that they are not under
the control of persons not resident in such countries. In Italy, authorities can provide a refund equal to the tax claimed limited to 4/9 of the Italian withholding tax if the recipient
can prove a tax is paid in his country on the dividend. In Luxembourg there is an exemption from withholding tax for EU and treaty partners if holding in company resident in
Luxembourg is at least 10%. Source: International Bureau of Fiscal Documentation.

location choices of foreign subsidiaries with low xed assets are more responsive to both host and parent country taxation. This could reect that
high xed assets are a barrier to choosing an alternative physical location. Finally, we nd that international location decisions are relatively
sensitive to international double taxation including the additional parent country taxation if the parent country does not allow the deferral of
foreign source income until dividend repatriation.
Existing studies of the inuence of corporate taxes on multinationals location have in general paid little attention to the role played
by parent country taxes. For instance, Devereux and Grifth (1998) investigate how host country taxation affects the subsidiary location decisions of US multinationals in several large European countries (France,
Germany, and the United Kingdom) over the period 19801994. They
nd that conditional on the choice to locate production abroad
host country average effective tax rates (but not marginal effective tax
rates) are important in determining foreign location choice, even if taxation does not appear to affect the earlier choice to locate abroad or to
export. Buettner and Ruf (2007) in turn nd that location choices of
German multinationals across 18 potential host countries between
1996 and 2003 are affected more by host country statutory tax rates
than effective average tax rates, while they nd no effect of marginal effective tax rates.
Several authors, however, have previously found a role for additional parent country taxation to affect the location of FDI. For US
multinationals, Kemsley (1998) nds that the host country tax only
affects the ratio of US exports to foreign production over the period
19841992 if the multinationals nd themselves in excess credit

positions. 1 Analogously, a role of parent country taxation in affecting

FDI into the United States is found by Hines (1996) who shows that
foreign countries with worldwide taxation invest relatively much in
US states with high state taxes. This reects that multinationals located in countries with worldwide taxation may be able to obtain foreign
tax credits for US state corporate income taxes. Altshuler and Grubert
(2001) evaluate the implications of a hypothetical switch to dividend
exemption by the US to conclude, in contrast, that there is no evidence that such a change would materially affect international location decisions of US multinationals. Egger et al. (2009) construct an
effective tax rate on a bilateral basis that reects overall host and
home country taxation, and nd that this bilateral effective tax rate
has a negative impact on bilateral FDI stocks after controlling for
host and parent country unilateral effective tax rates. 2 However,
none of these papers studies a rm's location decision in a multilateral setting of N by N countries, as we do.

1
US multinationals are subject to worldwide taxation in the United States. Thus,
they have to pay tax in the United States on their foreign-source income, subject to
the provision of a foreign tax credit for taxes already paid in the host country. The foreign tax credit, in practice, is limited to the amount of US tax due on the foreign-source
income. This implies that the overall tax on the foreign income is the host country tax if
this tax exceeds the US tax, while it is the US tax if this tax is the higher of the two. US
taxes on foreign source income can be deferred until the income is repatriated.
2
Repatriation taxes more broadly can affect multinational rms' behavior. Desai et
al. (2001) analyze the effect of repatriation taxes on dividend payments by the foreign
afliates of US multinational rms to nd that 1% lower repatriation tax rates are associated with 1% higher dividends.

S. Barrios et al. / Journal of Public Economics 96 (2012) 946958

949

Greece

Hungary

Iceland

Ireland

Italy

Latvia

Lithuania

Luxembourg

Malta

Netherlands

Norway

Poland

0
0
10
5
0
15
0
0
0
0
0
X
10
15
20
0
10
15
0
0
0
20
15
0
10
15
15
15
0
0
5
16.5
0

10
10
10
5
0
5
5
0
5
5
5
0
X
15
0
10
10
15
0
0
5
10
10
15
5
10
5
10
5
0
10
11
0

25
25
15
15
0
5
0
0
0
5
5
0
20
X
20
27
5
5
0
0
0
0
5
15
10
15
15
15
5
0
5
16.5
0

0
0
5
15
0
5
0
0
0
0
0
0
5
15
X
0
5
5
0
0
0
0
0
0
3
10
0
5
0
0
10
16.5
0

0
0
10
10
0
15
0
0
0
0
0
0
10
15
0
0
10
5
0
0
0
15
10
0
10
5
15
10
0
0
15
16.5
0

25
25
15
5
0
5
5
0
0
5
5
0
20
5
0
27
X
0
20
0
5
5
5
30
10
15
10
5
15
0
5
16.5
0

25
25
15
15
0
5
5
0
0
5
5
0
20
5
0
5
0
X
20
0
5
5
5
30
10
15
10
5
15
0
5
11
0

0
0
5
15
0
5
0
0
0
0
0
0
5
5
0
0
10
15
X
0
0
5
5
0
5
10
5
5
0
0
0
16.5
0

15
15
0
5
0
5
0
0
0
5
5
0
5
15
20
15
5
15
0
X
5
15
5
15
5
15
5
15
15
0
35
16.5
0

0
0
5
5
0
0
0
0
0
0
0
0
5
0
0
0
5
5
0
0
X
0
0
0
5
5
0
5
0
0
0
16.5
0

5
5
15
15
0
5
0
0
0
0
0
0
10
0
0
15
5
5
0
0
0
X
5
15
10
10
5
15
10
0
5
16.5
0

10
10
10
5
0
5
0
0
0
5
5
0
10
5
0
10
5
5
0
0
0
5
X
15
5
10
5
5
5
0
5
11
0

Using information on international M&As, Huizinga and Voget

(2009) nd that the international ownership pattern that emerges reects an effort to avoid international double taxation of foreign
source income. Desai and Hines (2002) argue that international double taxation imposed by the US can explain inversions of existing US
multinationals, whereby the original US parent becomes a subsidiary
and the earlier foreign subsidiary becomes the new parent rm. Voget
(2011) nds that a multinational is more likely to relocate its headquarters abroad, if it is located in a parent country with high taxation
of foreign source income. This varied evidence on how international
double taxation affects the structure of the multinational rm is
consistent with a pattern of foreign subsidiary location that avoids
additional parent country taxation, as found in the present paper.
Huizinga et al. (2012) demonstrate that the international double
taxes that are triggered by an international M&A are to a large extent
capitalized into a lower takeover price. This also suggests that international double taxes are economically burdensome to multinational
rms, providing them with an incentive to own foreign subsidiaries
that are subject to low parent country taxation.
At a theoretical level, Hartman (1985) makes a distinction between
mature foreign afliates that are nanced at the margin by retained
earnings, and immature foreign afliates that rely on funding from
their parent rms. The afliate's required rate of return is shown to reect parent country taxation only if it is immature. Parent country taxation thus should be particularly important for the newly established
foreign subsidiaries that are examined in this paper.
In the remainder of this paper, Section 2 describes the tax treatment of the foreign source income of multinational rms. Section 3
discusses our rm-level data. Section 4 presents estimates of the impact of international taxation on the location of foreign subsidiaries.
Finally, Section 5 concludes.

2. The international tax system

This section describes the corporate tax system applicable to a
multinational company with foreign subsidiaries. 3 Consider a multinational company with a parent located in home country p and a subsidiary located in host country s. Both home and host countries may
tax the subsidiary's income. First, the host country may levy a corporate income tax at a rate ts on this income. Table 1 shows the statutory
corporate income tax rates for the 33 European countries in our sample for the year 2003. 4 These statutory tax rates are those on distributed prots and include local taxes and applicable surcharges. In our
sample, the corporate tax rate for 2003 ranges from a low of 12.5% in
Ireland to a high of 39.6% in Germany.
Next, the host country levies a non-resident dividend withholding
tax at a rate ws on the subsidiary's net of corporate tax income upon
repatriation of this income to the parent. Table 2 provides information on the applicable withholding tax rates on dividends paid by
fully owned subsidiaries to their non-resident parents in 2003. For
example, a dividend paid by a Belgian subsidiary to its parent company located in Estonia will be subject to a withholding tax of 25%, while
the withholding tax on a dividend paid by an Estonian subsidiary to
its Belgian parent company has a zero rate. The withholding tax

3
See Huizinga et al. (2008) for a more detailed description of corporate tax systems
as they apply to multinational companies. This study is limited to the impact of the international taxation of dividends on location decisions, even though the taxation of
other forms of income such as royalties or interest could also be important in corporate
location choices.
4
For illustrative purposes, the tables report taxation data for the year 2003 only, although we have collected these data for the entire period 19992003.

950

S. Barrios et al. / Journal of Public Economics 96 (2012) 946958

Table 2 (continued)
Subsidiary country
Austria
Belgium
Bulgaria
Croatia
Cyprus
Czech Rep.
Denmark
Estonia
Finland
France
Germany
Greece
Hungary
Iceland
Ireland
Italy
Latvia
Lithuania
Luxembourg
Malta
Netherlands
Norway
Poland
Portugal
Romania
Russia
Slovak Rep.
Slovenia
Spain
Sweden
Switzerland
Turkey
United Kingdom

Parent country

Portugal

Romania

Russia

Slovak Rep.

Slovenia

Spain

Sweden

Switzerland

Turkey

United Kingdom

0
0
10
15
0
10
0
0
0
0
0
0
10
10
0
0
10
15
0
0
0
10
10
X
10
10
15
15
0
0
10
16.5
0

15
5
10
5
0
10
10
0
0
10
10
0
5
15
0
10
10
10
0
0
0
10
5
15
X
15
10
15
5
0
10
16.5
0

5
10
15
5
0
10
10
0
0
10
5
0
10
15
0
5
10
15
0
0
5
10
10
15
10
X
10
10
5
0
5
11
0

10
5
10
5
0
5
15
0
0
10
5
0
5
15
0
15
10
10
0
0
0
5
5
30
10
10
X
5
5
0
5
5.5
0

5
5
15
15
0
5
5
0
5
5
15
0
10
15
0
10
5
5
0
0
5
15
5
30
5
10
5
X
5
0
15
16.5
0

0
0
5
5
0
5
0
0
0
0
0
0
5
5
0
0
10
5
0
0
0
10
5
0
10
5
5
5
X
0
10
16.5
0

0
0
10
15
0
0
0
0
0
0
0
0
5
0
0
0
5
5
0
0
0
0
5
0
10
5
0
5
0
X
0
16.5
0

0
10
5
5
0
5
0
0
0
5
0
0
10
5
0
15
5
5
0
0
0
5
5
15
10
5
5
5
10
0
X
16.5
0

25
15
10
10
0
15
15
0
15
15
15
0
10
15
20
15
10
10
20
0
5
20
10
30
10
10
5
15
15
0
35
X
0

0
0
10
5
0
5
0
0
0
0
0
0
5
5
0
0
5
5
0
0
0
5
5
0
10
10
5
5
0
0
5
16.5
X

rates for transactions involving two EU Member States currently are

zero on account of the EU Parent-Subsidiary Directive. 5
The net of withholding tax dividend received by the parent company is in principle taxed in the parent countrysubject to some
form of double tax relief as recommended by the OECD Model Tax
Treaty or as prescribed by the EU Parent-Subsidiary Directive. Some
countries operate an exemption system. In this instance, the dividend
is not taxed in the parent country, if the provided exemption is full.
The overall international rate of taxation on the subsidiary's income
is then given by 1 (1 ts)(1 ws) or ts + ws tsws.
In other instances, the home country may tax the worldwide income of its multinationals and subject the received dividend to corporate income taxation at a rate tp. Generally, a foreign tax credit is
provided for taxes paid in the host country, usually limited to the
amount of the home tax due on the foreign source income. Some
countries apply an indirect tax credit system under which both the
corporate tax and the withholding tax paid in the host country are
credited against the home corporate income tax. In case the home
country's corporate income tax tp is higher than the overall host
country tax rate ts + ws tsws, the rm pays income tax in the
home country at a rate tp [ts + ws tsws] so that the combined, effective tax rate is equal to tp. If instead the home country's corporate
income tax rate is lower than the overall host country's rate, the rm
is said to be in excess foreign tax credit and it will pay no further tax
in the home country (having reduced its home tax liability to zero by
using foreign tax credits). In this instance, the combined, effective tax
rate is ts + ws tsws. In summary, for home countries with an indirect
tax credit system, the combined, effective tax rate is equal to max
[tp, ts + ws tsws].

5
Note that in 2003 prior to their accession, many new EU Member States still
maintained non-zero rates vis--vis EU countries and vice versa.

Home countries may restrict the foreign tax credit to cover only
host country non-resident withholding taxes giving rise to a direct
tax credit system. In this case, the multinational has to pay tax in
the home country to the extent that tp exceeds ws and the combined,
effective tax rate is given by ts + (1 ts) max[tp, ws].
Alternatively, some home countries offer neither exemption nor a
foreign tax credit for taxes paid abroad, but instead allow foreign
taxes to be deducted from home country taxable corporate income.
This amounts to the deduction system with a combined, effective
tax rate of 1 (1 ts)(1 ws)(1 tp).
Finally, in some rather exceptional cases no double tax relief is
provided at all. With full double taxation, the combined, effective
tax rate becomes ts + ws tsws + tp.
Columns 3 and 4 of Table 1 indicate which double tax relief system
is applied by European countries in the sample. As seen in the table,
some countries provide different double tax relief to treaty partners
and non-treaty countries. Thus, we need to know whether there
exist double tax treaties among the countries in our sample. On a bilateral basis, this information is provided in Table 3 with the value 1 indicating the existence of such a treaty and 0 its absence. The table
indicates that for many countries the treaty network is not complete.
For example, in 2003 the Czech Republic has a treaty with all countries
in the sample except Malta and Turkey. From Table 1, we see that this
implies that dividends from all foreign subsidiaries paid to a Czech
parent benet from an indirect tax credit, except for those paid by a
Maltese or a Turkish subsidiary where the deduction system applies.
Information from Tables 13 allows us to calculate the combined
effective tax rate on foreign dividends for any pair of home and host
countries. To x ideas, consider the case of a dividend paid by a Maltese subsidiary to its Czech parent in 2003. Table 1 shows that the
statutory corporate tax rate in Malta is 35%. We infer from Table 2
that net prots paid as a dividend to a foreign company are never
subject to a non-resident dividend withholding tax in Malta. As

S. Barrios et al. / Journal of Public Economics 96 (2012) 946958

already mentioned, Table 3 indicates that no tax treaty was in force

on income repatriated from Malta to the Czech Republic in 2003 so
that from Table 1 we see that incoming foreign dividends benet
from a deduction system in Czech Republic. Finally, the same table
indicates that the applicable corporate tax rate in this country is
31%. From the formula above, the combined, effective tax rate equals
1 (1 0.35) (1 0) (1 0.31) = 55.2%. This rate is considerably
higher than the Maltese corporate tax rate of 35%. This suggests
that the additional taxation of multinational rms, in the form of
withholding taxes and home country corporate income taxation, potentially has an independent and signicant impact on international
location decisions.
Below, we will investigate the independent inuences of host
country corporate income and dividend withholding taxation and
home country corporate income taxation on corporate location decisions. To make parameter estimates comparable across tax measures,
it is useful to construct all three tax measures as shares of the foreign
subsidiary's pre-tax income. The host country tax rate is already dened as a share of the subsidiary's pre-tax income. Our withholding
tax measure will be (1 ts) ws to reect that the withholding tax applies to the subsidiary's income net of the host country corporate tax.
Finally, the additional parent country corporate income tax as a
share of the subsidiary's pre-tax income is computed as the difference between the combined, effective tax rate and ts + ws tsws.
We will dene the international tax to be the sum of the withholding
tax and additional parent country corporate tax both expressed as
shares of the subsidiary's pre-tax income. Equivalently, the international tax is the difference between the combined, effective tax and
the host country corporate income tax.
Unlike host country corporate income taxes, withholding taxes and
home country corporate income taxes are generally deferred until the
foreign source income is repatriated to the parent in the form of dividends. Deferral reduces the present value of taxation. Thus, withholding taxes and home country corporate income taxes are expected to
bite less than host country corporate income taxes. Whether the deferral of withholding taxes and home country corporate income taxes
serves to make these taxes immaterial for location decisions is an empirical matter. This is what we turn to in the empirical section below.
3. Multinational enterprise data
Data on the structure of multinational rms in Europe are taken
from the Amadeus database. This database provides standard accounting data and data on ownership relationships within corporate
groups. 6 We have data on multinational rms operating in 33 European countries over the years 19992003. The ownership data enable us
to match European rms with their domestic subsidiaries and foreign
subsidiaries located in other European countries. 7 We have ownership information for the years 1999, 2001, and 2003. A rm is called
a subsidiary if at least 50% of the shares are owned by a single other
rm. A subsidiary is taken to be new in its year of incorporation.
6
The database is created by collecting standardized data received from 50 vendors
across Europe. The local source for these data is generally the ofce of the Registrar
of Companies.
7
The Amadeus database only contains information on European rms and we therefore only cover the European operations of the multinationals in our sample. In 2001,
intra-EU25 FDI ows amounted to 54% of all outward FDI ows from EU25 countries,
while intra-EU25 FDI stocks were 57% of outward FDI stocks of EU25 countries
(according to Eurostat). Thus, European multinationals have most of their FDI outstanding in other European countries. All the same, our conditioning on a European location of new foreign subsidiaries potentially affects estimated sensitivities of location
to taxation. In particular, by conditioning on European locations rather than locations
worldwide, we increase the expected probability of location in any one country and
potentially also the sensitivity of location to taxation. New foreign subsidiaries tend
to be large relative to the overall assets and capital of the multinational. In our sample,
a new foreign subsidiary on average represents 23% of the assets and 24% of the capital
of the expanded multinational rm including the new foreign subsidiary.

951

In our benchmark analysis below, we consider 909 new foreign subsidiaries. Information on the number of parent and subsidiary countries
involved in these new locations is provided in Panel A of Table 4. Our
benchmark sample excludes new locations where the parent company
becomes an intermediate company as it is a subsidiary itself of another
parent company. The United Kingdom with 115 new parent companies
has most new parent companies, followed by France with 84 new parent companies. Each subsidiary has a home country (where its parent is
located) and a host country (where it is located itself). For each country,
the table lists the number of subsidiaries by home country and by host
country. The table indicates that, for example, France, the Netherlands,
and the United Kingdom are the home country to relatively many
subsidiaries. Hence, there are relatively many subsidiaries with a parent
rm in one of these countries. Denmark, Germany and the United
Kingdom, on the other hand, are the host country to relatively many
subsidiaries.
Our subsequent empirical work on foreign subsidiary location
aims to predict the location of a new foreign subsidiary in 1 of 32 foreign European countries. The dependent variable, called subsidiary
location, takes on a value of one if a particular country is selected as
a subsidiary's location and it is zero otherwise.
Summary statistics on the subsidiary location variable, the tax variables, and some controls are provided in Panel B of Table 4 (see Table
A1 in the Appendix for variable denitions and data sources). The
26,648 observations reported in the table are identical to the number of
observations in the basic regression 1 of Table 5.8 The mean value of the
overall effective tax is 0.35. This mean effective tax, in effect, is the sum
of a mean host country tax of 0.30 and a mean international tax of 0.05.
Among the control variables, GDP bilateral is the ratio of the GDP of a
potential host country and the sum of the GDPs of all other potential foreign (but not domestic) locations. This variable captures market size,
and it is expected to exert a positive impact on the probability of subsidiary location in a host country.
Contiguity is a dummy variable signaling a common border between host and home countries. A common border is expected to
make location in the host country more likely.
Difference in labor costs is the log of the ratio of labor costs in the
home country and labor costs in the host country, expressed in a common currency. The impact of higher labor costs in the host country on
the probability of location is in principle ambiguous (see Kimino et al.
(2007) for a review of the empirical evidence). Higher labor costs in
the host country (or a lower difference in labor costs variable) are
expected to discourage location for a given level of labor productivity
in the host country. In contrast, they potentially encourage location to
the extent that they signal high labor skills and productivity that are
sought after by the multinational rm.
Economic freedom is an index of the extent of soundness of the legal
system, absence of trade barriers, absence of price controls, and transfers
and subsidies as a share of GDP. Economic freedom should make a country attractive as a subsidiary location.
Finally, EU membership is a dummy variable agging EU membership of a prospective host country. EU membership, to the extent that
is signals commitment to high EU standards of dealing with foreign investors, could engender subsidiary location. Moreover, establishing a
subsidiary in one EU country allows non-EU rms to trade freely across
all other EU countries by taking advantage of the EU common market.
Panel C of Table 4 provides correlation coefcients among the location, tax and control variables. Interestingly, location is positively
and signicantly related to the host country tax, but negatively and
signicantly to the international tax. The rst correlation possibly reects that subsidiaries tend to be located in larger countries, which
tend to have relatively high corporate income taxes. In the table,
the host country tax is indeed positively and signicantly correlated
8
Note that the mean value of the location variable is not exactly 1/32 due to the absence of data for some specic combinations of countries and years.

952

S. Barrios et al. / Journal of Public Economics 96 (2012) 946958

Table 3
Existence of bilateral tax treaties for European country pairs in 2003.
Income
To:
Austria
Belgium
Bulgaria
Croatia
Cyprus
Czech Rep.
Denmark
Estonia
Finland
France
Germany
Greece
Hungary
Iceland
Ireland
Italy
Latvia
Lithuania
Luxembourg
Malta
Netherlands
Norway
Poland
Portugal
Romania
Russia
Slovak Rep.
Slovenia
Spain
Sweden
Switzerland
Turkey
United Kingdom

From: Austria Belgium Bulgaria Croatia Cyprus Czech Denmark Estonia Finland France Germany Greece Hungary Iceland Ireland Italy
Rep.
X
1
1
1
1
1
1
0
1
1
1
1
1
0
1
1
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

1
X
1
1
1
1
1
0
1
1
1
1
1
0
1
1
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

1
1
X
0
1
1
1
0
1
1
1
1
1
0
1
1
0
0
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1

1
1
1
X
1
1
1
0
1
1
1
1
1
0
0
1
1
1
0
1
1
1
1
0
1
1
1
0
0
1
0
1
1

1
1
1
1
X
1
1
0
0
1
1
1
1
0
1
1
0
0
0
1
0
1
1
0
1
1
1
1
0
1
0
0
1

1
1
1
1
1
X
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1

1
1
1
1
1
1
X
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

1
0
0
0
0
1
1
X
1
1
1
0
0
1
1
1
1
1
0
1
1
1
1
0
0
0
0
0
0
1
0
0
1

1
1
1
1
0
1
1
1
X
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

1
1
1
1
1
1
1
1
1
X
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

1
1
1
1
1
1
1
1
1
1
X
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

1
1
1
1
1
1
1
0
1
1
1
X
1
0
0
1
0
0
1
0
1
1
1
1
1
0
1
0
1
1
1
0
1

1
1
1
1
1
1
1
0
1
1
1
1
X
0
1
1
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

0
0
0
0
0
1
1
1
1
1
1
0
0
X
0
0
1
1
1
0
1
1
1
1
0
0
0
0
1
1
1
0
1

1
1
1
0
1
1
1
1
1
1
1
1
1
0
X
1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
1
0
1

1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
X
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

Notes: This table addresses whether a tax treaty was in effect to deal with income received by countries listed in the rows and originating from countries listed in the columns.
Specically, 1 denotes that a bilateral tax treaty was applicable and 0 denotes that a tax treaty was not applicable. The table is not exactly symmetric as the dates of rst application
of a treaty may slightly differ between two treaty partners. Source: International Bureau of Fiscal Documentation and various ministries websites.

with the GDP bilateral variable as an index of host country relative

size. The negative correlation between location and the international
tax could reect that subsidiary location is chosen so as to mitigate international double taxation. The host country tax and the international tax are negatively correlated, perhaps reecting the operation of
the foreign tax credit mechanism.

4. Empirical results on taxation and subsidiary location

In this section, we estimate the impact of host and additional parent
country taxation on the choice of location of a new foreign subsidiary. In
our estimation, we explain subsidiary location choice only in the initial
year of establishment to ensure that observations regarding different
location episodes are independent. We condition on a foreign subsidiary location, and hence examine the impact of international taxation
on the location choice among competing foreign locations.
Our sample consists of new foreign subsidiaries that incorporate
in one of the years from 1999 to 2003. The data set encompasses 33
European countries, which implies that for a multinational resident
in a particular home country there are 32 foreign location options. Accordingly, for each new foreign subsidiary we construct 32 binary
variables that take on a value of one if the actual location is in a certain foreign country and zero otherwise. Regarding each potential location for each new foreign subsidiary, there thus is a binary choice.
Location choice is assumed to be determined by the various countries
tax rates and a range of other location or country characteristics. 9 The
9
Note that we explain changes in the rm's structure by the level of international
taxation. Expanding rms may have a need to establish new subsidiaries even if there
are no changes in the level of taxation.

underlying binary choice model is estimated using the conditional

logit approach of McFadden (1974, 1976). 10 In this approach, explanatory variables vary across alternative outcomes. All regressions include
year xed effects, which are not reported.11 We also control for possible
clustering of the errors at the level of the potential host country to allow
for the possibility that shocks to the attractiveness of host countries affect more than a single subsidiary location choice.12 Specically, we report heteroscedasticity-consistent standard errors that are adjusted for
clustering at the level of the host country.
In regression 1 of Table 5, we relate foreign subsidiary location only
to the effective tax rate. The number of observations is 26,648 rather
than 29,088 (or 909 new subsidiaries times 32 potential location countries) due to missing observations of the effective tax variable for some
combinations of countries and years. The effective tax variable enters
with a negative coefcient of 0.87 that is statistically insignicant.
10
The conditional logit model imposes the axiom of independence of irrelevant alternatives (IIA), which implies that adding a third option or changing the characteristics
of a third alternative does not affect the relative odds between any two options considered. As a robustness check, we have estimated a nested logit model that allows us to
relax the IIA assumption yielding results broadly similar to the ones reported in Table 5
(unreported).
11
To enable estimation of the year xed effects, we simultaneously consider how a
multinational selects the location of its foreign subsidiaries over the entire period
19992003, rather than 1 year at a time.
12
The possibility that multinationals' location choices might be correlated both geographically and over time has been suggested in a number of recent empirical studies.
Crozet et al. (2004), for instance, nd evidence of positive spillovers affecting the location choices of French rms investing abroad suggesting that French multinationals
cluster geographically in a gradual process of agglomeration and market penetration.
Head and Mayer (2004) study the determinants of the location choices of Japanese
multinationals in Europe and nd a similar tendency for these multinational rms to
cluster in the same countries.

S. Barrios et al. / Journal of Public Economics 96 (2012) 946958

953

Latvia Lithuania Luxembourg Malta Netherlands Norway Poland Portugal Romania Russia Slovak Slovenia Spain Sweden Switzerland Turkey United
Rep.
Kingdom
0
0
0
1
0
1
1
1
1
1
1
0
0
1
1
0
X
1
0
1
1
1
1
0
1
0
1
1
0
1
1
0
1

0
0
0
1
0
1
1
1
1
1
1
0
0
1
1
1
1
X
0
0
1
1
1
0
1
0
1
1
0
1
1
1
1

1
1
1
0
0
1
1
0
1
1
1
1
1
1
1
1
0
0
X
1
1
1
1
1
1
1
1
1
1
1
1
0
1

1
1
1
0
1
0
1
0
1
1
1
0
1
0
0
1
0
0
1
X
1
1
1
1
0
0
1
0
0
1
1
0
1

1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
X
1
1
1
1
1
1
1
1
1
1
1
1

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
X
1
1
1
1
1
1
1
1
1
1
1

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
X
1
1
1
1
1
1
1
1
1
1

1
1
1
0
0
1
0
0
1
1
1
1
1
1
1
1
0
0
1
1
1
1
1
X
1
1
1
0
1
1
1
0
1

The corresponding estimate of the marginal effect of the effective tax

rate on the probability of location is 0.22, meaning that a one percentage point increase in the effective tax rate (from a level of zero) reduces
the probability of location by 0.22%.
Next, regression 2 includes a host of non-tax control variables, and
it yields an estimated coefcient for the effective tax rate variable of
4.29 that is signicant at the 1% level. The corresponding estimate
of the marginal effect of the effective tax rate on the probability of
location is 0.78. Among the controls, a country's relative GDP is estimated to increase the probability of subsidiary location with significance at the 1% level. Contiguous countries are also more likely to
receive foreign subsidiaries with signicance at the 1% level. The difference in labor costs variable enters with a negative coefcient that
is signicant at 1%, perhaps suggesting that high wages in prospective
host countries signal sought after labor skills. The economic freedom
and EU membership variables both enter with positive coefcients
that are signicant at 10% and 1%, respectively. 13

13
The presence of intangible assets or R&D intensity could similarly affect the sensitivity of international location decisions to taxation. Recent evidence provided by
Dischinger and Riedel (2011) shows that corporate taxation signicantly affects the international location of intangible assets given a multinational rm's structure. To test
this, we distinguished between rms that belong to sectors with low and medium or
high R&D intensity using information from the OECD Science, Technology and Industry
Scoreboard (see OECD, 2003). Following the OECD taxonomy, medium or high R&D intensity sectors have R&D expenditure exceeding 2% of total value added. We estimated
separate regressions analogous to regression 2 of Table 5 for the samples of rms belonging to the two groups of sectors, yielding very similar parameter estimates for
the effective tax rate variable (unreported).

1
1
1
1
1
1
1
0
1
1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
X
1
1
1
1
1
1
1
1

1
1
0
1
1
1
1
0
1
1
1
0
1
0
1
1
0
0
1
0
1
1
1
1
1
X
1
1
1
1
1
1
1

1
1
1
1
1
1
1
0
1
1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
1
1
X
1
1
1
1
1
1

1
1
0
0
1
1
1
0
1
1
1
0
1
0
1
1
1
1
1
0
1
1
1
0
1
0
1
X
1
1
1
0
1

1
1
1
0
0
1
1
0
1
1
1
1
1
1
1
1
0
0
1
0
1
1
1
1
1
1
1
1
X
1
1
0
1

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
X
1
1
1

1
1
1
0
0
1
1
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
X
0
1

1
1
0
0
0
0
1
0
1
1
1
0
1
0
0
1
0
0
0
0
1
1
1
0
1
1
1
0
0
1
0
X
1

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
X

Regression 3 substitutes the host country corporate tax rate for the
effective tax rate. The estimated parameter on the host country tax variable has a value of 3.42 and it is signicant at the 1% level. In line with
this, a one percentage point increase in the host country tax rate is estimated to reduce the probability of location by 0.63%. The control variables enter regression 3 in qualitatively the same way as before.
Regression 4 in turn includes the international tax variable
reecting both non-resident withholding taxation in the host country
and additional parent country corporate taxation with an estimated
coefcient of 1.87 that is signicant at 5%. The corresponding marginal effect of the international tax rate on the probability of location
is estimated to be relatively small in absolute value at 0.19.
Next, regression 5 jointly includes the host country tax rate and
the international tax rate, yielding estimated coefcients of 4.38
and 3.93, that are both signicant at 1%, with corresponding estimated marginal effects of 0.82 and 0.73.
Finally, regression 6 splits up the effective tax rate into its three components: the host country corporate tax rate, the non-resident withholding tax rate, and the additional parent country corporate tax rate.
Parameter estimates for the host country tax rate and parent country corporate tax rate are negative and statistically signicant at 1%, while the
non-resident dividend withholding tax rate obtains a negative coefcient
that is statistically insignicant. A one percentage point increase in the
host country and additional parent country tax rates are estimated to reduce the probability of location by 0.90% and 1.07% respectively, while the
analogous estimated effect of the non-resident withholding tax is 0.06%.
Our results suggest that host country and additional parent country
taxation both play a signicant role in multinationals' location choices.

954

S. Barrios et al. / Journal of Public Economics 96 (2012) 946958

To evaluate the implications of tax rate changes in host and parent countries for location decisions, one needs to recognize that such changes generally alter both our host country and additional parent country tax
variables. An increase in the host country corporate tax rate, for instance,
would lead to an increase in the host country tax variable and an
off-setting reduction in the additional parent country tax variable if a
full foreign tax credit is available (this is the case if the host country tax
rate remains below the parent country tax rate absent nonresident withholding taxes). In this instance, the effective tax rate would remain
unchanged, and the implied estimated impact on location would be
small, as the estimated marginal effects of the host and additional parent
country tax variables on location probabilities of 0.90 and 1.07 are
very similar. An increase in the host country corporate tax rate, however,
increases the host country tax variable while leaving the additional parent country tax variable unchanged, if the host country tax rate exceeds
the parent country tax rate. In this instance, a higher host country corporate tax rate discourages location in the host country, while increasing the
combined, effective tax rate. Overall, our estimation is consistent with a
negative relationship between the host country corporate tax rate and
subsidiary location as generally reported in the literature surveyed by
De Mooij and Ederveen (2006).
Our nding of a large role for the additional parent country tax in
foreign subsidiary location is new, and perhaps surprising, as many
countries in the world allow parent country taxes on foreign source
income to be deferred until dividend repatriation, diminishing its potential to affect location decisions. Similarly, some countries allow
so-called worldwide income averaging. This practice allows multinationals resident in, for instance, the U.S. to claim foreign tax credits for
foreign taxes paid in high-tax countries against U.S. taxes due on income from low-tax countries, potentially reducing the burden of parent country corporate income taxation.
The result that foreign subsidiary location is sensitive to additional
parent country taxation reects that parent country taxation is rather discriminatory, as it only applies to parent rms residing in the pertinent
parent country. Additional parent country taxation thus discourages multinational rm ownership of foreign subsidiaries that are subject to additional parent country taxation. The estimated negative impact of parent
country location on subsidiary location implies that multinationals act
on an apparently signicant incentive to bring about an international
ownership pattern of subsidiaries that is internationally tax-efcient.
Substantial withholding taxes also put particular foreign owners at a
comparative disadvantage at owning local assets vis--vis any other foreign owners that are lowly taxed and local owners. All the same, we nd
that nonresident dividend withholding taxes are statistically insignicant in determining subsidiary location decisions. To explain this, rst
note that the EU Parent-Subsidiary Directive provides that no withholding tax shall be levied on dividend payments between related crossborder companies. This makes the application of a withholding tax rare
in our dataset. Second, companies can potentially avoid withholding
taxes by creating conduit companies or by letting the subsidiary company provide the parent company with a loan instead of a dividend.14
Next, Table 6 reports several robustness checks to gain additional
insight in the impact of the international tax system on foreign
subsidiary location. 15 First, we recognize that the calculation of our

14
The conversion of dividend payments into a loan to the parent potentially also
eliminates parent country taxation on repatriated income. Under US CFC rules, however, such loans could be labelled deemed dividends and trigger parent country taxation. In the EU, transactions of this kind may be subject to transfer pricing rules, but
adjustments of the transfers to qualify as dividends and subsequent adjustments of
parent country taxation seem to be the exception rather than the rule.
15
As additional robustness checks, we have performed regressions that inter alia (i)
exclude holding companies, (ii) include intermediate companies that are both parent
and subsidiary, and (iii) include tax variable interactions with a dummy variable signaling a previously established subsidiary in the country. Our main results on the impact of the international tax system on foreign subsidiary location are unaltered in each
of these regressions (unreported).

effective tax variable is based on countries top statutory tax rates,

thus ignoring potential differences in tax base denitions regarding,
for instance, interest deductibility that equally affect the tax burden
in a particular country. As an alternative effective tax measure, regression 1 of Table 6 includes the Effective Average Tax Rate (EATR) for
outbound FDI as computed by ZEW (2008) starting from regression
2 of Table 5. The EATR reects statutory information on tax rates, interest deductibility, as well as allowances for capital depreciation,
and it is constructed as the overall tax liability triggered by a project
divided by the overall returns generated by the project in the absence
of taxation (see Devereux and Grifth, 1999, 2003). The EATR approximates an average tax on FDI, unlike our effective tax rate based on top
statutory tax rates, and thus can be expected to inuence international
location decisions. In regression 1, the EATR variable obtains a negative coefcient of 2.51 that is statistically insignicant. The marginal
effect in the EATR regression is 0.60. The ZEW (2008) data do not
provide a split up the EATR variable into host and parent country components and hence we cannot perform regressions using EATR-like tax
variables analogous to regressions 36 of Table 5.
Next, we investigate whether the responsiveness of subsidiary location is different for subsidiaries with a low or high return on assets.
Subsidiaries with a low return on assets presumably pay low taxes,
suggesting that there is a reduced role of taxation to affect location choices. Alternatively, subsidiary location of foreign subsidiaries
with a high return on assets could be rather insensitive to taxation,
if the high return on assets reects the realization of rents that are location and owner specic. To investigate this, columns 2 and 3 of
Table 6 report regressions for the samples of subsidiaries with a rate
of return on assets below and above the sample median. Otherwise,
these two regressions are analogous to regression 6 of Table 5. The
host country tax variable and the additional parent country tax
enter with negative and statistically signicant coefcients in both
regressions 2 and 3. Interestingly, the estimated marginal effects of
the host and additional parent country taxes on the probability of location are both smaller in absolute terms for the high-ROA sample,
which suggests that subsidiaries with a high rate of return experience
rents that are location or owner specic.
Next, we consider whether the responsiveness of location is different for subsidiaries with low and high xed assets. Subsidiaries that
use high xed assets may be difcult to relocate physically, and
hence one expects their location to be less sensitive to the host country tax. Regressions 4 and 5 are based on samples of subsidiaries with
ratios of xed assets to total assets below and above the median. In
regressions 4 and 5, the host country tax and the additional parent
country tax both enter with negative and statistically signicant coefcients (albeit at different levels of signicance). The implied marginal effects of both tax rates on the probability of location are relatively
large in absolute terms for the sample of subsidiaries with low xed
assets. These results suggest that the location of subsidiaries with
low xed assets is relatively more sensitive to host and parent country taxation. 16
As discussed, the option to defer parent country taxes may make
foreign subsidiary location less responsive to the additional parent
country tax. Previously, Huizinga and Voget (2009) have collected information on whether the deferral option is available for parent rms
located in a particular home country. We can use this information to
check whether the deferral option indeed mutes the responsiveness
of location to additional parent country taxation. To start, regression
6 in Table 6 introduces an interaction of the effective tax and a

16
We tested for the equality of coefcients on the tax variable across the two regressions that result from dividing the sample into low ROA/high ROA and low xed assets/
high xed assets subsamples as reported in Table 6 using a Wald test. The results indicate that only the withholding tax variable obtains signicantly different coefcients in
each of the two pairs of regressions. In all four regressions, the withholding tax variable
tax variable, however, obtains a coefcient that itself is insignicant.

S. Barrios et al. / Journal of Public Economics 96 (2012) 946958

955

Table 4
Descriptive statistics for subsidiaries of European multinationals.
Panel A. Number of parent companies and subsidiaries used in basic regression
Country

Number of parent companies by home country

Austria
Belgium
Bulgaria
Croatia
Cyprus
Czech Republic
Denmark
Estonia
Finland
France
Germany
Greece
Hungary
Iceland
Ireland
Italy
Latvia
Lithuania
Luxembourg
Netherlands
Norway
Poland
Portugal
Romania
Russia
Slovak Republic
Slovenia
Spain
Sweden
Switzerland
Turkey
United Kingdom
Total

9
29
9
0
0
0
26
5
19
84
56
33
0
2
9
64
1
0
1
64
33
5
19
0
1
1
6
49
69
13
0
115
722

Number of subsidiaries
By home country

By host country

9
36
11
0
0
0
28
6
23
116
75
48
0
3
11
78
2
0
1
89
45
6
21
0
1
1
6
59
76
14
0
144
909

6
34
2
1
0
5
110
1
30
36
97
4
6
4
39
38
2
0
2
51
74
39
11
9
0
0
0
88
68
6
0
146
909

Panel B. Summary statistics for variables in basic regression

Variable

Mean

Standard dev.

Min

Max

Subsidiary location
Effective tax
International tax
Host country tax
GDP bilateral
Contiguity
Difference in labor costs
Economic freedom
EU membership

26,648
26,648
26,648
26,648
26,648
26,648
26,648
26,648
26,648

0.034
0.353
0.051
0.302
0.033
0.166
0.270
6.430
0.464

0.181
0.073
0.070
0.083
0.054
0.371
0.563
1.021
0.499

0
0.125
0
0
0.0003
0
2.098
3.800
0

1
0.750
0.550
0.567
0.264
1
2.373
8.425
1

Panel C. Correlation matrix for variables in basic regression

Subsidiary location
Effective tax
International tax
Host country tax
GDP bilateral
Contiguity
Difference in labor costs
Economic freedom
EU membership

Subsidiary
location

Effective
tax

International
tax

Host
country tax

GDP
bilateral

1.0000
0.0059
0.0798**
0.0628**
0.1669**
0.0733**
0.0923*
0.1030**
0.1402**

1.0000
0.3369**
0.5976**
0.2662**
0.0794**
0.0397*
0.1442**
0.0622**

1.0000
0.5536**
0.2907**
0.1098**
0.1522*
0.1667**
0.4990**

1.0000
0.4830**
0.1638**
0.0945*
0.0144**
0.4800**

1.0000
0.2324**
0.2416
0.1943**
0.4975**

Contiguity

1.0000
0.1455
0.0794**
0.2092**

Difference in
labor costs

Economic
freedom

EU
membership

1.0000
0.6670*
0. 4547**

1.0000
0.4891*

1.0000

Notes: Subsidiary location is a dummy variable equaling 1 if a subsidiary is located in a country and 0 otherwise. Effective tax is the tax rate on dividend income generated in the potential
subsidiary country resulting from corporate income taxation in host and parent countries and non-resident withholding taxation in the host country. International tax is the difference between the effective tax and the host country corporate tax. Host country tax is the corporate income tax in the subsidiary country, including local taxes and possible surcharges. GDP bilateral
is the ratio of the GDP in a potential host country to the sum of GDP of all potential host countries. Contiguity is a binary variable equal to 1 if the parent and potential host countries have a
common border. Difference in labor costs is the log of the ratio of labor costs in the home country and labor costs in the potential host country. Economic freedom is an index scaled between
0 and 10 reecting the following Fraser indicators of the potential host country: soundness of legal system, absence of trade barriers, and absence of price controls. EU membership is a binary
variable equal to 1 if the potential host country is a member of the European Union. Basic regression refers to regression 2 in Table 5. * denotes signicance at 5%; ** signicance at 1%.

deferral option dummy in regression 2 of Table 5. The effective tax

variable now obtains a coefcient of 9.17 that is signicant at 1%
and lower than the corresponding coefcient of 4.29 in regression

2 of Table 5. The interaction of the effective tax with the deferral option dummy obtains a positive coefcient of 4.85 that is signicant at
1%. The corresponding marginal effects on the probability of location

956

S. Barrios et al. / Journal of Public Economics 96 (2012) 946958

Table 5
Taxation and foreign subsidiary location. The dependent variable is subsidiary location equaling 1 if a subsidiary is located in a country and 0 otherwise. Effective tax is the tax rate
on dividend income generated in the potential subsidiary country resulting from corporate income taxation in host and parent countries and non-resident withholding taxation in
the host country. Host country corporate tax is the corporate income tax in the subsidiary country, including local taxes and possible surcharges. International tax is the difference
between the effective tax and the host country corporate tax. Withholding tax is the non-resident withholding tax burden imposed by the host country. Additional parent country
corporate tax is the additional corporate tax burden imposed by parent country after double tax relief. GDP bilateral is the ratio of the GDP in a potential host country to the sum of
GDP of all potential host countries. Contiguity is a binary variable equal to 1 if the parent and potential host countries have a common border. Difference in labor costs is the log of
the ratio of labor costs in the home country and labor costs in the potential host country. Economic freedom is an index scaled between 0 and 10 reecting the following Fraser
indicators of the potential host country: soundness of legal system, absence of trade barriers, absence of price controls, and transfers and subsidies as a share of GDP. EU membership is a binary variable equal to 1 if the potential host country is a member of the European Union. All regressions include year xed effects that are not reported. Sample reects
location decisions of new foreign subsidiaries. Estimation is by conditional logit model. Standard errors that are heteroskedasticity-consistent and adjusted for clustering at the level
of the host country are reported between brackets. Marginal effect is the slope of the probability curve with respect to an included tax variable evaluated at zero while other explanatory variables are evaluated at their mean. * denotes signicance at 10%; ** signicance at 5% and *** signicance at 1%.
(1)

(2)

(3)

(4)

(5)

(6)

Variables (expected sign)

Tax
only

Basic
regression

Host country
corporate tax

International
taxation

Host country corporate and

international taxes

Withholding and parent

country corporate tax

Effective tax ()

0.871
(0.574)

4.292***
(0.634)
4.806***
(0.654)

1.869**
(0.938)

4.376***
(0.624)
3.927***
(1.042)

3.424***
(0.500)

Host country corporate tax ()

International tax ()
Withholding tax ()
Additional parent country corporate
tax ()
GDP bilateral (+)

26,648
0.001

7.420***
(0.575)
0.396***
(0.084)
0.692***
(0.179)
0.140*
(0.073)
0.952***
(0.105)
26,648
0.135

0.218

0.784

Contiguity (+)
Differences in labor costs (+/)
Economic freedom (+)
EU membership (+)
Observations
Pseudo R-squared
Marginal effects of tax variables
Effective tax
Host country corporate tax
International tax
Withholding tax
Additional parent country
corporate tax

7.196***
(0.548)
0.372***
(0.085)
0.768***
(0.190)
0.139*
(0.073)
1.192***
(0.119)
26,648
0.133

5.404***
(0.475)
0.372***
(0.086)
0.573***
(0.174)
0.291***
(0.067)
0.814***
(0.117)
26,648
0.127

7.470***
(0.566)
0.395***
(0.084)
0.705***
(0.186)
0.134*
(0.074)
0.981***
(0.123)
26,648
0.136

0.189

0.816
0.732

0.632

are 1.63 and 0.86. These results suggest that the deferral option reduces the responsiveness of location to the effective country tax by
about half.
In analogous fashion, regression 7 of Table 6 includes an interaction of the international tax variable with the deferral option
dummy in regression 5 of Table 5. The international tax variable
and its interaction with the deferral option dummy enter with negative and positive coefcients, respectively, that are signicant at
1% and 5% respectively. The implied marginal effects of the two variables on the probability of location are 3.10 and 2.48, which suggests that the deferral option reduces the responsiveness of location
to the international tax variable by 80%. Finally, regression 8 of
Table 6 includes an interaction of the additional parent country
tax with the deferral option dummy in regression 6 of Table 5.
The additional parent country tax obtains a negative coefcient
that is signicant at 5%, while its interaction with the deferral option dummy obtains a positive coefcient that is statistically insignicant. Estimated marginal effects of the two variables on the
probability of location suggest that the deferral option reduces the
responsiveness of location to the additional parent country tax by
about 72%.

0.322
(1.400)
5.731***
(1.400)
7.620***
(0.571)
0.402***
(0.084)
0.737***
(0.181)
0.133*
(0.074)
1.107***
(0.129)
26,648
0.137

0.895
0.060
1.070

Overall, the robustness checks conrm that international double

taxation, as reected in our international tax and additional parent
country tax variables, is important in explaining location decisions regarding new foreign subsidiaries. More specically, our results suggest that foreign subsidiary location is relatively sensitive to
international double taxation for subsidiaries with a low return on assets, low xed assets and in the absence of the deferral option.
5. Conclusions
This paper provides evidence on the implications of international
taxation for the location of the foreign subsidiaries of multinational
rms. Our analysis uses panel data on the structure of multinational
rms in 33 European countries over the period 19992003. This rich
data set allows us to estimate the separate effects of host country
corporate income taxation, host country non-resident dividend withholding taxation, and additional parent country corporate income
taxation on the location decisions of multinational rms.
Our main result is that the additional parent country corporate
taxation of foreign source income has an independent, strongly negative effect on the probability of foreign subsidiary location in potential

S. Barrios et al. / Journal of Public Economics 96 (2012) 946958

957

Table 6
Taxation and foreign subsidiary location: robustness tests. The dependent variable is subsidiary location equaling 1 if a subsidiary is located in a country and 0 otherwise. Effective
tax is the tax rate on dividend income generated in the potential subsidiary country resulting from corporate income taxation in host and parent countries and non-resident withholding taxation in the host country. Host country corporate tax is the corporate income tax in the subsidiary country, including local taxes and possible surcharges. International
tax is the difference between the effective tax and the host country corporate tax. Withholding tax is the non-resident withholding tax burden imposed by the host country. Additional parent country corporate tax is the additional corporate tax burden imposed by parent country after double tax relief. GDP bilateral is the ratio of the GDP in a potential host
country to the sum of GDP of all potential host countries. Contiguity is a binary variable equal to 1 if the parent and potential host countries have a common border. Difference in
labor costs is the log of the ratio of labor costs in the home country and labor costs in the potential host country. Economic freedom is an index scaled between 0 and 10 reecting
the following Fraser indicators of the potential host country: soundness of legal system, absence of trade barriers, absence of price controls, and transfers and subsidies as a share of
GDP. EU membership is a binary variable equal to 1 if the potential host country is a member of the European Union. All regressions include year xed effects that are not reported.
The EATR is the cross-border Average Effective Corporate Tax Rate taken from ZEW (2008). Deferral is a dummy variable equal to 1 if deferral is possible and zero otherwise. Regressions 2 and 3 split the sample into subsidiaries that have respectively a ratio of pre-tax prots to total assets below or equal to and above the median value. Regressions 4 and 5
split the sample into subsidiaries that have respectively a ratio of xed assets to total assets below or equal to and above the median value. Sample reects location decisions of new
foreign subsidiaries. Standard errors that are heteroskedasticity-consistent and adjusted for clustering at the level of the host country are reported between brackets. Marginal effect
is the slope of the probability curve with respect to an included tax variable evaluated at zero while other explanatory variables are evaluated at their mean. * denotes signicance at
10%; ** signicance at 5% and *** signicance at 1%.
(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

Variables (expected sign)

EATR

Low ROA

High ROA

Low xed
assets

High xed
assets

Deferral

Deferral
two-way
split

Deferral
three-way
split

EATR

2.513
(1.883)

4.504***
(0.622)
16.61***
(5.620)

4.882***
(0.652)

9.166***
(1.798)

Effective tax ()
Host country corporate
tax ()
International tax ()

4.367***
(0.806)

5.238***
(0.912)

5.911***
(0.955)

3.930***
(0.869)

Withholding tax ()

0.383
(1.725)
5.951***
(1.802)

1.330
(2.068)
5.594***
(1.982)

1.036
(2.191)
6.803**
(3.239)

1.114
(1.627)
5.277***
(1.475)

Additional parent country

corporate tax ()
Deferral Effective tax (+)
Deferral International
tax (+)
Deferral Additional parent
country tax (+)
GDP bilateral (+)
Contiguity (+)
Labor cost (+/)
Economic freedom (+)
EU membership (+)
Observations
Pseudo R-squared
Marginal effects of tax variables
Effective tax
Host country corporate tax
International tax
Withholding tax
Additional parent country
corporate tax
Deferral Included
tax variable

0.202
(1.400)
17.850**
(8.167)
4.853***
(1.711)
13.26**
(5.580)

4.765***
(0.810)
0.664***
(0.109)
0.197
(0.693)
0.038
(0.280)
1.446***
(0.159)
12,175
0.108

7.066***
(0.779)
0.448***
(0.114)
1.027***
(0.199)
0.041
(0.080)
0.902***
(0.168)
14,176
0.119

8.094***
(0.772)
0.347***
(0.119)
0.353
(0.300)
0.351***
(0.104)
1.407***
(0.185)
12,384
0.170

9.367***
(0.811)
0.408***
(0.122)
1.081***
(0.330)
0.048
(0.113)
1.552***
(0.213)
11,658
0.192

5.980***
(0.771)
0.396***
(0.113)
0.575***
(0.214)
0.183*
(0.094)
0.849***
(0.158)
14,990
0.106

1.093

0.366

1.342

0.640

0.097
1.486

0.094
0.391

0.237
1.545

0.174
0.854

0.603

7.365***
(0.574)
0.412***
(0.084)
0.682***
(0.178)
0.149**
(0.072)
0.897***
(0.107)
26,648
0.137

7.478***
(0.561)
0.411***
(0.084)
0.713***
(0.186)
0.133*
(0.074)
0.977***
(0.126)
26,648
0.137

12.86
(8.180)
7.616***
(0.568)
0.416***
(0.084)
0.731***
(0.180)
0.133*
(0.073)
1.112***
(0.131)
26,648
0.138

1.625

host countries, despite the fact that parent country taxation can generally be deferred until income is repatriated. This result may reect
that a parent country's taxation is rather discriminatory as it only applies to parent rms residing in the parent country. The high sensitivity of foreign subsidiary location to parent country taxation suggests
that this tax is particularly distortive. Paradoxically, the parent country tax may be highly distortive on account of the foreign credit
mechanism aiming to alleviate international double taxation as
the foreign tax credit mechanism produces a high variability of the

0.841
3.100

0.910
0.037
3.329

0.860

2.475

2.398

(post credit) parent country tax across foreign location choices. The
sensitivity of foreign subsidiary location to parent country taxation
strengthens the case for abolishing worldwide taxation in favor of
territorial taxation.
Additional parent country taxation may not only distort the foreign subsidiary location decision, but also the capital investment decision once location is determined. The impact of parent country
taxation on capital investment abroad would be an interesting area
for future research.

958

S. Barrios et al. / Journal of Public Economics 96 (2012) 946958

Appendix A

Table A1
Variable denitions and data sources.
Variable

Denition

Source

Subsidiary location
Effective tax

Dummy variable equal to 1 if a subsidiary is located in a country and zero otherwise

Combined, bilateral tax on the subsidiary's income if repatriated in the form of dividends
as a share of the subsidiary's pre-tax income
Sum of withholding tax and parent country corporate income tax as a share of the
subsidiary's pre-tax income
Withholding tax calculated as a share of the subsidiary's pre-tax income

Amadeus database (Bureau Van Dijk)

International Bureau of Fiscal Documentation
and various ministries
International Bureau of Fiscal Documentation and
various ministries
International Bureau of Fiscal Documentation and
various ministries
International Bureau of Fiscal Documentation and
various ministries

International tax
Withholding tax

Additional parent
Residual parent tax calculated as the difference between the effective tax and the sum of the
country corporate tax host country corporate income tax and the withholding tax, as a share of the subsidiary's
pre-tax income
GDP bilateral
Ratio of the GDP of the host country to the sum of GDP of all possible foreign locations.
Contiguity
Binary variable taking the value 1 if the parent and the subsidiary countries have a common
border
Difference labor costs
Log of the ratio of labor costs in the home country and labor costs in the potential host country.
Labor costs are labor compensation per unit of output in USD of 1990.
Economic freedom
Time-varying index scaled between 0 and 10 and computed as the average of the following
Fraser indicators for the subsidiary country: impartial courts, absence of trade barriers,
absence of price controls, and transfers and subsidies as a share of GDP
EU membership
Binary variable taking the value 1 if a potential host country is member
of the European Union
EATR
Effective average tax rate
Deferral
Dummy variable equal to 1 if deferral is possible and zero otherwise
ROA
Fixed assets

Ratio of pre-tax prots to total assets

Ratio of xed assets to total assets

References
Altshuler, R., Grubert, H., 2001. Where will they go if we go territorial? Dividend exemption and the location decisions of U.S. multinational corporations. National
Tax Journal 54, 787809.
Barrios, S., Grg, H., Strobl, E., 2005. Foreign direct investment, competition and industrial development in the host country. European Economic Review 49, 17611784.
Buettner, T., Ruf, M., 2007. Tax incentives and the location of FDI: evidence from a panel
of German multinationals. International Tax and Public Finance 14, 151164.
Crozet, M., Mayer, T., Mucchielli, J.-L., 2004. How do rms agglomerate? A study of FDI
in France. Regional Science and Urban Economics 34, 2754.
De Mooij, R., Ederveen, S., 2006. What a difference does it make? Understanding the
empirical literature on taxation and international capital ows. Economic Papers
No. 261, European Commission. D-G for Economic and Financial Affairs, Brussels.
Desai, M., Foley, C., Hines, J., 2001. Repatriation taxes and dividend distortions. National
Tax Journal 54, 829851.
Desai, M.A., Hines, J.R., 2002. Expectations and expatriations: tracing the causes and
consequences of corporate inversions. National Tax Journal 55, 409440.
Devereux, M.P., Grifth, R., 1998. Taxes and the location of production: evidence from a
panel of US multinationals. Journal of Public Economics 68, 335367.
Devereux, M.P., Grifth, R., 1999. The taxation of discrete investment choices. Working
Paper No. W98/16. Institute for Fiscal Studies, London.
Devereux, M.P., Grifth, R., 2003. Evaluating tax policy for location decisions. International Tax and Public Finance 10, 107126.
Devereux, M.P., Mafni, G., 2007. The impact of taxation on the location of capital, rms
and prot: a survey of empirical evidence. Oxford University Centre for Business
Taxation Working Paper 07/02.
Dischinger, M., Riedel, N., 2011. Corporate taxes and the location of intangible assets
within multinational rms. Journal of Public Economics 95, 691707.
Egger, P., Loretz, S., Pfaffermayr, M., Winner, H., 2009. Bilateral effective tax rates and
foreign direct investment. International Tax and Public Finance 16, 822849.
Grg, H., Greenaway, D., 2004. Much ado about nothing? Do domestic rms really benet from foreign direct investment? World Bank Research Observer 19, 171197.

Ameco database (European Commission)

Authors calculation
OECD
Fraser Institute Economic Freedom of the
World Index
Authors calculation
ZEW (2008)
Huizinga and Voget (2009),
Table IA.IV in internet appendix
Amadeus database (Bureau Van Dijk)
Amadeus database (Bureau Van Dijk)

Hartman, D.G., 1985. Tax policy and foreign direct investment. Journal of Public Economics 26, 107121.
Head, K., Mayer, T., 2004. Market potential and the location of Japanese rms in the
European Union. Review of Economics and Statistics 86, 959972.
Hines, J.R., 1996. Altered states: taxes and the location of foreign direct investment in
America. American Economic Review 86, 10761094.
Huizinga, H., Voget, J., 2009. International taxation and the direction and volume of
cross-border M&As. Journal of Finance 64, 12171249.
Huizinga, H., Laeven, L., Nicodme, G., 2008. Capital structure and international debt
shifting. Journal of Financial Economics 88, 80118.
Huizinga, H., Voget, J., Wagner, W., 2012. Who bears the burden of international
taxation? Evidence from cross-border M&As. Journal of International Economics.
88 (1), 186197.
Kemsley, D., 1998. The effect of taxes on production location. Journal of Accounting Research 36, 321341.
Kimino, S., Saal, D., Drifeld, N., 2007. Macro determinants of FDI inows to Japan: an
analysis of source country characteristics. World Economy 30, 446469.
Mayer, T., Ottaviano, G., 2007. The happy few: new facts on the internationalisation of
European rms. Bruegel-CEPR EFIM 2007 Report, Bruegel Blueprint Series.
McFadden, D., 1974. Conditional logit analysis of qualitative choice behavior. In: Zarembka,
P. (Ed.), Frontiers in Econometrics. Academic Press, New York, pp. 105142.
McFadden, D., 1976. A comment on discriminant analysis versus logit analysis. Annals
of Economic and Social Measurement 5, 511523.
OECD, 2003. OECD Science, Technology and Industry Scoreboard 2003. Organization for
Economic Cooperation and Development, Paris.
Voget, J., 2011. Headquarter relocations and international taxation. Journal of Public
Economics 95, 10671081.
ZEW, 2008. Effective levels of company taxation within an enlarged EU. Study for the
European Commission.