There's No Place Like Home: The Profitability Gap between Headquarters and their Foreign Subsidiaries

• Author: Bodo Knoll,Nadine Riedel,Matthias Dischinger
• Date: 01 June 2014
• Published date: 01 June 2014
• DOI: http://doi.org/10.1111/jems.12058

There’s No Place Like Home: The Profitability Gap

between Headquarters and their Foreign

Subsidiaries

MATTHIAS DISCHINGER

Department of Economics,

University of Munich,

Akademiestr. 1/II, 80799 Munich, Germany

dischinger@lmu.de

BODO KNOLL

Institute of Economics (520D),

University of Hohenheim,

70593 Stuttgart, Germany

bodo.knoll@uni-hohenheim.de

NADINE RIEDEL

Department of Economics,

University of Hohenheim,

70593 Stuttgart, Germany

nadine.riedel@uni-hohenheim.de

Using data on European firms, this paper provides evidence that an overproportional fraction of

multinational group profitsaccrues with the corporate headquarters. Quantitatively, the estimates

suggest that headquarters are by around 25% more profitable than their foreign subsidiaries,

whereas this gap tends to decline over time. The effect turns out to be robust against controlling for

observed and unobserved heterogeneity between the entities. Analogous (although quantitatively

smaller) effects are found for national groups. We discuss various welfare implications of our

findings.

1. Introduction

With increasing economic integration, the importance of multinational entities (MNEs)

has steadily grown over the last decades. Today, more than one third of international

trade flows through intrafirm channels and global FDI totalled $1.3 trillion in 2006

(see e.g., OECD, 2007). Thus, it is not surprising that the emergence of multinational

enterprises and the determinants of FDI are well studied in the economic literature (see

e.g., Barba Navaretti and Venables, 2006; Brakman and Garretsen, 2008).

Although the literature largely focusses on the investment structures of MNEs,

welfare implications of multinational activity are often related to the international dis-

tribution of corporate profits rather than assets as the former impact directly on welfare

We are indebted to Johannes Becker, Ronald Davis, Michael Devereux, Clemens Fuest, Andreas Haufler,

Harry Huizinga, Joel Slemrod, Johannes Voget and to participants of the Zeuthen Workshopat the University

of Copenhagen, the workshop Taxes & the Financial & Legal Structureof Firms at the University of Vienna, the

CESifo Public Sector Conference and seminars at the Universities of Munich, Tilburg and Dublin for helpful

comments. Financial support of the German Research Foundation (DFG) is gratefully acknowledged.

C2014 Wiley Periodicals, Inc.

Journal of Economics & Management Strategy, Volume23, Number 2, Summer 2014, 369–395

370 Journal of Economics & Management Strategy

components like the firms’ local tax payments and workers’ wages (see e.g., Budd et al.,

2005; Devereux and Maffini, 2007). Consequently, differences in the distribution of real

activity and multinational profits across locations may have important welfare implica-

tions. Taking up this idea, recent public finance papers show that multinational firms

distort the location of profits toward low-tax affiliates in order to reduce their overall

tax burden (see e.g., Clausing, 2003; Devereux and Maffini, 2007; Huizinga and Laeven,

2008; Dischinger and Riedel, 2011).

Along the same line, our paper suggests that an overproportional fraction of the

multinational profit accrues with the corporate headquarters. This hypothesis is estab-

lished by a set of theoretical contributions. Papers on “vertical” foreign direct investment

(FDI), for example, propose that a profit bias in favor of the headquarter firms may arise

due to agency costs faced by the headquarters’ management if valuable assets and func-

tions are located with geographically separated subsidiaries (see e.g., Chang and Taylor,

1999; Hamilton and Kashlak, 1999; O’Donnell, 2000). Studies on “horizontal” FDI imply

that investments at the parent location may be more profitable because MNEs have

advantages when operating in their home market as they know the language, culture

and customs better than foreign competitors (see e.g., Dunning, 1977; Brakman and

Garretsen, 2008).1

To test for a profitability gap between headquarters and their foreignsubsidiaries,

we exploit a large European firm data set which is available for the years 1999–2006.

Our results indicate that the profitability of headquarters’ investments outweighs the

profitability of investments at foreign subsidiaries by around 25%. Evaluated at the sam-

ple median, this corresponds to a difference in pretax profits per total assets of around

2-percentage points. The results turn out to be robust against the use of different prof-

itability measures and the inclusion of a largeset of control variables that absorb observed

and time-constant unobserved differences between multinational groups and affiliates.

The estimations furthermore account for potential endogeneity of firm-specific regres-

sors by using an instrumental variables approach and distinguish between greenfield

and M&A investment. The results also turn out to be robust against using propensity

score matching techniques.

The analysis moreover suggests that this profitability gap is not unique to the in-

ternational context but prevails between headquarters and their domestic subsidiaries

(whereas the bias is quantitatively smaller in domestic groups). Furthermore, we show

that the profitability gap between parents and their foreign subsidiaries declines over

time. In our sample period from 1999 to 2006, the profitability bias drops by around

1-percentage point per year, which may reflect a decline in agency costs as new techno-

logical developments, like the invention of the internet and the mobile phone, have led

to reductions in communication and monitoring costs.

On top, the profitability gap shows up in subsamples of MNEs with “vertical”

and “horizontal” FDI (proxied by subsidiaries which operate in different and the same

4-digit NACE industry than their parent, respectively). As the agency cost argument

1. Note that our analysis focuses on the profit distribution within multinational firms and thus on foreign

capital links. The idea of the paper, however,extends to other contractual relationships. For example, agency

problems are likely to be even more severe if corporations have contracts with third parties, for example,

suppliers, in foreign countries. Analogously, abstracting from transport costs, firms may earn lower profits

from exports to foreign countries due to the home market effect described above. Although, it would certainly

be interesting to incorporate an analysis of these activities in our study, we are restricted by our firm data

which unfortunately does not provide information on corporate exports and imports and other contractual

ties to third parties.