Taxing the Financially Integrated Multinational Firm

• Author: NIELS JOHANNESEN
• Date: 01 August 2016
• Published date: 01 August 2016
• DOI: http://doi.org/10.1111/jpet.12192

TAXING THE FINANCIALLY INTEGRATED MULTINATIONAL FIRM

NIELS JOHANNESEN

Department of Economics, University of Copenhagen

Abstract

This paper develops a theoretical model of corporate taxation in the

presence of financially integrated multinational firms. Under the as-

sumption that multinational firms use some measure of internal loans

to finance foreign investment, we find that the optimal corporate tax

rate is positive from the perspective of a small, open economy. This

finding contrasts the standard result that the optimal-source-based cap-

ital tax is zero. Intuitively, when multinational firms finance invest-

ment in one country with loans from affiliates in another country, the

burden of the corporate taxes levied in the latter country partly falls

on investment and thus workers in the former country. This tax ex-

porting mechanism introduces a scope for corporate taxes, which is

not present in standard models of international taxation. Accounting

for the internal capital markets of multinational firms thus helps re-

solve the tension between standard theory predicting zero capital taxes

and the casual observation that countries tend to employ corporate

taxes at fairly high rates.

1. Introduction

At the heart of the multinational firm is the internal capital market that allocates funds

inside the firm. Affiliates of multinational firms typically finance investment with a com-

bination of funds from external sources such as bank loans and bond emissions and

funds from internal sources such as equity injections and loans from related entities.

The internal capital market thus creates financial linkages within the multinational firm

in the sense that affiliates are tied together by internal loans and equity stakes. The

multinational firm is financially integrated.

The MiDi data set collected by the German Central Bank and recently made avail-

able for research provides a unique opportunity to assess the nature and size of these

financial linkages within the multinational firm. A remarkable feature highlighted by

this data set is the prevalence of internal loans. Buettner and Wamser (2013) report

that foreign affiliates of German firms have an average total debt-asset ratio of around

Niels Johannesen, Department of Economics, University of Copenhagen, Denmark (niels.johannesen@

econ.ku.dk).

I would like to thank Peter Birch Sørensen and Marko Koethenbuerger for valuable comments

and suggestions. I gratefully acknowledge financial support from the Danish Council for Independent

Research.

Received June 11, 2012; Accepted February 18, 2015.

C2016 Wiley Periodicals, Inc.

Journal of Public Economic Theory, 18 (4), 2016, pp. 487–510.

487

488 Journal of Public Economic Theory

0.60, which breaks down into an external debt-asset ratio of 0.35 and an internal debt-asset

ratio of 0.25. This suggests that internal loans represent a source of financing that is

quantitatively almost as important as external loans. The internal debt-asset ratio fur-

ther breaks down into a parent debt-asset ratio of 0.15 and an other affiliate debt-asset ratio

of 0.10. This suggests that parents as well as other affiliates are important lenders in the

internal capital markets of multinational firms.1

The main contribution of this paper is to show that financial linkages within multi-

national firms have important implications for optimal taxation of capital. The paper

develops a model of corporate taxation in the presence of financially integrated multi-

national firms while assuming that foreign investment of multinational firms is financed

with some measure of internal loans and that these internal loans are not exclusively

motivated by profit shifting. These assumptions find strong support in the empirical

facts about German firms documented in the MiDi data set. Not only do foreign affili-

ates of German firms receive around one quarter of their capital in the form of loans

from related entities, but most of these loans derive from affiliates in high-tax countries:

Buettner and Wamser (2013) report that around half of the internal loans derive from

the parent in Germany where the corporate tax rate has consistently been very high

by international standards, while Ruf and Weichenrieder (2012) report that the other

half mostly derive from affiliates in high-tax countries (e.g., the United States and the

United Kingdom) and only to a much smaller extent from affiliates in low-tax havens

(e.g., Cayman Islands, Ireland, Switzerland, and Luxembourg). To the extent that in-

ternal lending was predominantly driven by profit shifting, we should observe internal

loans flowing from low-tax countries to high-tax countries; however, we actually observe

most internal loans flowing out of high-tax countries. This pattern is not inconsistent

with the large body of literature finding that the financial structure of multinational

firms is tax-sensitive, but it certainly seems to imply that other considerations than profit

shifting play an important role in determining flows of capital within the multinational

firm.

The main finding of the paper is that internal loans serving other purposes than

profit shifting introduce a tax exporting mechanism that causes the optimal corporate

tax rate to be positive from the perspective of a small, open economy. This contrasts

with the standard result that the optimal tax on investment is zero (Gordon 1986).

Intuitively, small, open economies are facing a perfectly elastic supply of capital; hence,

any tax that raises the cost of investment is fully shifted to the workers. In the standard

model with only domestic firms, this implies that corporate taxes are strictly dominated

by labor taxes because they distort the labor supply to the same extent as labor taxes

and add a distortion of the capital–labor ratio. In our model, multinational firms partly

finance investment with loans from foreign affiliates. This is effectively shifting the tax

base from the country of the investing affiliate to the countries of the lending affiliates

because interest payments are deductible in the former country and taxable in the latter.

Tothe extent that corporate taxes fall on interest income from loans to foreign affiliates,

1These patterns are consistent with the findings of other papers. Also, based on the MiDi data set, Ramb

and Weichenrieder (2005) report that German affiliates of non-German firms have an average total debt-

asset ratio of 0.53, which breaks down into an external debt-asset ratio of 0.23 and an internal debt-asset ratio

of 0.29. Based on data from the U.S. Internal Revenue Service, Altshuler and Grubert (2002) report an

average total debt-asset ratio of 0.54 for foreign affiliates of U.S. firms with an average parent debt-asset ratio

of 0.11. Desai, Foley, and Hines (2004) report similar figures based on data from the U.S. Bureau of

Economic Analysis. The U.S. data sources do not distinguish between loans from other affiliates than

the parent and loans from third parties and therefore do not contain sufficient information to compute

internal debt-asset ratios.