The (Im)possibility of Distinguishing Horizontal and Vertical Motivations for FDI

• Date: 01 February 2014
• DOI: http://doi.org/10.1111/rode.12074
• Published date: 01 February 2014
• Author: Maureen Lankhuizen

The (Im)possibility of Distinguishing Horizontal and

Vertical Motivations for FDI

Maureen Lankhuizen*

Abstract

Skill differences between parent and host countries are considered a key variable for distinguishing hori-

zontal and vertical motivations within aggregate foreign direct investment (FDI). This paper tests the

robustness of the skill difference term in the knowledge-capital model for FDI in a sample of Organisation

for Economic Co-operation and Development (OECD) countries. The evidence in this paper indicates that

skill differences per se do not properly explain FDI: the skill level of the host country is also important. This

paper argues that both horizontal and vertical FDI may increase in the skill level of the host. It follows that

the distinction between vertical and horizontal motivations for FDI with respect to skills is less straight-

forward than generally assumed in the literature.

1. Introduction

The theoretical literature on foreign direct investment (FDI) generally distinguishes

two main models of FDI. In the horizontal model of FDI (originally developed by

Markusen, 1984) firms set up plants in multiple markets to exploit firm-specific assets

and to avoid transport costs and trade barriers. The vertical model (Helpman, 1984)

entails the geographical separation of production and headquarter activities so as

to exploit factor-cost differentials caused by different relative factor supplies. The

knowledge-capital model (Markusen, 2002) integrates horizontal and vertical motiva-

tions for FDI in a single general-equilibrium model of multinational enterprises

(MNEs). Much of the empirical literature on FDI has focused on whether horizontal

or vertical motivations for FDI dominate. This debate has centered largely on the role

of countries’ relative skill endowments. According to the theory, horizontal FDI

dominates when countries have similar relative skill endowments. Vertical FDI domi-

nates when countries differ in relative skill endowments. Headquarter activities are

assumed to be more skilled-labor intensive than production. Therefore, when hosts

are skill-scarce relative to the parent country, this will motivate firms to relocate pro-

duction away from headquarters.

In line with the theoretical predictions, the empirical specifications of FDI in the lit-

erature usually include a skill difference term, measuring the difference in relative

skill endowments between a parent and a host country. This paper states that control-

ling for skill differences in this manner entails imposing the linear restriction that

parent- and host-country skill levels have an equal but opposite effect on FDI. The

paper tests this restriction for a sample of Organisation for Economic Co-operation

and Development (OECD) countries using the empirical specification of the

knowledge-capital model of Carr et al. (2001). I use FDI stocks and, to test the

*Lankhuizen: Department of Spatial Economics, VU University Amsterdam, De Boelelaan 1105, 1081 HV

Amsterdam, The Netherlands. Tel: +31-20-598-6088; Fax: +31-20-598-6004, E-mail: m.b.m.lankhuizen

@vu.nl. Useful comments and suggestions by Henri L. F de Groot, Eelke de Jong, Gert-Jan Linders and an

anonymous referee are gratefully acknowledged. The usual disclaimer applies.

Review of Development Economics, 18(1), 139–151, 2014

DOI:10.1111/rode.12074

© 2014 John Wiley & Sons Ltd