Review of Industrial Organization 16: 385–397, 2000.
© 2000 Kluwer Academic Publishers. Printed in the Netherlands.
R&D, Foreign Direct Investment and Technology
CATHERINE Y. CO
Department of Economics, University of Central Florida, Orlando, FL 32816, U.S.A.
Abstract. The U.S. experienced a surge in foreign direct investment (FDI) during the 1980s. To
date, we do not fully understand what the research and development (R&D) effects of FDI are. FDI
can affect the host country’s R&D activities in at least two ways. First, if domestic ﬁrms perceive
foreign entrants as serious threats to their domestic market share, they will change their R&D policies
hence industry-wide R&D intensity also changes. Second, the increase in R&D intensity may be
attributed to foreign ﬁrms’ R&D activities to acclimatize to U.S. conditions. Using a panel of 4-
digit SIC industries from 1981–91, there is some evidence that the latter scenario may be a possible
explanation for the observed changes in industry-wide R&D intensity.
Key words: Foreign direct investment, research and development, technology sourcing.
JEL Classiﬁcation: F23, L60, 030.
I. Introduction and Background
Although global outﬂows of foreign direct investment (FDI) reached a high of
$232 billion in 1990, to date, we do not yet completely understand what motivates
ﬁrms to invest abroad and what the consequent effects of FDI are. The traditional
explanation for why ﬁrms invest abroad assumes that ﬁrms have ﬁrm-speciﬁc assets
that give them an advantage over domestic ﬁrms in the host market.
As Neven and
Siotis (1996, pp. 546–547) argue, when ﬁrms have ﬁrm-speciﬁc knowledge, they
presumably would want to engage in greenﬁeld (new plant) FDI to minimize the
chance that others would gain access to this knowledge. The traditional explanation
assumes that foreign ﬁrms have complete information regarding the level of tech-
nological sophistication of domestic ﬁrms. If this assumption is violated, even if
foreign ﬁrms have a technological advantage, they may be deterred from entering
upon observing a sufﬁciently “high” level of research and development (R&D)
I thank an anonymous referee whose comments improved this paper. All remaining errors are
mine. I would like to thank Jian Dong for excellent research assistance.
See Corden (1967), Horst (1971), Buckley and Casson (1976), Hymer (1976), Brecher and
Diaz-Alejandro (1977) and Markusen (1984) for early contributions. Caves (1996) and Markusen
(1995) are recent surveys of this literature.