Multinational Enterprises, Development and Globalization: Some Clarifications and a Research Agenda

Abstract

This paper considers how economic globalization has affected opportunities and challenges for developing countries in following a multinational enterprise (MNE)-assisted development strategy, revisiting an earlier article by the authors. The growing share of industrial activity owned and/or controlled by MNEs has not—by and large—led to a proportional increase in sustainable domestic industrial growth. Particular attention is paid to how MNEs have responded proactively to globalization by modifying their strategies, spatial organization and the modalities by which they interact with host economic actors, and how these changes alter our understanding of MNEs and development. What has been learnt over the last decade about embeddedness, institutions, inertia, absorptive capacity, spillovers and linkages, and how they can explain the success of some countries (or regions) in promoting growth, and the failure of others, is examined. The need to link MNE and industrial policies systematically is highlighted. Attracting the “right kinds” of MNE activity remains important, but greater heterogeneity of MNE activity and host locations requires greater customization of policy tools.

Notes

Professor Dunning passed away on 29 January 2009. The final version of this paper was written after Professor Dunning's death, incorporating his comments on a first draft. While I have tried to stay within the bounds of what he might have agreed, I take full and sole responsibility for all errors and omissions.

Research assistance by Ping-Shan Cheng is gratefully acknowledged. On the final version, comments from Philippe Gugler, Lou Anne Barclay, Gabriel Benito, Isabel Alvarez and the anonymous referees have been invaluable.

¹ Although cross-section studies serve to illustrate important issues, panel data sets across countries and specific proxies (such as NOI and GDP) that are used to test the IDP raise several methodological and measurement challenges. Variables such as NOI represent an aggregation of inward and outward FDI, which are themselves also aggregate variables across a variety of sectors and industries, both of which seek to proxy the intensity of MNE activity. Furthermore, all such analyses—with almost no exception—have utilized nominal values of FDI and GDP, and in the case of FDI, variously utilizing stocks, flows, sum-of-flows and average flows as substitutes, a practice for which very little empirical evidence exists. Much the same can be said about the dangers of GDP as a proxy for development.

² See e.g. Chowdhury & Mavrotas (2006) and Hansen & Rand (2006).

³ One might even suggest that multinational activity is merely concatenated with host country growth, rather than being responsible for such growth. In other words, multinational activity may represent a placebo effect, indicative of improving domestic activity rather than being the most important cause of it. However, it is not our intention to investigate this possibility.

⁴ Although there may be growth in GDP, such growth does not imply industrial development.

⁵ See e.g. Narula & Dunning (2000), Barry et al. (2003), Liu et al. (2005) and Galan et al. (2007).

⁶ Others may derive from the difficulties of aggregating to the industry level, as spillovers may accrue to actors external to formally organized firms and sectors. Furthermore, economies may also have peculiar inter-industry relationships and interdependencies, some of which may lie in the informal sector.

⁷ See e.g. Zhang & Van Den Bulcke (1996).

⁸ Of all registered FDI by the end of 2004, the coastal region had 87.8%, the central region had 8.3% and the western region had 3.9%.

⁹ On a GDP per capita basis, Shanghai was US$5280, 10 times greater than that of Guizhou Province in 2004.

¹⁰ For a discussion of these policies, see Amsden (2001) and Lall (1996).

¹¹ For a discussion of these FDI policy stances, see Lall (1996).

¹² Although some research indicates that subsidiaries with greater-than-average competence levels are able to exert higher levels of autonomy (Pearce, 1999; Taggart & Hood, 1999); others have argued that because of the strategic importance of such a subsidiary, headquarters will attempt to exert tighter control (Martınez & Jarillo, 1991).

¹³ There is a considerable literature on the individual MNE's choice of mode of entry (see Meyer et al., 2009).

¹⁴ Iguchi (2008) found that subsidiaries are more likely to create backward linkages with the host economy when the affiliate is seen to yield positive returns.

¹⁵ It may also occur in non-related industries where similar techniques for efficiency gains may be applied through what is known as the demonstration effect. In addition there are numerous spillover effects, which will be discussed in another section. Indeed, it is difficult in practice to separate these different means by which MNE subsidiaries may affect domestic industry.