TOWARDANECLECTICTHEORY OFINTERNATIONALPRODUCTION: SOME EMPIRICALTESTS JOHN H. DUNNING* University of Reading Abstract. This paper first sets out the main features of the eclectic theory of international production and then seeks to evaluate its significance of ownership- and location-specific variables in explaining the industrial pattern and geographical distribution of the sales of U.S. affiliates in fourteen manufacturing industries in seven countries in 1970. * There is now a consensus of opinion that the propensity of an enterprise to engage in international production-that financed by foreign direct investment- rests on three main determinants: first, the extent to which it possesses (or can acquire, on more favorable terms) assets1 which its competitors (or potential competitors) do not possess; second, whether it is in its interest to sell or lease these assets to other firms, or make use of-internalize-them itself; and third, how far it is profitable to exploit these assets in conjunction with the indigenous resources of foreign countries rather than those of the home country. The more the ownership-specific advantages possessed by an enterprise, the greater the in- ducement to internalize them; and the wider the attractions of a foreign rather than a home country production base, the greater the likelihood that an enter- prise, given the incentive to do so, will engage in international production. This eclectic approach to the theory of international production may be summa- rized as follows.2 A national firm supplying its own market has various avenues for growth: it can diversify horizontally or laterally into new product lines, or ver- tically into new activities, including the production of knowledge; it can acquire existing enterprises; or it can exploit foreign markets. When it makes good eco- nomic sense to choose the last route (which may also embrace one or more of the others), the enterprise becomes an international enterprise (defined as a firm which services foreign markets).However, for it to be able to produce alongside in- digenous firms domiciled in these markets, it must possess additional ownership advantages sufficient to outweigh the costs of servicing an unfamiliaror distant environment [Hirsch 1976]. The function of an enterprise is to transform, by the process of production, valu- able inputs into more valuable outputs. Inputs are of two kinds. The first are those which are available, on the same terms, to all firms, whatever their size or nation- ality, but which are specific in their origin to particular locations and have to be used in that location. These include not only Ricardian type endowments-natural resources, most kinds of labor, and proximity to markets,3 but also the legal and commercial environment in which the endowments are used-market structure, and government legislation and policies. In classical and neoclassical trade theories, differences in the possession of these endowments between countries fully explain the willingness and the ability of enterprises to become interna- tional;4 but since all firms, whatever their nationality of ownership, were assumed to have full and free access to them (including technology), there were no advan- tages to be gained from foreign production. INTRODUCTION The Underlying Theory *John H. Dunning is Professor of InternationalInvestment and Business Studies at the Uni- versity of Reading. He has been working in the field of international investment and the multinational enterprise since the mid 1950s and has published several books and numer- ous articles on the subject. The author is much indebted to Professor Guy Landry of Brandon University,Winnipeg who was responsible for most of the computational workbehind Tables 3-6 and who assisted in writing the first draft of pages 12-23. 9 Palgrave Macmillan Journals is collaborating with JSTOR to digitize, preserve, and extend access to Journal of International Business Studies www.jstor.org ®