مالکیت خارجی و بهره وری از جهت علیت تا آشکار

This paper estimates the effect of foreign ownership on productivity in a very general setting where all potential endogeneity sources are controlled for. In particular we apply the GMM-System estimator to estimate TFP for a large sample of firms located in Italy. After controlling for unobserved heterogeneity, input simultaneity and measurement errors, foreign ownership has no effect on productivity. When we also control for the simultaneity of the ownership variable we find that nationality matters since firms under US ownership tend to be more productive than firms under national ownership. Therefore we do not find widespread empirical support to the standard internalization theory of foreign direct investment. In particular, the transfer of technology seems to occur only if the difference between the recipient and the investment country is sufficiently pronounced. Our results also highlight the importance of controlling for simultaneity of the foreign ownership variable.

The existence of Multinational Enterprises (MNEs) and the characteristics of their foreign affiliates have been extensively examined by the economic literature. The most widely accepted theory on MNEs, the so-called “internalization theory” (see Caves, 1996), provides some insights to both issues. It suggests that MNEs exist so that they can exploit superior knowledge (e.g. managerial expertise and superior technological skills) in foreign markets, which compensates for the higher costs induced by operating in a foreign environment. In turn, MNEs' foreign affiliates should benefit from the transfer of this knowledge and might therefore display higher productivity and profitability levels with respect to domestic-owned firms. The alleged superior performance of MNEs' subsidiaries compared with domestic-owned firms has been widely documented in early empirical works, mostly using aggregate data, and has become a “stylised fact” in the literature on MNEs (Conyon et al., 2002).

Both received theory on multinational firms and conventional wisdom point out that subsidiaries of foreign firms should operate more efficiently than local firms. However, this may occur for different reasons. Indeed, it may be explained by the fact the MNEs possess superior managerial and/or technological capabilities that can be transferred to their foreign affiliates. However, it might also be the outcome of a preference for MNEs to acquire the best locals or to operate in the most productive industries. Finally, as pointed out by Griffith, 1999a and Griffith, 1999b it may simply be that both groups of firms are drawn from the same distribution but only the best foreign owned firms have chosen to locate in a given country.