Search

This paper exploits an international bilateral data set over the period 1963-1998 to investigate the relationship between foreign direct investment (FDI) and foreign-educated labor in an FDI host country. Workers educated abroad acquire country-specific human capital that is more productive in the host country of study. A foreign subsidiary sharing a parent firm's technology will invest more if it has more foreign-educated labor, since it can utilize this labor more productively because of the country-specific human capital. Consistent with our predictions, our empirical findings show that foreign-educated labor accounted for a sizable portion of growth in FDI flows. [ABSTRACT FROM AUTHOR]

Earlier work found evidence for geographic linkages of aggregate foreign direct investment across countries and country‐pairs. From a theoretical point of view, such linkages at the macroeconomic level may root in between‐firm as well as within‐firm linkages and originate from information spillovers across or within firms in exploring unknown markets, and vertical linkages between production plants across different locations within the firm. We use data on the universe of German multinational enterprises (MNEs) to empirically explore how marginal investments at one foreign affiliate depend on investments at other affiliates within the same MNE. The empirical approach employs two channels or modes of cross‐affiliate interdependence: mere geography (capturing horizontal linkages through correlated learning and horizontal competition within the firm) and input–output relationships within or across industries (which capture vertical linkages). Adding to earlier findings at the aggregate level, we find evidence of a significant interdependence of investments within the firm. In the firm‐level data at hand, vertical linkages appear to be more important than horizontal ones. Investments at one location tend to stimulate investments at other locations of the same MNE, particularly if input linkages are strong. The opposite seems to be true for output linkages. Beyond vertical linkages, mere geographic proximity matters only to a minor extent. This suggests that evidence of linkages through geographic closeness at aggregate data levels accrue mainly to reasons of vertical linkages within networks of affiliates. (JEL C31, D22, F21, F23, F68, G31, H32) [ABSTRACT FROM AUTHOR]