Comments and Discussion

Robert Z. Lawrence: The authors of this informative paper present their work as a challenge to literature on multinational firms. Their paper explores the distinction between horizontal foreign direct investment (FDI), which exists because "trade barriers make exporting costly," and vertical FDI, which arises "to take advantage of international factor-price differences." The authors argue that vertical FDI is more prevalent than the literature has suggested. They also find that the motives for FDI are more complicated than the simple distinction between vertical and horizontal FDI implies. In particular, they point out the importance of FDI to seek export platforms, outsource, and provide wholesale services. In these comments, I will first react to the issue of vertical versus horizontal FDI and then turn to some comments on the findings of more diverse motivations for FDI.

My reading suggests that the FDI literature is far more eclectic than the authors imply. This leads me to view the results here as less novel than some of the discussion in the paper indicates. Beyond the very narrow field of general equilibrium economic modeling, the motives for FDI have long been recognized as complicated. ¹ Export platforms and outsourcing have long been recognized as important phenomena. The motivations for U.S. multinationals that invested extensively in Belgium in the late 1950s and early 1960s were surely to service the European market. Likewise, more recently, an important driving force behind the proliferation of regional trade agreements has been to attract FDI to service regional markets. Similarly, if anything, in general discussions about multinational activities, the importance of outsourcing has probably been exaggerated rather than ignored. Nonetheless, the authors have done us a service in pointing out the dangers of the oversimplification prevalent in the general equilibrium literature, although I suspect that those who work in this area have been driven by the demands of analytical tractability rather than ignorance in the modeling strategies they have adopted.

It might be useful to expand on why this debate over the motives for FDI is relevant to current politically charged concerns about globalization. In general, vertical FDI operates as a complement to trade. If FDI is sensitive to relative factor prices, therefore, as with trade, it could put downward pressure on wages, particularly of unskilled and unionized workers in developed countries. But today's concerns are broader, since such FDI could also impose pressures for "a race to the bottom" with respect to regulations and taxes. ² On the other hand, horizontal FDI could be more benign. It could actually substitute for trade and thus mitigate some of these effects. Indeed that is precisely the reason for what Bhagwati has called quid pro quo investment.

Which type of FDI is more important today? Typically, when econometric specifications permit a horse race to test whether the vertical or the horizontal story explains more of the phenomenon, the horse with horizontal colors wins handsomely. This is still the case. As the regressions the authors run suggest, in explaining aggregate sales, or FDI, the horizontal story still prevails. It does best at explaining FDI on average.

This is not surprising, as even a cursory glance at the data suggests. The authors report in table 1, for example, that measured by sales almost 80 percent of U.S. FDI is in OECD countries. It is hard to obtain a precise number for vertical FDI. But we know it is particularly important in transportation, electronics, and machinery. Adopting an expansive definition which includes affiliate activities in all developing countries in the transportation, industrial machinery, and electronics sectors, employment in developing countries in 1997 accounted for just 12.9 percent of all majority owned foreign affiliate employment in 1997 and just 3.2 percent of the FDI position in 1998 valued at historic cost. ³ Thus vertical outsourcing is still a relatively small proportion of overall activity.

Nonetheless, the authors are surely correct in pointing to the growing role for vertical FDI. Indeed, this...