Foreign investment and migration: Extensions of the basic model

The seminal work by Ramaswami (1968) comparing the advantages of allowing international labor inflows with those of foreign investment elicited an immediate response from Webb (1970). and subsequent elaborations by Bhagwati (1979). Calvo and Wellisz (1983), Bhagwati and Srinivasan (1983), and Ruffin (1984). This work served to confirm Ramaswami's suggestion that if international factor markets are in operation, it is preferable for a capital-rich country to admit foreign labor than to export some of its own capital to foreign countries if such labor could be obtained at the low wage rate prevailing abroad. The model developed in this literature was stripped down to such bare elements as to be dubbed the "basic model" in Jones, Coelho, and Easton (1986), an article which extended the reasoning employed by Ramaswami to analyze optimal policy when an active country can control jointly the international flows of labor and of capital. Simple assumptions often lead to striking conclusions, and optimal policy in the "basic model" was shown to involve subsidizing capital inflows from the low-wage, high-rent country. Indeed, first-best policy on the part of the active, capital-rich country was shown to involve almost a total "buy-out" of the other country's factor supplies at their autarky wages and rentals. The key assumptions of the basic model supporting such a strong result include: (i) The active home country can obtain cheap labor abroad at low foreign wage rates. The Ramaswami reasoning and the "buy-out" strategy break down if foreign workers on net receive the high wages prevailing at home. (ii) Technological knowledge and factor skills are the same in both countries. Both Ruffin (1984). and Calvo and Wellisz (1983) remark on the question of technological parity, but seem to disagree on the consequences of allowing differences in technology. (iii) All (both) factors are assumed to be internationally mobile. If the dimensions of the model were to be extended to allow the existence of some factor (say land) that was immobile, the complete "buy-out" strategy would obviously be thwarted. But would there nonetheless be a tendency for the active home country to attempt to obtain at least some quantities of those factors that are internationally mobile, even if this entails subsidies to get expensive factors to move? An analysis of this issue was presented by Jones and Coelho (1985) in the context of a model with many commodities and (even more numerous) factors. Kuhn and Wooton (1986). and Bond (1986) address this issue in the more simple context in which only one commodity is produced (as in the basic model), both labor and capital can flow between countries (subject to taxation or encouraged by subsidies), but a third factor (land) is immobile in each country. In the present paper we investigate the fate of the Ramaswami type of reasoning, which leads to a buy-out optimal strategy in the basic model, when technologies differ or when a third non-mobile factor is introduced. As in the basic model, international factor flows affect welfare in the active country both by the changes they bring about in world output and by induced changes in the "terms of trade", the factor prices paid to internationally mobile labor and capital. If the terms of trade effect were absent, the active home country would capture all the gains in world output. Our analysis proceeds by considering separately a set of world-output contours in factor space, and a set of terms-of-trade contours. A contract curve of maximal points connects these two sets of contours, and the optimal point for the home country is located on such a contract curve. This graphical device is employed in section I to review the argument for the basic model, in section II to investigate the case in which technology is inferior in the active home country, and in section III to establish the sense in which the Ramaswami reasoning is still in evidence even with immobile land.