ABSTRACT
It is more than 250 years since Montesquieu (1749) wrote his “Lettre à William Domville”. In that essay Montesquieu discussed what would now be called the welfare implications of international trade. The chief novelty of the essay lay in its focus on the well-being not of the Prince but of the People; that is, of the population at large. The central questions suggested by it concern the sense in which a country may be said to benefit from the opportunity to trade with other countries and the variety of circumstances under which trade is indeed beneficial. The first of these questions was answered, although not to everyone’s satisfaction, by Pareto (1894), at the end of the nineteenth century. Early in the twenty-first century the second question still awaits a complete answer. However, much progress has been made, especially during the past thirty years or so. In particular, it has been shown by Grandmont and McFadden (1972) and by Kemp and Wan (1972) that, for a single country, free trade coupled with a suitable scheme of compensatory lump sum transfers confined to that country would leave each resident of the country better off than in autarchy. Both demonstrations were conducted under assumptions of a type made familiar by Arrow and Debreu (1954) and McKenzie (1954), building on Walras (1874). Moreover, since 1972, the proposition has benefited from a considerable weakening of assumptions. It is now known to accommodate incomplete markets, symmetrical cash-in-advance monetary economies, chaotic trading equilibria and trade-dependent preferences and technologies; it is also known that, subject to existence, the proposition accommodates non-convex production sets and the associated oligopolistic competition. On the other hand, all contributions to date have been based on the conventional neoclassical model of household behaviour. In that model, consumption is constrained by household preferences and by a single financial budget. Thus the model neglects the fact that all consumption takes time and that each household is subject to a second time budget of twenty-four hours a day. The importance of the time constraint was first emphasized by Gossen (1854). Gossen’s contribution was virtually ignored during his lifetime but, a generation later, was acclaimed by Edgeworth (1896) and Pantaleoni (1889). Even
earlier, Jevons (1879) and Walras (1885) had warmly praised Gossen’s work, but without fully appreciating the central importance of Gossen’s focus on the constraint of time. More recently, Gossen’s time constraint has been discussed, in an appreciative and illuminating way but primarily in a context of closed economies, by Georgescu-Roegen (1983, 1985) and by Steedman (2001).2 Closely related to the 1972 propositions is the so-called Kemp-Wan proposition concerning the possibility of forming Pareto-improving customs unions; see Kemp (1964, p. 176), Vanek (1965), Kemp and Wan (1976, 1986) and Ohyama (1972). Like the 1972 propositions, the Kemp-Wan proposition was established under assumptions of the Walras-Arrow-Debreu-McKenzie type. However, as in the case of the 1972 propositions, it has been possible to relax those assumptions and accommodate incomplete markets, symmetrical cash-in-advance monetary economies, increasing returns to scale and oligopolistic competition; see Kemp (1995, chs 5 and 7) and Kemp (2001, ch. 20). Given the current popularity of preferential trading arrangements, the KempWan proposition is of special interest to policy makers. Indeed, to meet the immediate needs of policy makers, the proposition has recently been extended to accommodate free trade associations that are not also customs unions and nonmember countries that adjust their tariffs in response to the formation of free trade associations; see Kemp (2007a) and Kemp and Shimomura (2001a), respectively. It will be noted that both the Kemp-Wan proposition and the subsequent generalizations of that result can also accommodate Gossen’s time constraint. The final section of this chapter contains cautionary remarks concerning the propositions put forward in the previous section. Specifically, it is emphasized that all propositions have been derived from finite general-equilibrium models of the Walras-Arrow-Debreu-McKenzie type and that those models are internally consistent only if households and firms are endowed with a degree of ignorance or irrationality, or both.
