Do Japanese Firms Price Discriminate in North America?

INCE the early 1980s, there has been a continuing, energetic and 45 sophisticated effort to assess the economic consequences of the Canadian-US Free Trade Agreement (CUSTA) and the North American Free Trade Agreement (NAFTA). ' This work has been conducted both in an academic setting, in government and in private sector consulting firms. The most interesting of these studies have employed computable general equilibrium (CGE) models to characterise the US, Canadian and Mexican economies and their interrelationships. The dimensions specified in these models typically include a number of industrial sectors, the form of technology, demand, the clearing of factor markets and goods markets, external relationships and trade policy. Most of these studies find that CUSTA and NAFTA, while having little effect on the United States, can have a significant impact on the structure and operation of the Canadian and Mexican economies. While much of the CGE research agrees as to conclusions, these findings rest very heavily on how technology, productivity, and most importantly, on how the behaviour of capital is modelled. By any reasonable criterion, much of the modelling in these areas is crudely done.2 Moreover, by the standards of sectoral research, these models are highly aggregated and non-tariff trade policy, in particular, is treated at an extremely rudimentary level. It has not been generally recognised that another tradition of research in international economics might be helpful for evaluating the consequences of trade liberalisation and FTAs. The slow pace of adjustment of exchange rate changes in the 1970s and again in the midand late 1980s triggered substantial research efforts which continue today. In empirical work explaining the slow response of