[Journal:] Swiss Journal of Economics and Statistics [ISSN:] 2235-6282 [Volume:] 145 [Year:] 2009 [Issue:] 4 [Pages:] 387-403
The classical Armington (1969) assumption that consumption is differentiated by the location of production is often needed to match aggregate trade patterns. Also more recent findings that there are pronounced systematic patterns in the quality composition of production, trade, and consumption or that intra-industry trade volume can best be explained by factors that are specific to country-pairs are not easily rationalized in frameworks based on the Dixit and Stiglitz preference framework. It is worthwhile to examine which underlying preference structure of rational agents can explain these patterns of trade and to then analyze how trade liberalization affects the aggregate economy once these preference are properly modeled. For example, as famously conjectured by Linder (1961), it is likely that "[t]he more similar is the demand structure of two countries, the more intensive, potentially, is the trade between these two countries". Implicit in Linder's hypothesis is the argument that domestic firms tend to produce goods that are optimized for the local taste and less for the taste of foreign consumers. Consequently, differences in the tastes and products impede trade and reduce the welfare gains from trade liberalization.