A large empirical literature finds that there is too little international trade and too much intranational trade to be rationalized by observed international trade costs, such as tariffs and transport costs. This paper investigates whether a model in which the nature of production can change in response to trade costs—a framework with multistage production—can better explain the home bias in trade. The calibrated model can explain about two-fifths of the Canada border effect, about two-and-one-half times that of a model with one production stage. The model also explains a significant fraction of Canada-US “back-and-forth,” or vertical specialization, trade.

The striking growth in the trade share of output is one of the most important developments in the world economy since World War II. Two features of this growth present challenges to the standard trade models. First, the growth is generally thought to have been generated by falling tariff barriers worldwide. But tariff barriers have decreased by only about 11 percentage points since the early 1960s; the standard models cannot explain the growth of trade without assuming counterfactually large elasticities of substitution between goods. Second, tariff declines were much larger prior to the mid 1980s than after, and yet, trade growth was smaller in the earlier period than in the later period. The standard models have difficulty generating this nonlinear feature. This paper develops a two-country dynamic Ricardian trade model that offers a resolution of these two puzzles. The key idea embedded in this model is vertical specialization, which occurs when countries specialize only in particular stages of a good's production sequence. The model generates a nonlinear trade response to tariff reductions and can explain over 50 percent of the growth of trade. Finally, the model has important implications for the gains from trade.

International Trade and Business Cycles: Is Vertical Specialization the Missing Link? by M. Ayhan Kose and Kei-Mu Yi. Published in volume 91, issue 2, pages 371-375 of American Economic Review, May 2001

Narayana R. Kocherlakota, Kei-Mu Yi, A Simple Time Series Test of Endogenous vs. Exogenous Growth Models: An Application to the United States, The Review of Economics and Statistics, Vol. 78, No. 1 (Feb., 1996), pp. 126-134

A common view is that cross-border vertical linkages played a key role in the 2008–2009 collapse of global trade. This paper presents two accounting results from a global input-output framework that shed light on this channel. We feed in observed changes in final demand and find that trade in final goods fell by twice as much as trade in intermediate goods. Nevertheless, intermediate goods account for more than two-fifths of the trade collapse. We also find that vertical specialization trade fell 13 percent, while value-added trade fell by 10 percent, because declines in demand were largest in highly vertically-specialized sectors.