The theoretical result that there are gains from trade is a central tenet of international economics. Assuming perfect competition and no market failures, trade acts like a technological improvement that expands the set of feasible allocations and enables Pareto superior outcomes to be achieved. A recent body of research has sought to quantify the magnitude of these welfare gains. In a class of standard trade models that feature a constant elasticity of bilateral trade with respect to bilateral trade costs, the welfare gains from trade can be computed using only the economy’s domestic trade share and the elasticity of trade with respect to variable trade costs (see Arkolakis, Costinot and Rodriguez-Clare 2012). One of the main findings from this literature is that the welfare gains from trade are relatively modest. For example, in a study of 19 OECD countries, Eaton and Kortum (2002) find that the welfare cost of moving to autarky ranges from 0.2-10.3%.
A number of extensions to this quantitative approach have been considered, including the introduction of input-output linkages (e.g. Caliendo and Parro 2012) and multiple sectors (e.g. Ossa 2012). Some of these extensions can increase the predicted welfare gains from trade, as surveyed in Costinot and Rodriguez-Clare (2013), which reports a range of potential values for the welfare gains from trade.
Trade-induced changes in domestic productivity
In our research (Melitz and Redding 2014), we suggest a channel for welfare gains that this standard quantitative approach typically abstracts from, i.e. trade-induced changes in domestic productivity. Domestic productivity directly affects welfare in both the closed and open economy (separately from trade flows and relative prices). Therefore it provides an additional potential channel for trade to affect welfare. We develop a simple model featuring sequential production along a supply chain that highlights how trade endogenously increases domestic productivity through a more efficient organisation of production within the supply chain. We show that the contribution of this additional channel to the overall welfare gains from trade can be large.
In our model, the world consists of many countries that can differ in terms of their bilateral trade costs, size and productivities. Each country consumes a single final good that is produced using a sequence of traded intermediate inputs indexed by their stage of production s from 1 to S. The final good is the output from stage of production S and is produced using the intermediate input from stage S-1. The intermediate input from stage S-1 is produced using the intermediate input from stage S-2, and so on. The intermediate input from stage 1 is produced using a primary input that is manufactured from labour. Each stage of production must be completed for the final good to be produced.
In this setting, opening the economy to trade generates welfare gains from trade at each stage of production. If one falsely assumes a single stage of production, when production is in fact sequential, these gains from trade at each intermediate stage show up as an endogenous increase in measured domestic productivity. The gains from trade can be arbitrarily large as the number of production stages increases. This captures the idea that trade involves myriad changes in the organisation of production throughout the economy and the welfare costs from forgoing this pervasive specialisation can be large.
Holding the length of the supply chain fixed, the gains from trade also become arbitrarily large as differences in technological capabilities across countries for any given stage of production are magnified. This captures the idea that some countries may have strong comparative advantages in some stages of production and the welfare losses from forgoing this specialisation can be large. This result for sequential production has similarities and differences with Ossa (2012)’s result in a multi-sector model that the presence of sectors with low trade elasticity’s can generate large aggregate welfare gains from trade. In contrast, our result holds even if all production stages have the same trade elasticity, because each production stage has to be completed for the final good to be produced.
One general conclusion from these findings is that the abstraction from endogenous changes in domestic productivity can result in substantial underestimates of the welfare gains from trade. These welfare gains are distinct from any dynamic gains that would arise due to a spillover effect from trade to technology. Another more specific conclusion relates to the fragmentation of stages of production across national borders as a source of endogenous changes in measured domestic productivity. This international fragmentation of production is a distinctive feature of contemporary globalisation relative to previous eras of globalisation, such as that at the end of the nineteenth century. Existing research has explored the implications of this fragmentation for a variety of issues, including the growth of world trade and the border effect (e.g. Yi 1997). Our research highlights that it also has important implications for the aggregate welfare gains from trade.
Arkolakis, Costas, Arnaud Costinot and Andres Rodriguez-Clare (2012) “New Trade Models, Same Old Gains,” The American Economic Review, 102(1), 94-130.
Caliendo, Lorenzo and Fernando Parro (2012) “Estimates of the Trade and Welfare Effects of NAFTA,” NBER Working Paper, 18508.
Costinot, Arnaud and Andres Rodriguez-Clare (2013) “Trade Theory with Numbers: Quantifying the Consequences of Globalization,” Handbook of International Economics, Volume 4, forthcoming.
Eaton, Jonathan and Kortum, Samuel (2002) “Technology, Geography, and Trade,” Econometrica, 70(5), 1741-1779.
Melitz, Marc J and Stephen J Redding (2014) Missing Gains from Trade? CEPR Discussion Paper, 9821.
Yi, Kei-Mu (2003) “Can Vertical Specialization Explain the Growth of World Trade?” Journal of Political Economy, 111(1), 52-102.
Yi, Kei-Mu (2010) “Can Multistage Production Explain the Home Bias in Trade?” The American Economic Review, 100(1), 364-393.