Intertemporal Trade SBE-SES 0820439 Eaton and 0820338 Kortum (collaborative research)
Over the last half century the world economy has become more integrated in a number of dimensions: International trade has grown much faster than world output. Foreign assets constitute a growing share of portfolios. Countries rely more on international pools of saving to finance investment and government deficits. Because data on imports and exports are readily available and can be analyzed in a static framework, the first form of integration has been analyzed in greatest depth. In previous work the investigators developed a quantitative general equilibrium model of international trade that can incorporate substantial detail about individual countries and the geographic barriers that separate them. Because complex dynamic considerations are involved, investigation into financial integration remains much more primitive. Models have only two or three countries whose representation is very stylized. Hence they are incapable of incorporating the rich variety displayed in the data. This research builds a rich multicountry dynamic framework that integrates time into the static framework that the investigators developed in their previous work. The research has three components. The first incorporates trade imbalances to assess the impact of current account reversals on exchange rates, welfare, and income distribution within and across countries. A key policy question that the research addresses is the impact that a likely reversal of the U.S. current account deficit will have in the United States and abroad. The outcome reflects (i) adjustment in the set of goods that are traded (the extensive margin), (ii) the mobility of labor and other factors of production between traded and nontraded activities, and (iii) the role of geography in adjustment. With little ability to adjust, the relative GDP changes across countries associated with achieving current account rebalancing are large. Because of the pervasiveness of nontradability, the effect of exchange rate changes on standards of living is more modest. The second component introduces intertemporal maximization. The static framework is extended to allow for differences in factor endowments, technology, and trade barriers over time as well as across countries. Countries trade intertemporally to accommodate shocks. But this trade is costly in terms of the resources absorbed by geographic barriers. Calculations indicate that trade barriers of the magnitude implied by static gravity models impede consumption smoothing by amounts consistent with that observed in the data. The third component examines investment as well as consumption. Results indicate that geographic barriers can explain the high (but declining) correlation of savings and investment. The analysis provides a deeper understanding of the determinants and consequences of long-run swings in countries' trade balances. This research furthers economists' ability to model quantitatively a dynamic, multi-country world. It connects two literatures, the static modeling of trade flows and dynamic open-economy macroeconomics. The framework it develops can be used to address a host of key issues in international economic policy.
The major component of the research conducted under this grant is the analysis of the collapse in international trade which occurred during the Great Recession, roughly the period 2008-2009. This work, conducted with Samuel Kortum, Brent Neiman, and John Romalis, all at the University of Chicago, adapted a model of international trade, developed previously by Eaton and Kortum, to decompose changes in trade between individual countries into shocks due to the cost of trade itself, shocks to demand for different types of goods, shocks to productivity, and changes in trade deficits. Key ingredients for the model are trade and production data in durable and nondurable manufactures and nonmanufactures. Interindustry linkages among these three sectors play a major role. For the world it found that changes in the composition of final demand were the major source of the collapse, although shocks to trade costs appear to have played a role for some countries' trade, particularly China's. The analysis also shows how trade and production data can be used to infer the extent to which the international trading system propagates shocks emanating from different countries worldwide. We find, for example, that shocks from abroad played a major role in the decline in industrial production in the United States and China. Comparison with data from the trade collapse during the Great Depression suggest that increased trade barriers were a much more important source in the earlier period. Results from the research appeared in Vox while the main findings of the research are presented in a paper "Trade in the Great Recession" which is under revision for the American Economic Review. A second component of the research undertaken under this grant, done jointly with Costas Arkolakis of Yale University and Sam Kortum of the University of Chicago, is the modeling of trade dynamics. In the long run trade seems extremely responsive to changes in barriers such as tariffs and quotas, but the short-run response to high frequency changes such as exchange rates seems minimal or perverse in sign. While this phenomenon is well-known, incorporating these dynamics into a standard gravity model of international trade is challenging. The investigators are pursuing several modeling strategies to accommodate these effects. The results appear in a working paper "Staggered Adjustments and Trade Dynamics." Work on this project is ongoing. A third component of the research, joint with Kortum and Francis Kramarz of CREST INSEE, was the completion of a project on the export activity of French manufacturing firms. Access to customs and tax records of French firms enabled the researchers to construct a firm-level dataset showing patterns of where and how much firms export and how much they sell at home. A number of regularities emerge: Larger markets attract more firms (with an elasticity of around two thirds) and more sales per firm (with an elasticity of around one third). Firms that export more widely and to less popular markets systematically sell more in France as well. The distribution of sales across markets of different sizes is very similar in shape. The investigators adapt a standard model of firm export behavior to explain these features and estimate its parameters to match moments from this dataset. The results show the extent to which trade liberalization raises sales by firms in the top 10 percentile while average sales of smaller firms decline. Many small firms disappear. The results appear in the article "An Anatomy of International Trade: Evidence from French Firms" which appeared in Economtrica in September 2011.
