International Trade and Investment

Members of the NBER's International Trade and Investment Program met in Cambridge on December 7-8. Program Director Stephen J. Redding of Princeton University organized the meeting. These researchers' papers were presented and discussed:

Alberto Cavallo, Harvard University and NBER; Robert C. Feenstra, University of California, Davis and NBER; and Robert Inklaar, University of Groningen

Foreign and Domestic Trade Costs, Product Variety, and the Standard of Living Across Countries

Cavallo, Feenstra, and Inklaar extend the analysis of Arkolakis, Costinot and Rodríguez-Clare (AER, 2012) to allow for differences in domestic trade costs in addition to foreign trade costs. The domestic trade costs are measured by local transportation charges and wholesaling and retailing margins. By allowing for differences in domestic trade costs, as well as in country size, productivity and in fixed costs, the researchers are able to model both the welfare change between two equilibria and the welfare difference between two countries. They find that the extended ACR formula depends on: (a) the share of expenditure on domestic goods (reflecting in part foreign trade costs), (b) domestic trade costs, (c) the extent of product variety available to consumers. The researchers measure the extent to which differences in the cost of living between countries are explained by these terms. They find that domestic trade costs are of comparable importance to foreign trade costs and that differences in product variety are notably more important than both of these.

Rodrigo Adão, University of Chicago and NBER; Michal Kolesár, Princeton University; and Eduardo Morales, Princeton University and NBER

Shift-Share Designs: Theory and Inference (NBER Working Paper No. 24944)

Since Bartik (1991), it has become popular in empirical studies to estimate regressions in which the variable of interest is a shift-share, such as when a regional labor market outcome is regressed on a weighted average of observed sectoral shocks, using the regional sector shares as weights. In this paper, Adão, Kolesár, and Morales discuss inference in these regressions. They show that standard economic models imply that the regression residuals are likely to be correlated across regions with similar sector shares, independently of their geographic location. These correlations are ignored by inference procedures commonly used in these regressions, which can lead to severe undercoverage. In regressions studying the effect of randomly generated placebo sectoral shocks on actual labor market outcomes in U.S. commuting zones, the researchers find that a 5% level significance test based on standard errors clustered at the state level rejects the null of no effect in up to 45% of the placebo interventions. The researchers derive novel confidence intervals that correctly account for the potential correlation in the regression residuals.

Yuhei Miyauchi, Stanford University

Matching and Agglomeration: Theory and Evidence from Japanese Firm-to-Firm Trade

Why are economic activities geographically concentrated? In this paper, Miyauchi argues that increasing returns in firm-to-firm matching is an important source of agglomeration. Miyauchi opens by providing reduced-form evidence of increasing returns in matching using a panel of firm-to-firm trade data covering over a million Japanese firms. Using unexpected supplier bankruptcies as an instrument, it is shown that the new supplier matching rate upon a supplier loss increases in locations and industries when there are more alternative suppliers selling in the buyer's location, while this rate remains stable in the presence of other buyers looking for a match. Based on these findings, Miyauchi develops a new structural trade model that incorporates dynamic firm-to-firm matching across space in a standard Melitz model. In this economy, the presence of more input sellers increases input buyers' aggregate sales by improving the supplier matching rates and hence reducing their production cost, this, in turn, attracts more suppliers to sell in the location. The model is calibrated to match the reduced-form estimates, and it is shown that this type of circular causation explains 7% and 16% of spatial inequality of the firm density and the real wages in Japan, respectively. Lastly, Miyauchi analyzes policies to promote economically lagged areas, and it is found that (1) subsidies for input suppliers to sell in the target areas are much more effective in improving the welfare of these areas than subsidies to produce in these areas, and (2) improving transportation infrastructure initially decreases and then increases the welfare of the target areas.

Zhen Huo, Yale University; Andrei A. Levchenko, University of Michigan and NBER; and Nitya Pandalai-Nayar, University of Texas at Austin

The Global Business Cycle: Measurement and Transmission

Huo, Levchenko, and Pandalai-Nayar examine the role of both technology and non-technology shocks in international business cycle comovement. Using industry-level data on 30 countries and up to 28 years, they first provide estimates of utilization-adjusted TFP shocks, and an approach to infer non-technology shocks. They then set up a quantitative model calibrated to the observed international input-output and final goods trade, and use it to assess the contribution of both technology and non-technology shocks to international comovement. The researchers show that unlike the traditional Solow residual, the utilization-adjusted TFP shocks are virtually uncorrelated across countries. Transmission of TFP shocks across countries also cannot generate noticeable comovement in GDP in the sample of countries. By contrast, non-technology shocks are correlated across countries, and the model simulation with only non-technology shocks generates substantial GDP comovement. While shocks transmit across countries through production networks, the contribution of trade openness to comovement depends also on whether more or less correlated sectors get larger as the country opens to trade. The researchers conclude that in order to understand international comovement, it is essential to both model and measure non-TFP shocks in a framework with international production networks.

Dominick G. Bartelme, University of Michigan; Arnaud Costinot and Dave Donaldson, MIT and NBER; and Andrés Rodríguez-Clare, University of California, Berkeley and NBER

External Economies of Scale and Industrial Policy: A View from Trade

Bartelme, Costinot, Donaldson, and Rodríguez-Clare develop a new empirical strategy to estimate external economies of scale using trade data. Across 2-digit manufacturing sectors, their baseline estimates of scale elasticities range from 0 to 0.23 and average 0.06. They then use their estimates of external economies of scale across sectors to explore the structure and implications of optimal industrial policy. The researchers find that gains from optimal industrial policy are only around 0.1% on average across the countries in our sample -- this is small relative to the gains from optimal trade policy, which are around 0.6% on average.

Jeronimo Carballo, University of Colorado; Kyle Handley, University of Michigan; and Nuno Limão, University of Maryland and NBER

Economic and Policy Uncertainty: Export Dynamics and the Value of Agreements (NBER Working Paper No. 24368)

Carballo, Handley, and Limão examine the interaction of economic and policy uncertainty in a dynamic, heterogeneous firms model. Uncertainty about foreign income, trade protection and their interaction dampens export investment. This can be mitigated by trade agreements, which are particularly valuable in periods of increased demand volatility. The researchers use firm data to establish new facts about U.S. export dynamics in 2003-2011 and estimate the model. They find a significant role for uncertainty in explaining the trade collapse in the 2008 crisis and partial recovery in its aftermath. Consistent with the model predictions, the researchers find that the negative effects worked (1) through the extensive margin, (2) in destinations without preferential agreements with the U.S. (accounting for over half its trade) and (3) in industries with higher potential protection. U.S. exports to non-preferential markets would have been 6.5% higher under an agreement — equivalent to an 8% foreign GDP increase. These findings highlight and quantify the value of international policy commitments through agreements that mitigate uncertainty, particularly during downturns.

Pablo Fajgelbaum, University of California, Los Angeles and NBER; Pinelopi K. Goldberg, Yale University and NBER; Patrick Kennedy, University of California, Berkeley; and Amit Khandelwal, Columbia University and NBER

The Return to Protectionism: Causes and Consequences of the 2018 Trade War

Protectionism has returned. After decades of leading efforts to reduce global trade barriers, the United States enacted several waves of tariff increases on trading partners in 2018. In response, U.S. trade partners retaliated. What is the impact of tariffs on the U.S. economy? What explains the structure of protection? Fajgelbaum, Goldberg, Kennedy, and Khandelwal analyze the impacts of U.S. and retaliatory tariffs on trade, production and prices. Using tariff variation, they estimate demand and supply elasticities in a standard international trade framework. Using the estimated model, the researchers compute the general-equilibrium impacts of tariffs on real income across factors of production and U.S. regions. Fajgelbaum, Goldberg, Kennedy, and Khandelwal also recover government Pareto weights over regions and sectors that rationalize the observed pattern of tariff changes, and use them to discriminate amongst competing explanations behind the structure of protection. They detect immediate quantity responses of trade flows to U.S. and retaliatory tariffs. However, the researchers find no evidence of terms-of-trade effects, suggesting that U.S. consumers bear the full incidence of the tariffs.

Levent Celik, Higher School of Economics, Moscow; Bilgehan Karabay, RMIT University; and John McLaren, University of Virginia and NBER

Fast-Track Authority: A Hold-Up Interpretation (NBER Working Paper No. 24427)

A central institution of US trade policy is Fast-Track Authority (FT), by which Congress commits not to amend a trade agreement that is presented to it for ratification, but to subject the agreement to an up-or-down vote. Celik, Karabay, and McLaren offer a new interpretation of FT based on a hold-up problem. If the U.S. government negotiates a trade agreement with the government of a smaller economy, as the negotiations proceed, businesses in the partner economy, anticipating the opening of the US market to their goods, may make sunk investments to take advantage of the U.S. market, such as quality upgrades to meet the expectations of the demanding U.S. consumer. As a result, when the time comes for ratification of the agreement, the partner economy will be locked in to the U.S. market in a way it was not previously. At this point, if Congress is able to amend the agreement, the partner country has less bargaining power than it did ex ante, and so Congress can make changes that are adverse to the partner. As a result, if the U.S. wants to convince such a partner country to negotiate a trade deal, it must first commit not to amend the agreement ex post. In this situation, FT is Pareto-improving.

NBER Videos

National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us