VoxEU Column International trade

The collapse of US trade: In search of the smoking gun

US trade has experienced an unexpectedly large drop – seven standard deviations more than that predicted by theory. We evaluate three leading hypotheses on its causes: the vertical linkages effect, the compositional effect, and the credit effect. Using highly disaggregated US trade and production data, we show that between 50% and 100% of the drop is due to a “compositional effect”, i.e. that trade fell systematically more in sectors that also experienced larger domestic output reductions. The trade drop was also particularly concentrated in sectors marked by strong vertical linkages. We find no evidence that US trade was significantly hindered by trade credit problems.

A remarkable feature of the recent crisis is the collapse in international trade. This collapse is global in nature and dramatic in magnitude (WTO 2009).

Figure 1 US trade to GDP ratio, 1947-2009.

Note: Shaded bars show recessions.

Source: National Income and Product Accounts.

During the crisis, US GDP declined by 4% from its peak, while its real imports fell by 19% and real exports fell by 15%. This is unprecedented in the postwar period, as Figure 1 shows. Though protectionist pressures inevitably increased over the course of the recent crisis, it is widely believed that the collapse is not due to newly erected trade barriers (Baldwin and Evenett, 2009).1

Is the trade collapse a puzzle?

Though the reduction in international trade flows has been drastic, it is not a priori clear that it is in any sense exceptional or represents a puzzle. Trade tends to decline during global recessions. The key question is:

  • Is this trade drop in some sense abnormal?

To evaluate this question, we need a benchmark to establish “normal”. To this end, we follow a simple “wedges” methodology – an approach that is standard in open economy macroeconomics (Cole and Ohanian 2002 and Chari, Kehoe, and McGrattan 2007).

We derive a simple import demand equation, where imports depend upon total domestic demand (absorption) and relative prices (import and domestic prices). The “wedge” is the extent to which actual imports deviate from what the import demand equation predicts. (Details can be found in our underlying research – Levchenko, Lewis, and Tesar 2009 – which was prepared for the IMF Economic Review special issue, “Economic Linkages, Spillovers and the Financial Crisis.”)

Figure 2 depicts the wedge’s evolution from 1984 to the second quarter of 2009. Two features of this series stand out:

  • The current value of the wedge exhibits a drastic deviation from the norm; by the second quarter of 2009 it reached 54%.
  • Compared to the historical experience this is indeed exceptional; over the past 25 years the mean value of the wedge is less than 9%, with a standard deviation of 8.7%.
  • Since the current value of the wedge is seven standard deviations away from the mean and six standard deviations away from zero, we can say that the recent trade collapse is a puzzle.

In short the collapse in trade is abnormal; it is well in excess of what the pace of economic activity and prices would predict.

Figure 2 The international trade wedge.

Source: Levchenko, Lewis, and Tesar (2009).

Why did trade collapse? Three leading suspects

We focus on the three main hypotheses that have been mooted:2

  1. Trade is collapsing because of the transmission of shocks through vertical production linkages.
  2. The collapse in trade is due to compositional effects.
  3. Trade is collapsing because of the contraction in credit.

To confront this hypothesis with data, we compare the behaviour of disaggregated trade data with characteristics of the sectors concerned. Specifically, we look at US exports and imports at the 6-digit NAICS level of disaggregation (about 450 distinct sectors) and see the extent to which the differential export performance can be explained by sector-specific characteristics that proxy for the three main hypotheses.

The supply-chain/vertical linkages hypothesis

When demand for final goods drops, the demand for intermediate inputs suffer. In sectors marked by extensive international supply chains, this logic means that the value of trade drops more than final demand; a dollar drop in imported final-goods purchases can lead to more than a one dollar drop in total trade.3 To test for this possibility, we build several measures of intermediate input linkages at the detailed sector level based on the US Input-Output tables.

  • We find strong evidence in favour of the vertical linkages explanation. After controlling for a variety of other industry characteristics, trade fell systematically more in sectors that are used intensively as intermediate inputs.
Compositional effects hypothesis

If international trade occurs disproportionately in sectors whose domestic absorption (or production) collapsed the most, we should expect trade to fall more than GDP.4 To explore this, we collect measures of US industrial production at the most disaggregated level available, and correlate them with reductions in trade.

  • We find strong evidence of compositional effects. As it turns out, trade in this crisis fell systematically more in sectors that also experienced larger domestic output reductions.
The credit crunch hypothesis

Firms need credit – capitalism without capital doesn’t work. As the global crisis – especially the period that coincidences with the steepest decline in trade – is marked by a massive and global credit crunch, it is natural suppose that the lack of credit contributed to the great trade collapse. Indeed there is evidence of this in past crises. Raddatz (2009) shows that there is greater co-movement between sectors that have stronger trade credit links, while Iacovone and Zavacka (2009) demonstrate that in countries experiencing banking crises, exports fell systematically more in financially dependent industries. Amiti and Weinstein (2009) show that exports by Japanese firms in the 1990s declined when the bank commonly recognized as providing trade finance to the firm was in distress.

To evaluate whether this channel mattered in the current crisis, we use US firm-level data from the COMPUSTAT database to construct measures of the intensity of trade credit use in each sector. Taking as given that credit became scarce at the end of 2008 and first half of 2009, we should expect to see trade falling particularly much in sectors that rely particularly heavily on trade credit.

  • By contrast to the previous two explanations, we find no evidence that trade credit played an independent role in the trade collapse. Sectors that receive, or extend more trade credit did not experience systematically larger reductions in either imports or exports during the current episode.

We can also examine the time evolution of trade credit directly.

Figure 3 depicts the evolution of the most standard measure of trade credit extended to firms, i.e. ‘Accounts Receivable’ relative to the ‘Cost of Goods Sold’ for the firms in COMPUSTAT data. The dashed line represents the raw series. As there is substantial seasonality in the raw series, the solid black line represents this following seasonal adjustment. The horizontal line plots the mean value of this variable over the entire period.

Figure 3 The evolution of US trade credit.

Source: Levchenko, Lewis, and Tesar (2009).

While there is indeed a contraction in trade credit during the recent crisis, its magnitude is very small. It was 55% in the first quarter of 2009, just 1% below the period average of 57%, and only 3 percentage points below the most recent peak of 58% in the fourth quarter of 2007.

In summary, the typical firm in the COMPUSTAT data experienced at most a small contraction in trade credit.

How much does each factor matter?

We can push our reasoning further and assess how far each effects accounts for the trade collapse. As it turns out, the compositional effect is by far the most important. We compare percentage reductions in exports and imports that would be expected if compositional effects were the only effect in operation (i.e. the reduction in trade that should have occurred if exports and imports fell by the exact same percentage as domestic industrial production).

  • Our work suggests that compositional effects account for between 50% and 100% of the reduction in US trade, depending on the details of how the measure is constructed.

A number of caveats should however be considered in order to interpret the results. Most importantly, this is an accounting exercise rather than an economic explanation.
We do not know why trade flows are systematically biased towards sectors that experienced larger output reductions, nor do we know why some sectors experienced larger output drops than others. Nonetheless, this exercise suggests further evidence of compositional effects.

Summary and implications

Using a theory-based benchmark, we demonstrate that the US cross-border trade has indeed experienced a significant disruption: it fell much more than what would be expected given the observed reduction in aggregate demand. We also find that part of the collapse can be accounted for by compositional effects and vertical linkages.

The result that the state of international trade today is far from “business as usual” underscores the importance of resisting protectionist pressures (that are also on the rise in the current crisis). Going forward, we simply do not know how long it will take for international trade to recover. Policymakers should therefore be especially careful not to impede this recovery process further.


1 For detailed descriptions of the various features of the current collapse in trade, see Francois and Woerz (2009) and Ferrantino and Larsen (2009).

2 Another possibility is a rise in trade barriers (see, e.g. Campbell, Jacks, Meissner, and Novy, 2009). However, so far there is no evidence that countries are adopting protectionist measures severe enough to generate a reduction of this magnitude. In addition, actual shipping costs have plummeted, following a collapse in oil prices and a drop in demand for shipping (Economist 2009).

3 Hummels, Ishii, and Yi (2001) and Yi (2003) document the dramatic growth in vertical trade in recent decades, and di Giovanni and Levchenko (2009) demonstrate that greater sector-level vertical linkages play a role in the transmission of shocks between countries.

4 Two special cases of the compositional story are investment goods (Boileau, 1999, Erceg, Guerrieri, and Gust, 2008) and durable goods (Engel and Wang 2009). Since investment and durables consumption are several times more volatile than GDP, trade in investment and durable goods would be expected to experience larger swings than GDP as well.


Amiti, Mary and David E. Weinstein, “Exports and Financial Shocks,” September 2009. mimeo, Federal Reserve Bank of New York and Columbia University.

Baldwin, Richard and Simon Evenett, “Introduction and Recommendations for the G20,” in Richard Baldwin and Simon Evenett, eds., The Collapse of Global Trade, Murky Protectionism, and the Crisis: Recommendations for the G20, London: CEPR, 2009, chapter 1, pp. 1–9.

Boileau, Martin, “Trade in capital goods and the volatility of net exports and the terms of trade,” Journal of International Economics, August 1999, 48 (2), 347–365.

Campbell, Douglas L., David Jacks, Christopher M. Meissner, and Dennis Novy “Explaining two trade busts: Output vs. trade costs in the Great Depression and today,” VoxEU.org, 19 September 2009.

di Giovanni, Julian and Andrei A. Levchenko, “Putting the Parts Together: Trade, Vertical Linkages, and Business Cycle Comovement,” IMF Working Paper 09/181, July 2009, Forthcoming, American Economic Journal: Macroeconomics.

Economist, “Shipping in the Downturn: Sea of Troubles,” June 30 2009.

Engel, Charles and Jian Wang, “International Trade in Durable Goods: Understanding Volatility, Cyclicality, and Elasticities,” July 2009. mimeo, University of Wisconsin and Federal Reserve Bank of Dallas.

Erceg, Christopher J., Luca Guerrieri, and Christopher Gust, “Trade adjustment and the composition of trade,” Journal of Economic Dynamics and Control, August 2008, 32 (8), 2622–2650.

Ferrantino, Michael J. and Amy Larsen “Transmission of the global recession through US trade”, VoxEU.org, 29 August 2009.

Francois, Joseph, and Julia Woerz, “The big drop: Trade and the Great Recession” VoxEU.org, 2 May 2009.

Hummels, David, Jun Ishii, and Kei-Mu Yi, “The Nature and Growth of Vertical Specialization in World Trade,” Journal of International Economics, June 2001, 54, 75–96.

Iacovone, Leonardo and Veronika Zavacka, “Banking Crises and Exports: Lessons from the Past,” May 2009. World Bank Policy Research Working Paper 5016.

Levchenko, Andrei A., Logan Lewis, and Linda L. Tesar, “The Collapse of International Trade During the 2008-2009 Crisis: In Search of the Smoking Gun,” October 2009. Prepared for the IMF Economic Review special issue, “Economic Linkages, Spillovers and the Financial Crisis.”

Raddatz, Claudio, “Credit Chains and Sectoral Comovement: Does the Use of Trade Credit Amplify Sectoral Shocks?,” 2009. Forthcoming, Review of Economics and Statistics.

WTO, “World Trade Report,” July 2009.

Yi, Kei-Mu, “Can Vertical Specialization Explain the Growth of World Trade?,” Journal of Political Economy, February 2003, 111 (1), 52–102.

1,365 Reads