The global crisis was accompanied by a severe fall in trade, characterised by Baldwin (2009) as “sudden, severe, and synchronised… the sharpest in recorded history and the deepest since WWII.”
Trade can reduce across one of two lines – the intensive margin or the extensive margin. The speed and durability of the recovery of international trade could depend on which of these two margins has been the most affected by the crisis.
- The intensive margin of trade refers to changes in values of goods already being traded.
- The extensive margin refers to changes in the number of goods exported and changes in the number of destinations to which a country exports goods.
If there are significant fixed costs associated with exporting or importing new products (i.e., the extensive margin), then identifying the relative size of changes along the extensive and intensive margins is useful in predicting the speed of the recovery. Evidence from US and French firms (Peter Schott 2009 and Bricongne et al. 2009, respectively) suggests that the intensive margin was more affected during this crisis than the extensive margin. These results are consistent with Bernard et al. (2009), who analyse past recessionary periods. Yet none of these studies have examined price changes.
Demand shock or supply shock?
One key question is to what extent the collapse in world trade was driven primarily by demand-side shocks or supply-side shocks.
- Supporters of the demand shock hypothesis argue that the fall in trade has been the result of a postponement of purchases, especially of durable consumer and investment products. Eaton et al. (2010) and Levchenko et al. (2010) are among those arguing that the collapse in trade was primarily the result of demand-side shocks.
- Supporters of the supply shock hypothesis suggest that the collapse in trade has been a consequence of the sudden financial arrest, which froze global credit markets and spilled over on the specialised financial instruments that finance international trade.
As the trade fell, was it from falling prices, quantities, or both?
When trying to decide whether a particular quantity shock came from shifts in supply curve or in the demand curve, the price is a useful indicator. We would expect that if the decline in trade were mostly driven by a negative demand shock, then both prices and quantities would be negatively affected. However, if supply-side shocks were important (for instance, a reduction in credit), then we would have expected less downward pressure, and possibly some upward pressure, on prices.
Our latest research (Haddad et al. 2010) follows this intuition by separating changes in trade flows into traded prices and quantities. We decompose the fall in international trade into product entry and exit, price changes, and quantity changes for imports by Brazil, the EU, Indonesia, and the US. (All effects are defined relative to the same quarter of the previous year, to avoid issues with seasonally traded goods.)
We find that imports by the US and the EU fell 25% from $4 trillion in 2008 to $3 trillion in 2009. Imports by Brazil and Indonesia dropped nearly 20% from $135 billion to $109 billion. (Here and elsewhere, we report figures in real US dollars, deflated by each importer’s monthly Consumer Price Index.) Figure 1 shows the decomposition of the change in total import value along each margin, where each margin is shown as a percentage of the total value in 2008.
Figure 1. Change in total import value by margin
As this figure shows, the effects along the intensive margin (quantity and price) dramatically outweighed the effects along the extensive margin (exiting and new products). The reduction in quantity was large in all countries. It accounted for 15.9% out of the total 25.2% reduction in the total value change in the US and EU. The decline in prices accounted for only 5.5%. Net entry, the sum of exit and entry, was negative but accounted for only a small portion of the total value change. For Brazil and Indonesia, quantity changes accounted for 18.5% out of the total 18.9% drop in the value of trade, with product net exit accounting for 1% and a small price increase partially offsetting these effects.
The aggregate figures, however, mask enormous differences across different products. The following figures (see Haddad et al. 2010 for details) show the difference between commodities and manufactures, where each margin is again shown as a percentage of the change total import value in 2008. The differences across product classes are striking.
- The negative price effect, apparent for the US and EU when aggregating across all goods, is still evident for commodities but not for manufactures.
- With commodity prices falling during the crisis, it is not surprising that the price effect was large for commodities. It is, however, noteworthy that the price effect was generally contained to commodities, indeed the price effect for manufactures was positive.
Since we know that demand for manufactures fell during the crisis, the effect on prices can tell us something about what happened to supply. Where prices rose or where they fell only slightly, it is plausible that supply shifted in. Thus the evidence on manufactures, contrasted with commodities and particularly in the case of Brazil and Indonesia, points to a negative supply shock in manufactures in addition to the negative demand shock. This negative supply shock could be from fragmentation of the global supply chain or from reductions in trade finance.
Figure 2. Product level variations
We also examine evidence on credit constraints, by adopting the classification scheme of Bricongne et al. (2009) to separate products according to sectoral dependence on external finance. We restrict our analysis to manufactures.
- For the US, price increases were most significant in sectors which are typically credit constrained, partially counteracting the large quantity effect.
- The EU does not show this effect.
- For Brazil, the overall import value for the finance-dependent sectors dropped (by 3%) but rose for the low-dependence sectors (by 9%).
- The overall import value also fell more for finance-dependent sectors (19% as opposed to 6%) in Indonesian imports.
Were any of these trends present before the crisis?
We examine whether these findings are unique to the crisis, or whether they represent the continuation of historical trends. The following figures show changes in each margin (in billions of $) for US and Indonesian imports across quarters for 2007 to 2009. Price and quantity changes are now defined relative to the previous quarter, rather than the same quarter of the previous year. Hence the magnitudes of these changes do not match the magnitudes given in the other figures, but they do show the specific timing of the collapse in trade.
Figure 3. Changes in each margin by country
As expected, entry and exit do not play much of a role in US imports (before and after the crisis) but are more important in a developing country like Indonesia where trade relations are thinner and less established.
The quantity and price effects are not part of broader historical trends, rather they match the timing of the global economic crisis. Manufacturing imports to the US level off in the third quarter of 2008, as the crisis is beginning, then plummet in the following two quarters. For this whole period, the negative quantity effect dominates, and there is a smaller positive price effect for Q4 2008.
Manufacturing imports to Indonesia follow a similar pattern, with a negative quantity effect beginning in Q4 2008 and an initial positive price effect. Commodity imports to the US also plummet in Q4 2008, but here the negative price effect dominates. For commodity imports to Indonesian, both quantity and price effects are negative and begin around Q4 2008.
Were there differences across income groups?
While the global crisis originated in high income countries, its effects on trade were rapidly transmitted to low-income countries. It has been hypothesised that the effects of constrained trade finance could vary by exporter income (Malouche 2009, Berman and Martin 2010) and by geographic region (Berman and Martin 2010). While high-income countries with well-developed markets were most affected in the global crisis, on the other hand low-income exporters with less-developed financial markets may be more reliant on trade finance originating in their trading partners. Countries with different levels of income export different baskets of goods, which embody different levels of quality and variety.
To analyse how the response in trade volumes changes with the income level of the exporting country, we classify trading partners in four categories: high-, upper-middle-, lower-middle-, and low-income, according to the World Bank’s country classification. We also add China and Sub-Saharan Africa as separate categories. We restrict the sample to manufacturing.
- Overall, high and upper-middle-income exports to developed countries were most affected by the crisis, with falls in the value of their exports reaching 25% to the US and EU.
- Low-income countries were able to increase their exports to the US and EU by 7%.
- The impact of the crisis on the exports of countries of various income groups to Brazil and Indonesia was very small.
- The most striking result is the large increase in exports of low-income countries to Brazil and Indonesia by nearly 30% between 2008 and 2009.
- China’s exports to the US and the EU dropped by only 8%, in line with other lower-middle income countries; but increased by 5% in Brazil and Indonesia.
This confirms that south-south trade is becoming increasingly important and was reinforced during the crisis. Sub-Saharan Africa, however, was not able to take advantage of these South-South trade opportunities as its exports to Brazil and Indonesia dropped 27%.
Figure 4. Imports by Brazil-Indonesia and US-EU from various income country groups, percentage changes from 2008 to 2009
The causes of the greatest fall in international trade since World War II have yet to be fully understood, but economists are making progress in understanding the relative magnitudes of the possible factors. Initial concerns about protection have given way to a general sense that, while protection did increase somewhat, this does not explain much of the fall in trade (Bown 2009).
Evidence from a number of countries points to a relatively small fall on the extensive margin: while countries generally exported fewer products to fewer countries, much more value was lost as products that continued to be traded fell in quantity. The hope is that this will lead to a faster recovery, so long as ramping quantities back up involves fewer fixed costs than establishing new export relationships or re-establishing dropped relationships.
Finally, there is some evidence that both the demand side and the supply side mattered. With falls in real GDP, consumers postponed the purchase of durables and companies postponed the purchase of investment goods. This led to negative pressure on both the quantity of goods traded and their prices. Yet we find evidence that supply-side frictions must also have played a role, particularly in the trade of manufactures, since positive price pressure is apparent during the crisis. This may have been the result of supply-chain fragmentation; it may also have been the result of the drying up of credit. Indeed, our research finds that credit-dependent sectors in the US experienced more upward price pressure than other sectors, and that falls in overall value were more dramatic for credit-dependent sectors in Brazil and Indonesia than for other sectors.
Baldwin, Richard (ed.) (2009), The Great Trade Collapse: Causes, Consequences and Prospects, VoxEU.org Ebook, 27 November.
Bernard, Andrew B, J Bradford Jensen, Stephen J Redding, and Peter K Schott (2009), “The Margins of US Trade”, American Economic Review: Papers & Proceedings, 99(2):487-493.
Berman, Nicolas and Philippe Martin (2010), “The Vulnerability of Sub-Saharan Africa to the Financial Crisis: The Case of Trade”, CEPR Discussion Paper 7765.
Bown, Chad (2009), “The Global Resort to Antidumping, Safeguards, and Other Trade Remedies Amidst the Economic Crisis”, World Bank Policy Research Working Paper No.5051.
Bricongne, Jean-Charles, Lionel Fontagne, Guillaume Gauler, Daria Taglioni, and Vincent Vicard (2009), “Firms and the Global Crisis: French Exports in the Turmoil”, Banque de France Document de Travail No.265.
Eaton, Jonathan, Sam Kortum, Brent Neiman, and John Romalis (2010), “Trade and the Global Recession”, Working Paper.
Haddad, Mona, Ann Harrison, and Catherine Hausman (2010), “Decomposing the Great Trade Collapse: Products, Prices, and Quantities in the 2008-2009 Crisis”, NBER Working Paper 16253.
Levchenko, Andrei A, Logan T Lewis, and Linda L Tesar (2010), “The Collapse of International Trade during the 2008-2009 Crisis: In Search of the Smoking Gun”, NBER Working Paper 16006.
Malouche, Mariem (2009), “Trade and Trade Finance Developments in 14 Developing Countries Post September 2008: A World Bank Survey”, World Bank Policy Research Working Paper No. 5138.
Schott, Peter (2009), “US Trade Margins During the 2008 Crisis”, VoxEU.org, 27 November.