It is well understood that a weak dollar is the natural consequence of the long string of current account deficits run by the US in the past. Already in the year 2000, leading economists such as Obstfeld and Rogoff had warned about the need for real dollar depreciation in the range of 35% or 50%, on a multilateral basis (see e.g. Obstfeld and Rogoff 2005). Since then, from its peak in 2002 and October 2007, on a multilateral basis the dollar has fallen approximately 28% in nominal terms, about 25% in real terms.
The big questions are: How much real dollar depreciation should be expected, and, most important, to what extent the dollar fall will be accompanied by a global realignment of Asian currencies, supposedly lifting the pressure on the euro? These are important questions, yet the strong interest in exchange rate forecasts (by the way, perhaps a hopeless quest) should not overshadow the need to understand the specific mechanisms by which real dollar depreciation is an essential step towards global adjustment. After all, it is these mechanisms that will shape the macroeconomic outlook in the next few years.
The ABCs of USD depreciation: terms of trade vs. domestic price adjustment
The dollar exchange rate in real terms is defined as the price of US consumption relative to consumption abroad. It is usually measured by multiplying the nominal exchange rate by the ratio of the consumer price indices (CPI’s), abroad and in the US. While the nominal and the real exchange rate move closely together in the data, broadly speaking changes in the real exchange rate can be decomposed into changes in the price of US imports in terms of US exports – usually referred to as the terms of trade – and the domestic relative price of US goods that are typically not traded across the border (the US nontradables), in terms of goods that are traded across the border (the tradables).
Now, closing the US current account deficit may require both types of relative prices to change. US goods which are exported overseas must get cheaper than their foreign rivals to boost US exports – this is a terms of trade adjustment. But also inside the US, the price of goods which are typically not exported (nontradables) may need to fall relative to the price of tradables in order to induce US consumers to shift their demand away from imports – this would be an adjustment in domestic relative prices.
In their well-known work, Obstfeld and Rogoff (2005, 2007) propose the following scenario. Closing the US external deficit to within 5% of the US GDP will require the US terms of trade to fall between 5% and 15% – a surprisingly contained movement. By contrast, the fall in the internal relative price should be 3 to 5 times larger, namely the relative price of nontradeables inside the US must get between 20% and 30% cheaper.
To translate these figures into the current macroeconomic stance, keep in mind that, over time, productivity growth is faster in manufacturing (producing most tradables) than in services (mostly nontradables). These productivity differentials across sectors mean that the price of manufacturing decline steadily in terms of services. Now, relative to these long-run trends, we should see the price of US services drop by about one third in terms of US manufacturing, as the US eliminates their current account deficit. While there are other possible scenarios for adjustment (for instance, the nontraded goods prices could adjust substantially in the rest of the world, rather than in the US), these considerations raise an interesting issue.
Is a sizeable change in internal prices likely to happen in the US?
Let’s consider some evidence. An episode of major dollar depreciation and current account adjustment occurred in the 1980s. The dollar started to depreciate in 1985, followed with some delay by a moderate improvement in the US current account. The evidence during this year is surprisingly at odds with the calculations above: movements in the internal relative price of nontraded goods were actually quite small. The relative price did not fall significantly and certainly less than the terms of trade.
To wit: in 3-year period from 1985 to 1988, the dollar depreciated by about 35% , as opposed to a cumulative appreciation as high as 20 percent in the previous three years. Over the whole period, the internal relative price of tradables to nontradeables (as proxied by the ratio of the producer price index, PPI, to the consumer price index, CPI, services) fell steadily. Actually the fall is faster after 1985 than before, when the dollar was appreciating (9% vs. 7%, respectively). Obviously, cyclical considerations may have heavily influenced these numbers. Yet, one may safely conclude that there is no strong evidence of a sizeable internal depreciation of nontradables.
In this respect, another way to make the same point is to stress that the real exchange rate remained strongly correlated with the US terms of trade over the adjustment period. In the period 1985-1987, the US terms of trade (based on export deflators) deteriorated by about 40% against OECD partners.
So, the evidence of the 1980s suggests that closing the US current account imbalance could take place without strong movements in internal prices. I should stress here that this evidence is still consistent with the theory underlying Obstfeld and Rogoff calculations, although it requires a modification of the specific features of their model. But discussing this point will take us away from the main message.
It is possible that in the next few years we will witness some effects of dollar depreciation on the relative rate of inflation in services and manufacturing, with the former falling somewhat behind the latter. Yet it would be highly surprising that these sectoral inflation differentials would change drastically relative to current trends.
The US terms of trade in the medium run
Without strong internal relative prices changes, the adjustment will fall on US export prices, hence directly affecting European and world firms (the US exports will become more competitive), but also European and world consumers (US goods will become cheaper). By how much? Ultimately (i.e. after short-run business cycle fluctuations or any more worrisome yet possible deterioration of the credit market are settled down), the fall in these prices could be sizeable, but not necessarily as large as 50 percent.
Results from numerical exercises developed in joint work with Martin, and Pesenti, suggests that closing the US current account deficit (from 5% of GDP to zero) could lead to a combination of lower US consumption (-6%), and higher US employment (+3%), relative to trend. This would then correspond to a rate of real dollar depreciation of the order of 20% – close to what we have experienced so far. Moreover, because of entry and exit of new firms and product varieties in the export market over time, the rate of dollar real depreciation could actually be smaller than 20% (even substantially smaller, see Corsetti Martin and Pesenti 2007), although the adjustment in consumption and employment is likely to remain of the same magnitude as above (-6% and +3% respectively).
I should stress here that these figures are clearly not a forecast. In the next few months, the world is likely to witness substantial swings in the currency market. Over time, however, what matters for the US is to achieve a sustainable external balance. A drop in relative prices of about 20% may well be all is needed.
Corsetti G., P. Martin and P. Pesenti (2007). “Varieties and the transfer problem: the extensive margin of current account adjustment”. European University Institute, mimeo
Obstfeld, M. and K. Rogoff (2005). "Global Current Account Imbalances and Exchange Rate Adjustments", Brookings Papers on Economics 1, 67-123
Obstfeld, M. and K. Rogoff, 2007, "The Unsustainable US Current Account Position Revisited," in R. Clarida (ed.), G7 Current Account Imbalances: Sustainability and Adjustment, Chicago, IL: University of Chicago Press, forthcoming.