American consumers are still buying goods and services from the rest of the world with gusto, despite a prolonged decline in the dollar’s value. Unfortunately it seems that such enthusiasm is not reciprocated by other countries. The result has been an increasing gap between the value of US imports and exports. The trade deficit - i.e. the difference between imports and exports - has been steadily worsening since the mid-1990s, and is now more than 5% percent of the American Gross Domestic Product (GDP). This is an unprecedented level in time of peace, and much higher than the average trade deficit since the end of World War II. Policymakers around the world are worried that the imbalance in American trade may cause instability in the world economy, leading to abrupt variations in exchange rates and sharp declines in the rate of economic growth. On top of this, trade imbalance generates political pressure towards protectionism, thus creating further problems for world economic stability.

Why did this happen and what should be done? In order to answer these questions we must start by understanding the determinants of American exports and imports of goods and services. Standard international trade theory states that a country's imports of goods and services positively depend on its income: when a country starts growing faster it needs more goods from abroad for investment purposes and, besides, a higher income stimulates the consumption of imported goods. Imports are also positively related to the strength of currency: a strong currency makes buying abroad cheaper.

Symmetrically, the export of goods and services positively depends on the income of the rest of the world, and negatively on the currency's strength. Notice that whether buying abroad is cheaper does not only depend on the exchange rate, but on the level of prices inside and outside the country too. Therefore, the empirical analysis looks at some measure of the “real exchange rate”, that is the exchange rate adjusted to take differences in the rates of inflation across countries into account.

Trade theory is pretty clear and convincing when identifying the main variables with an impact on the volume of international trade, but what is their quantitative impact? Can we identify a stable and relevant relationship between real exchange rates, GDP growth rates and the import and export of goods in the American economy? Answering these questions requires a careful econometric analysis of import and export functions, and this is the paper's main goal.

In order to do things properly certain technical issues must be considered; different price indices may be used to compute the real exchange rate, and the GDP of the rest of the world can be computed assigning different weights to countries which are important trading partner for the United States. Some general results emerge from the analysis. The most important is that imports do not appear to be stable, and have in fact fluctuated significantly over time. The opposite holds true for exports; in this case a stable relationship can be readily identified. The empirical estimates indicate that, in the long run, a 1% increase in the income of the rest of the world increases the volume of American exports by between 1.5% and 2%, while a 1% depreciation in the real exchange rate induces a 0.5% increase in exports.

What’s the problem with imports? Why is it so difficult to find a stable import function for the US economy? Delving deeper into the data, the author shows that a stable relationship can be found if we allow for a structural break in the 1990s. In other words, the econometric analysis shows that from the first quarter of 1995 the relationship changes. Once we allow for the structural break, it appears that a 1% change in US income increases US imports by more than 2%. Therefore, there is an asymmetry in the effect of income growth for exports and imports. Keeping all else equal, if the US economy were to grow at the same rate as the rest of world, US imports would grow more than exports. Closing the trade gap therefore requires a decrease in the rate of growth of the American economy or an increase in the rate of growth in the rest of the world.

The other possible mechanism for rebalancing trade, a change in the real exchange rate, does not seem to be very effective. The long-run effect on imports is statistically insignificant, while the impact on exports is statistically significant but very small. As a matter of fact, the depreciation of the dollar in the last three years has done little to improve the US trade balance.

What happened in the 1990s that changed the economics of US imports? A sector-based analysis of data shows that the increased weight of trade in computers and semiconductors goes a long way to explain it. Computers and semiconductors accounted for 3.5% of American imports in 1990; a decade later their weight was almost doubled, reaching 6%. Furthermore, the relative price of computers over this period of time decreased dramatically, which implies that aggregate price indices do little to capture the real impact of changing prices on this part of the economy. Once the momentous expansion of computers will have reached its peak, it will probably be possible to provide more stable estimates of imports for the US economy.

In the meantime, the main lesson from this analysis is that policy makers should not rely too much on movements of exchange rates to solve the trade imbalance between the United States and the rest of the world. The best hope for a soft landing is an increase in the rate of growth of world economy. Most of Asia is already growing at an impressive pace, and most of Africa seems to be somewhat stagnating at the moment. Thus, an increase in the rate of growth of Europe and Latin America seems to be the best way to solve the imbalance without creating waves of financial and economic instability.

Report Source:
"Doomed to Deficits? Aggregate U.S. Trade Flows Re-Examined", a Working Paper by: Menzie D. Chinn (University of Wisconsin, Madison and NBER).
La Follette School Working Paper No.2005-015, April 2005, pp.39