Fool’s Gold:
Projections of the U.S. Import Market

By Dean Baker and Mark Weisbrot[1]

January 8, 2004

Executive Summary

            For developing countries, the main selling point of new commercial agreements such as the Free Trade Area of the Americas (FTAA) or the Central American Free Trade Agreement (CAFTA) has been the lure of increased access to U.S. markets. And indeed these markets have grown considerably in recent years.

            However, it is not clear that most developing countries will be able to increase their exports to the United States in the foreseeable future. The United States is now running an unsustainable trade and current account deficit, with the latter now at more than 5 percent of GDP. Economists are in agreement that the United States cannot continue to sell financial assets at this rate to the rest of the world for an indefinite period of time. The implication of the required adjustment is that the market for exports to the United States will shrink.

            This paper looks at the U.S. market for imports over the next decade under the assumption that the United States' foreign borrowing, while not disappearing, will have to at least stabilize. The authors assume that the United States' foreign debt stabilizes at 50 percent of GDP, an unprecedented level of foreign debt for a developed country.

            The paper finds that under any plausible set of assumptions about the adjustment process, U.S. imports -- measured in non-dollar terms in order to reflect their value to other countries -- will actually decline. The estimated decline is between $90 billion and $375 billion, in today's (2003) dollars. The range of estimates varies with assumptions about the growth of U.S. exports, the elasticity of demand for U.S. imports and exports, and the percentage of currency depreciation that is "passed through" to import and export prices.

            This inevitable adjustment process has enormous implications for developing countries seeking access to U.S. markets. It means that, as a group, they can no longer expect to increase their exports earnings from U.S. markets in the foreseeable future, regardless of the terms of present or future commercial agreements. The only countries that can gain from increased access to U.S. markets will be those that do so at the expense of other countries.

            Commercial agreements such as the WTO, proposed FTAA, and bilateral agreements typically require developing countries to make important and sometimes economically costly concessions in such areas as intellectual property, rules governing investment and government procurement, or other policies. For most countries, the costs of such concessions can be expected to exceed any gain they might anticipate from increased access to a shrinking U.S. market for their exports.


Fool’s Gold: Projections of the U.S. Import Market

            Many developing country governments see access to the United States import market as a key part of their development strategy. This has been their main motivation in seeking out trade pacts such as the Central America Free Trade Agreement or the Free Trade Area of the Americas. The expansion in this market certainly has provided a path for growth for some developing countries in recent years. Over the last dozen years, annual imports into the United States have increased by nearly $780 billion, measured in 2003 dollars.[2] Since the real value of the dollar has appreciated against other currencies over this period, the increase in the value of U.S. imports measured in other currencies would be even larger. Measured against a weighted basket of other currencies, the increase in annual imports to the United States between 1991 and 2003 would be the equivalent of more than $860 billion 2003 dollars.[3]

            However, it is questionable whether access to the United States import market will turn out to be of much value in the coming decade. The extraordinary growth in U.S. imports over the last twelve years cannot possibly be repeated. This run-up in imports has already led the United States to become the world’s largest debtor nation, with a negative net asset position that will almost certainly exceed $3 trillion at the end of 2003.[4] At present, the U.S. current account deficit is running at an annual rate of more than $540 billion a year, a deficit that can be sustained only as long as the United States can sell this amount of financial assets to foreigners.

            The projections in this paper show that, in contrast to their extraordinary growth over the last decade, the annual value of U.S. imports will almost certainly shrink over the next decade, when measured in non-dollar currencies. If developing countries want to increase their exports to the United States over this period, they will have to beat out other exporters, like Mexico and China. On net, there will be no opportunity to gain market share in the United States at the expense of domestic production. If countries enter into trade agreements with the United States under the assumption that the import growth of the last dozen years will continue, then they will be seriously disappointed. 


The Construction of the Projections

            It is necessary to make a series of assumptions on the path of key economic variables in order to construct projections for imports. These variables include the growth path of foreign indebtedness, the growth of U.S. exports, the responsiveness of the demand for imports and exports to changes in prices, and the price response of imports and exports to changes in the value of the dollar. These assumptions are discussed in turn below.

            The basic assumption for the growth path of net foreign indebtedness is that it stabilizes at 50 percent of GDP in 2013. This implies an increase of approximately 23 percentage points from the 27 percent debt-to-GDP ratio that the United States is likely to have at the end of 2003. Since the current account deficit is currently more than 5 percent of GDP (which implies that the U.S. is borrowing an amount equal to 5 percent of GDP each year), the assumption that the debt-to-GDP ratio stabilizes at 50 percent in 2013 implies a fairly rapid pace of adjustment. While the debt-to-GDP ratio could stabilize at a lower level, this would almost certainly require a sharp downturn in the U.S. economy, which would lower imports and bring the trade deficit closer to balance. In the absence of a sharp downturn, it is difficult to imagine a scenario in which the debt-to-GDP ratio does not rise close to 50 percent by 2013.[5]

            It is entirely possible that the debt-to-GDP will have not yet stabilized by 2013. This would mean that imports could be higher than in the projections constructed in this paper in 2013, but will have to be lower at some future date. The rising debt-to-GDP ratio implies larger future interest/dividend payments to people living abroad. U.S. export earnings will have to be used to make these payments rather than paying for imports.

            The projections assume that the average real return on foreign holdings of U.S. assets is 3.5 percent. This assumption is considerably higher than the return that these assets are receiving at present. There are several reasons for assuming a higher return in the future. First, a large portion of foreign assets is currently held as official reserves by central banks. These are short-term deposits that offer very low interest rates. While it is likely that a substantial amount of dollars will continue to be held as short-term reserves, it is reasonable to believe that the amount held for this purpose will be reduced in the future.[6] In any case, the portion of U.S. assets held as reserves is almost certain to fall as the absolute size of foreign holdings increases. Second, the current interest rate structure in the United States is extremely low. Historically, short-term deposits have provided real returns of close to 2.0 percent (compared to minus 1.0 percent at present). The real return on government bonds has averaged close to 3.0 percent, with corporate bonds offering a return that averages more than a percentage point higher. If the return on equity remains near its historic average of 7.0 percent, then a conservative mix of assets should easily provide a real return of 3.5 percent.[7]

            Using the 50 percent ratio of foreign debt-to-GDP, and the assumption of a 3.5 percent real return on assets, by adding in a growth rate assumption it is possible to calculate the trade deficit that is consistent with a stable debt-to-GDP ratio. This analysis assumes that annual GDP growth averages 2.5 percent in the middle of the next decade, somewhat faster than the 2.0 percent growth rate projected by the Congressional Budget Office and considerably faster than the 1.8 percent growth rate projected in the Social Security Trustees' Report. The 2.5 percent growth rate implies that a current account deficit equal to 1.25 percent of GDP is consistent with a stable debt-to-GDP ratio, which in turn implies a trade surplus equal to 0.5 percent of GDP, given the necessary interest and dividend payments on the debt.[8]

            The next step is to project a path for exports. The simplest method for projecting baseline exports is to assume that the real growth of exports over the next decade is the same as it has been in prior decades. The central assumption in these projections assumes that real exports grow at the rate of 2.87 percent annually, their average rate of growth over the years from 1960 to 2002.[9] This baseline growth path implies that exports will be equal to $1,350 billion in 2013 (measured in 2003 dollars). Separate high import and low import scenarios alternatively assume that the growth rate of exports is half a percentage point higher (3.37 percent) or half a percentage point lower (2.37 percent). 

            Finally, it is necessary to specify an adjustment process that will allow the current account to adjust to a sustainable level. This requires assumptions about both the sensitivity of imports and exports to changes in price, and assumptions about the extent to which changes in the value of the dollar are reflected in the price of imports and exports. In the case of both imports and exports, the central assumption is that the elasticity of both imports and exports with respect to price changes is 2, which implies that a 10 percent increase in the price of imports will lead to a 20 percent decline in demand. An alternative high import scenario assumes that the elasticity of demand for both imports and exports is 2.5.

            The central assumption on the sensitivity of import and export prices to changes in currency values (the “pass through”) is 0.5, which means that the change in the price of the product is half of what would be implied by the change in the currency value. For example, this means that if the dollar falls by 10 percent, then import prices rise by an average of 5 percent. The implication is that the other half of the increase in costs is absorbed in the form of lower profits by suppliers. High import and low import scenarios assume alternatively that the pass through rate is 0.6 or 0.4, respectively.[10]

            Table 1 below shows projections for import levels in 2013 and the decline in the value of the dollar that will be needed to bring the current account deficit down to a sustainable level. The import levels are shown in both dollar terms and measured in real units of non-dollar currencies. Since the real value of the dollar will have to decline in order to bring the current account into balance, the dollars earned by these imports will have considerably less value elsewhere in the world than do dollars today – even after adjusting for domestic inflation in the United States. In addition to showing the middle import scenario, the table also shows the high import and low import scenarios described above.

Table 1

Projections of U.S. Imports in 2013

 

Imports

Decline in Dollar

Non-dollar value

Change 2003-2013

 

(2003 dollars)

 

 

(2003 dollars)

Middle

$1779 billion

-22.8%

1374 billion

-163 billion

Low

$2002 billion

-42.0%

1161 billion

-375 billion

High

$1678 billion

-13.8%

1446 billion

-90 billion

Source: Congressional Budget Office, Bureau of Economic Analysis, and authors' calculations; see text.

            All three scenarios show that the imports will only increase modestly when measured in dollar terms over the next decade, from their level in the second quarter of 2003 ($1,540 billion). However, because the dollar will have to fall substantially from its current level in order to bring the current account deficit to a sustainable level, the value of U.S. imports measured in real units of other currencies (the “non-dollar value” column in table 1) will fall in all three scenarios. For example, in the middle scenario, the projected value of U.S. imports will be $163 billion less in 2013 than it is today, when measured in a basket of foreign currencies. This means that if countries took the dollars earned from their exports to the United States and then traded these dollars for other currencies, the purchasing power of what they would be able to buy based on their 2013 exports would be $163 billion less (in 2003 dollars) than what they could buy based on 2003 exports.

            The fact that imports decline even in the high-import scenario shows the insoluble nature of the problem. The current level of the U.S. trade deficit is clearly unsustainable. The adjustment will require a sharp drop in imports in the near future. Optimistic assumptions about the future reduce the size of the implied decline in imports, but it is not possible to construct a plausible scenario in which the U.S. can even sustain its current level of imports.  Measured in real terms, the extraordinary growth in U.S. imports over the last twelve years clearly will not be repeated.

            It is worth noting that the projections in Table 1 actually understate the negative trend in the U.S. import market from the standpoint of developing countries. One of the assumptions used in the construction of the table is that declines in the dollar are not passed through by 100 percent in the form of higher prices in the U.S. The central assumption in the scenarios in the table was that the pass through rate was 50 percent. This assumption implies that half of the impact of a lower dollar is felt in the form of lower profit margins. This means in the case of the middle scenario, where the implied reduction in the value of the dollar is 22.8 percent, the reduction in profit margins will be equal to approximately 11.4 percent of the price of the products. Of course, not all of this reduction in margins will be felt by firms in the exporting country; much will be absorbed by shippers, wholesalers, retailers, and other intermediaries. However, the implied decline in the dollar means that not only will the value of exports to the United States decline over the next decade, but the unit profits of these exports will almost certainly fall as well. In short, increased access to the U.S. import market is not likely to be of great value over the next decade.


Conclusion

            This paper constructs a set of projections of U.S. imports for 2013. It shows that under plausible assumptions the U.S. import market will decline over the next decade, when measured using the real value of non-dollar currencies, such as the Japanese yen or the euro. The projections show that imports will decline even with optimistic assumptions: for example that the annual growth rate of U.S. exports is half a percentage point higher than its average over the last four decades (before taking account of the stimulus to exports provided by a falling dollar). Furthermore, the projected declines in the value of the dollar are also likely to squeeze the profit margins of exporters, as they absorb some of the effect of a falling dollar instead of passing it all through in the form of higher import prices in the United States.

            The United States current account deficit is clearly unsustainable. The unavoidable implication is that imports will have to shrink in the near future. In contrast to the huge growth in the U.S. import market over the last decade, there will be a substantial decline in the size of the market over the next decade. This means that efforts by most developing countries to gain access to the U.S. import market – if they involve important concessions in other areas (e.g. on intellectual property rights, investment or government procurement rules) – are likely to prove misguided. Except for the few that can increase their exports substantially to the United States by displacing other exporters, any significant concessions made in order to gain access to U.S. markets would lead to a net loss for the countries that make them.


[1] Dean Baker and Mark Weisbrot are co-Directors of the Center for Economic and Policy Research

[2] This calculation takes the nominal value of imports in the third quarter of 1991 ($625.3 billion) and converts it into 2003 dollars, using the GDP deflator, rather than the deflator for imports. This gives a better measure of what can be purchased with the income earned through exporting to the United States.

[3] This figure uses the Federal Reserve Board’s real broad index to adjust for the real appreciation of the dollar against other currencies over this 12-year period. The calculation effectively compares the real value of imports in 1991, measured in non-dollar currencies, with the real value of imports in 2003, also measured in non-dollar currencies. This difference is then converted back into dollars at the exchange rates that prevailed in the third quarter of 2003.

[4] The Bureau of Economic Analysis estimated that the U.S. net international investment position was negative $2.6 trillion (measured at market costs) at the end of 2002 (“U.S. Net International Investment Position at Yearend 2002,” Bureau of Economic Analysis, June 2003). With the current account deficit certain to exceed $500 billion in 2003, the negative net international investment position will almost certainly be pushed to about $3 trillion.

[5] If the trade deficit stays constant as a share of GDP over the next decade, and the average real return on foreign held assets in the United States is 3.5 percent, the ratio of foreign indebtedness to GDP would exceed 70 percent by the end of 2013.

[6] One reason for believing that amount of dollar reserve holdings will decline is that countries like China are unlikely to want to hold vast amounts of dollars indefinitely in order to depress the value of their currency. The second main reason is that the use of the euro as a reserve currency is increasing at a rapid pace. Given the current expansion plans for the EU, the size of euro zone economy will considerably exceed the size of the U.S. economy by 2013, especially if the United Kingdom adopts the euro. This should further increase the value of the euro as a reserve currency relative to the dollar.

[7] Historically the return on U.S. holdings of foreign assets has been substantially higher than the return on foreign holdings of U.S. assets. While this difference in rate of return has limited the extent to which the net income flows are negative, the difference in rates of return will matter less as foreign holdings of U.S. assets grow to levels that will be more than twice as large U.S. holdings of foreign assets.   

[8] The 3.5 percent return on a foreign debt that is equal to 50 percent of GDP implies that an amount equal to 1.75 percent of GDP will be paid out as dividends or interest each year. In order to bring the current account deficit down to 1.25 percent of GDP, the trade surplus will have to be equal to 0.5 percent of GDP. The slower growth rates assumed by the Congressional Budget Office and the Social Security trustees report would imply even larger trade surpluses  -- 0.75 percent of GDP in the case of the CBO projection and 0.85 percent of GDP in the case of the Social Security trustees projections. 

[9] This calculation uses the nominal value of exports and then deflates them using the GDP deflator. This method avoids some of the problems associated with pricing some important exports, most notably computers.

[10] Some estimates of pass through are significantly lower; this would make the import market in the U.S. contract by more than the estimates in this paper. See e.g., Bernake, Ben S., "Monetary Policy and the Economic Outlook: 2004, p.12, (Paper presented at the proceedings of the American Economic Association, January 2004), which assumes a pass through of 0.3. Bernake is a member of the Board of Governors of the Federal Reserve System.