THE FAST TRACK ( TO PROFITS FOR PNLK )
How to Gauge the Importance of Trade to the U.S. Economy
Speech to the National Economists Club, Oct. 7, 1997
Jeffrey Frankel, Member, Council of Economic Advisers
Working to open markets around the world is one of President Clinton's highest economic priorities, along with education and deficit reduction. In its first four years the Administration negotiated some 200 agreements, counting all of them -- large and small -- multilateral, regional and bilateral. We hope and expect to continue this momentum in the second term. Major successes in the last year include the Information Technology Agreement and a global telecommunications pact. We hope to have a Multilateral Agreement on Investment before long, and to continue to make progress in civil aviation, and a host of other areas.
Fast track
Currently, our single most important objective in the area of international economics is obtaining from Congress the authority to negotiate trade agreements on a fast-track basis. Every President since 1974 has gotten fast-track authority from Congress. Why is it so important? The answer is that fast track lends credibility to trade negotiations. The parties to a trade agreement know that any provisions that are agreed upon cannot later be renegotiated individually, nor can such renegotiation be threatened.
History tells us that fast-track negotiating authority is critical. Congress failed to ratify the Kennedy Round GATT agreement in its entirety, and, as a result the Europeans refused to negotiate further trade agreements with the United States. For this reason, in 1974, Congress developed and passed legislation providing the President with fast track negotiating authority, giving the President the credibility he needs to negotiate international trade agreements.
So, why do we need fast track authority today? What would the Administration use it for? We would focus on three key areas for its use.
The first use for fast-track authority would be to complete the "built-in" agenda that was established in the most recent Round of GATT negotiations. At the end of the Uruguay Round, the United States, along with the European Union and other countries, pushed for a timetable along which negotiations in different areas would resume. We did that because we wanted more out of the Uruguay Round than we got. This year, we begin again the talks on government procurement; in 1998, we will begin intellectual property rights discussions; in 1999 we will begin talks on agriculture; then in the year 2000 on services. Government procurement is a trillion-dollar market for the United States in Asia alone over the next decade; agriculture, a $600-billion market globally; services, $1.2 trillion market. The United States wants better access to those markets. And, we must have fast track authority before beginning discussions in these areas. Otherwise countries will not put meaningful offers for market access on the table.
The second major use for fast-track authority is to extend the Information Technology Agreement. The recently concluded ITA will reduce to zero tariffs on a range of information technology products: semiconductors, computers, telecommunications equipment, faxes, phones, integrated circuits -- a huge array of products in which U.S. firms tend to be highly competitive. The U.S. tariff barriers in those areas are either very low or zero. Asia's tariffs in these areas averaged 30 percent. In the ITA, the United States agreed -- with 43 other countries -- that tariffs on information technology products should be brought to zero across the board in all countries, by roughly the year 2000. We already have agreement among our trading partners for an ITA-2 -- which would expand the scope of the products encompassed by this extremely ambitious initiative. And, fast-track authority would be needed to expand that arrangement.
We are also considering seeking similar kinds of agreements in a number of other individual sectors where the United States is very competitive but global barriers tend to be high -- areas such as environmental equipment and services, medical equipment and technology, and transportation equipment. In all of these areas, it would be necessary to have fast track authority to negotiate a reduction of trade barriers.
Finally, the third area that fast-track will be used is for comprehensive market access agreements with individual countries. Thus far, Chile has been identified as the first and most likely country for this kind of arrangement. Today, every major economy in this hemisphere--save the United States--has a preferential trade agreement with Chile. In practice, this means that U.S. firms face tariffs in Chile that make their goods 11 percent more expensive than those of competing firms from other countries. This is a serious competitive disadvantage.
Countries around the world are negotiating preferential trade agreements at a rapid pace. The United States could easily be left behind through inaction. Since 1992, other countries have negotiated 20 preferential trade agreements in Latin America and Asia that exclude the United States. The European Union has begun a process that will culminate in a free trade agreement with Brazil, Argentina, and other MERCOSUR nations; President Chirac of France has even gone so far as to declare that the economic interests of Latin America "lie not with the United States, but with Europe." And within South America itself, the four MERCOSUR countries may be in the process of extending their PTA to encompass the entire continent. We do not want to be left out of these trading blocs. We hope to make progress toward the FTAA in Santiago next April. In short, fast-track authority is essential to our ability to negotiate a variety of important trade agreements.
How should we think about the benefits of trade?
What is the proper way of thinking about or measuring the benefits of trade and trade agreements to the U.S.? In public policy discourse, the measure most commonly cited is the trade balance and its purported implications for jobs. Like most economists, I think this choice of emphasis can be misplaced. Instead of focusing on the difference between exports and imports, I would prefer to gauge the effects and benefits in terms of increases in the levels of trade. U.S. exports, for example, increased at an annual average rate of growth of 7.8% [goods and services; merchandise alone was 8.6%] over the years 1992-96. Thus trade has grown much more rapidly than income. This has been the pattern throughout most of the world, throughout most of the post-war period (and also during the century preceding 1914. The collapse of trade during the two world wars and the period between them were the exception). In theory, one wants to look at the effect of trade on a measure of economic welfare, such as real income. It can be difficult, however, to sort out the effects of specific trade agreements, from other forms of liberalization and globalization, such as declining transportation costs or the rise of high-exporting countries in Asia and elsewhere. This is true even when trying to sort out the effects on trade flows. To estimate the effects of trade agreements on real income, one has to take into account myriad other influences on the economy as well. Statistical estimates are prone to a range of uncertainty. The public instinctively knows this, and often distrusts confident assertions from Washington that a particular agreement or program will raise economic welfare by a particular amount. To take a recent example, the legislation implementing NAFTA required that we send a report to Congress on its progress and effects this past July. To isolate the effects of NAFTA, one has to adjust for both the effects of the 1994-95 Mexican peso crisis, which were negative, and the effects of the 1993-97 expansion of the U.S. domestic economy, which were positive. To make such adjustments is not easy, given only a few years of post-NAFTA data. Drawing on existing academic research, as well as new research commissioned by the government, we felt we were able to conclude that NAFTA taken in isolation clearly had (modest) positive effects, both on U.S. trade and real income. Nevertheless, poll results show that not all Americans have that same interpretation of events. Many mix together the effects of NAFTA with the effects of the peso crisis.
In the political arena, the public demands measures that are more tangible than econometrically-estimated effects on real income or other theoretical constructs of economists. For better or worse, we are stuck with having to think about jobs and the trade balance. In the remainder of my remarks I am going to talk about both approaches, the economist's preferred focus on the level of trade and real income, as well as the public's focus on the trade deficit and employment.
More jobs, or better jobs?
I will flag here my most important message. The opening of foreign markets raises demand for our goods, which in turn raises our real income. The increase in real income can take different forms. In particular, increased exports can contribute to real income both by increasing output and employment when there is slack in the economy and by reshuffling the jobs in the economy towards better higher-paying jobs. (Alternatively, the increase in real incomes can take the form of a real appreciation of the dollar.) When President Clinton first took office, there was a lot of slack in the economy. Unemployment was over 7 per cent. The increase in global demand for American goods during the period 1993-1996 showed up primarily as higher output and employment in exporting industries. The rapid growth in exports during the first Clinton Administration contributed importantly to the 13 million jobs that have been created since January 1993. (This contribution is also significant because it resulted in the creation of jobs that were on average better than might otherwise have been.) But now, in 1997, the economy is close to full employment [with unemployment at 4.9%]. The pool of unemployed workers is sufficiently small that jobs created in one sector tend to pull workers away from other sectors. They tend to add little to total employment. But export jobs still add to American real incomes because, on average, they pay 13-16% higher wages than do other jobs. To be sure, some workers will suffer job loss as the economy adjusts to globalization. But ultimately opportunities will be enhanced by the availability of good export-oriented jobs. Furthermore, statistical estimates suggest that growing trade need not imply an increase in the total frequency of involuntary dislocation among American workers. Although trade increases worker displacements in import-competing industries, export opportunities subtract from the displacements that would otherwise take place in other industries.
In other words, market opening consistently raises real income. In the first Clinton Administration, much of the effect showed up primarily in the form of more jobs. In the second Clinton Administration, the effect is showing up primarily in the form of better jobs. That's the message.
Recent external deficits
Just as job-creation is less of a concern once we are close to the constraints of full employment, so the trade balance is less of a concern when we are up against the constraints of potential output. Nevertheless, I will talk a bit about recent external deficits. The trade deficit on Goods & Services is still well below the 1987 peak of $153 billion, especially as a share of GDP. But it has been rising since its 1991 low-point [of $30 billion], and in 1996 reached $114 billion. The current account deficit likewise remains slightly below the 1987 peak of $167 billion, especially as a share of GDP. But it too is well up from its 1991 low-point [of $10 billion], reaching $165 billion in 1996. What have been the causes of these movements over the last ten years? The 1985-87 depreciation of the dollar was a major cause of the subsequent trade balance improvement. The 1990-91 recession was another clear source of "improvement," by depressing imports. [The current account deficit actually disappeared in early 1991 because, on top of the recession, we received transfers from Japan, Saudi Arabia, Kuwait, and others, in association with the Gulf War.] I think that the most important reason why the trade deficit and current account deficit have grown over the last year, or for that matter over the last six years, is that the US has grown more rapidly than our trading partners. Higher income brings higher imports. [See chart.]
As a share of GDP, the current account deficit is about the same as in 1989, the last time that unemployment was [almost] this low. [If even a small share of the statistical discrepancy is judged to belong in the current account, then it is currently better than 1989.] As a matter of pure macroeconomics, I am not too worried about the "deterioration" of the last six years. That it represents rapid growth sucking in imports is a good thing. Better yet, both national saving and investment have increased steadily as shares of GDP. [See chart.] The explanation for the widening current account deficit evidently is that investment is increasing more rapidly than national saving. This too is a good thing. [National saving and investment rates generally rise during expansions, but in this case they have been increasing even on a cyclically adjusted basis.]
The recent situation is very different from the much larger current account deterioration that took place between the 1970s and 1980s. That was a fall in the national saving rate: the famous fiscal profligacy of the first Reagan Administration, and a continued fall in private saving in the 2nd Reagan and Bush Administrations. The declines in National Saving and Investment were grounds for concern.. Our low rates of National Saving and Investment are leading candidates to explain our low rate of productivity growth since 1973, and so the recent improvements are good news. So far in the 1990s we have made up a little of the earlier decline, but only a little. More needs to be done. [See table.]
Prognosis: Houthakker-Magee and Current account sustainability
What about 1997 and the future? The trend is likely to continue this year: growth was strong in the first two quarters, and shows little sign of having slowed down in the third. The trade deficit so far this year is running 3.7% above last year [Jan.-July]. The budget deficit this year is continuing its strong improving trend, so it is quite possible that national saving, investment, and the current account deficit will, all three, continue to increase. Longer-term forecasts regarding the trade balance seem to be divided over whether or not there will be a deterioration in the coming years. Differences in assumptions about exchange rates and income levels account for much of that. But differences in assumptions about income elasticities of import demand also make a difference.
Some models have the property that the U.S. elasticity of demand for imports with respect to our income is much higher than the elasticity of the Rest of the World's imports from the U.S. with respect to their income. This property goes back to Houthakker and Magee (1969). The implication is that, holding constant relative growth rates and other determinants, the U.S. trade balance will get steadily worse over time. Indeed, the observation of long-term secular decline in U.S. trade drives the empirical finding and also gives it its only real intuitive plausibility.
These doomsday forecasts certainly merit thinking about. But I have my reservations. It is hard to interpret the hypothesis in terms of saving-investment, even if one does not demand it meet the full rigors of the academically-fashionable theory of the current account based on intertemporal optimization. My critique starts from the observation that an equation in which imports are determined solely by the size of the importing economy, with no constraint from the size of the exporting economy, only makes sense for a short-term time series, for example, over the course of one or two business cycles. It will not do for the long run. Consider U.S. imports from Asia. Within any one decade, it is fine to estimate imports from Asia as a function of U.S. income and relative prices. But does anyone think that this equation, if estimated for the 1960s, would still fit in the 1990s, when Asian economies are many times larger and the U.S. imports correspondingly more goods from them? One needs some measure of the size of the exporting economy in the equation, some notion of supply. I fear that trade models that lack such supply-side elements may be inadequate for thinking about the long run.
A technical aside
Paul Krugman, in an insufficiently studied 1989 article, argued several propositions: (1) There is a remarkable pattern in which, despite the high import elasticities in some countries (like the U.S.), which would seem to portend secularly declining trade balances, these countries coincidentally seem to grow more slowly than their trading partners, with the result that their trade balances don't deteriorate as much as expected. (2) If one derives a trade equation from the modern theory of trade in imperfect substitutes, then the exporting country's output belongs in the equation, alongside the importing country's income. (3) If one nevertheless omits the exporting country's output, it can be shown that econometric estimates will (spuriously) make it look as if slow-growing countries have higher elasticities than they in fact do. [The missing effect of foreign growth is appropriated by the domestic term]. Thus the first paradox, the apparent correlation between export elasticities and growth rates, is explained. Krugman did not undertake a full-fledged test of his hypothesis by estimating a global gravity model of bilateral trade. The gravity model is so-called by analogy with the law of gravitational attraction between heavenly bodies: trade between two countries is proportionate to the product of their sizes and inversely related to the distance between them. I have now done this (with Shang-Jin Wei of Harvard University ), and the results seem to bear out Krugman s prediction. When one includes the output [and income per capita] of the exporting country, along with those of the importing country, the coefficient on the exporting country's income is highly significant, and is almost as high as the coefficient on the importing country's income. The key question, for determining the sustainability of the U.S. current account deficit, is whether the U.S. elasticities are different from those for other countries (and higher on the import side than the export side). The answer is "no." This applies regardless whether the equations are estimated in levels or first differences. More research is required. But this piece of evidence suggests that the Houthakker-Magee finding is an artifact; at least that it should not be used without caution to think about the long run or to conclude that the U.S. current account deficit is not sustainable.
High levels of exports and imports are good
The trade balance can be equivalently regarded as saving minus investment, and exports minus imports. I have explained why I would rather focus on the level of National Saving and Investment rather than the difference. I'd now like so say something about focusing on the level of exports and imports, rather than the difference.
In negotiating reductions of trade barriers, the Clinton Administration is continuing the approach of other post-war Administrations. These pro-trade policies have succeeded in promoting the integration of the world economy. I mentioned earlier that, as a result, trade has increased more rapidly than income. [Some other factors in addition to government free-trade policies have also played a role: declining transport costs due to technological and economic innovation, and also the rise of Newly Industrializing Countries in East Asia and elsewhere.] This trend will continue, if we continue to pursue the right policies.
How do we know that increases in the level of trade are good? The classical theory of comparative advantage tells us that there are gains from trade. The "New Trade Theory" tells us the same thing. The classical theory goes back to Adam Smith and David Ricardo. It assumes perfect competition, constant returns to scale, and fixed technology, assumptions that are not very realistic. As is well known, the classical theory concludes that all countries gain from trade. Firms specialize in producing what they are best at, while consumers benefit from the opportunity to consume the goods available on world markets [as do firms buying inputs]. Less well-known is that the "New Trade Theory" -- which more realistically assumes imperfect competition, increasing returns to scale, and changing technology -- can be viewed as providing equally strong, or stronger, support for the sort of free-trade policies that the U.S. has followed throughout the post-war period, that is, multilateral and bilateral negotiations to reduce trade barriers. To be sure, these theories say that, under certain very special conditions, one country can get ahead by interventions (e.g., subsidies to strategic sectors), provided the actions of other countries are taken as given. But these theories also tend to have the property that a world in which everyone is subsidizing at once is a world in which everyone is worse off, and that we are all better off if we can agree to limit subsidies or other interventions. Bilateral or multilateral agreements where other sides make concessions to our products, in return for whatever concessions we make, are virtually the only sorts of trade agreements we have made. Indeed, most recent trade agreements (like NAFTA) have featured much larger reductions in import barriers on the part of our trading partners than we are required to make ourselves. The reason for this is that their barriers were higher than ours to start with. But the natural implication is that such agreements raise foreign demand for our products by more than they raise our demand for imports. Hence we are likely to benefit from a positive "terms of trade effect." This just adds to the usual benefits of increased efficiency of production and gains to consumers from international trade. Moreover, new trade theory offers reason to believe that openness can have a permanent effect on a country's rate of growth, not just the level of real GDP. A high rate of economic interaction with the rest of the world speeds the absorption of frontier technologies and global management best practices, spurs innovation and cost-cutting, and competes away monopoly.
Citing theory is not a complete answer to the question, "How do we know that growth in the level of trade is good?" So I would like to add a piece of empirical evidence. There is a large literature testing the determinants of countries' growth rates. Investment in physical capital and investment in human capital are the two factors that emerge the most strongly. But other factors matter as well. Estimates of growth equations have found a role for openness, measured for example as the sum of exports and imports as a share of GDP. But major concerns of simultaneous causality between growth and trade have been expressed. Does openness lead to growth, or does growth lead to openness? Some research I completed with a co-author just before entering the government [Frankel and Romer (1996)] aimed to deal with the simultaneity by using exogenous determinants of trade (from the aforementioned gravity model of bilateral trade, such as proximity to trading partners, as instrumental variables). We looked at a cross-section of 100 countries during the period since 1960. We found that the effect of openness on growth is even stronger when we correct for the simultaneity of openness and growth as compared to standard (OLS) estimates. Our estimate of the effect of openness on income per capita [.34] times the increase in U.S. openness since the 1950s [.12] implies an effect of 4 percent on U.S. income. In other words, this calculation suggests that the globalization of our economy over the last forty years has added an estimated $1,730 to current median family income. The fact that trade can affect a country's growth rate--as opposed to affecting the level of its GDP in a "one-shot" fashion--makes the case for trade liberalization even more compelling. Crucially, however, it is the overall size of a country's external sector that matters here, not the level of the trade balance. In other words, if the "new trade theory" view is right, then exporting four percent of GDP and importing five percent (like the U.S. in the 1960s) may be less beneficial than exporting thirteen percent and importing fifteen (like the U.S. today). Even though the first situation implies a trade deficit only half as large, it also implies an economy whose overall degree of openness to trade is far smaller--implying that it will enjoy a correspondingly lower level of benefits from trade.
I will close with three summary points. First, trade matters for an economy by raising real incomes. When the economy is close to full employment, the increase in real income mostly takes the form of better, higher-paying jobs. Second, there is no reason to fear that a current account deficit similar in magnitude to that of the United States today is the mark of a weak economy. Borrowing in order to invest is something that successful corporations do every day, and it really matters little whether the source of capital is domestic or foreign. And third, there is a significant role for our government in promoting trade--not through providing subsidies to export-sector firms or placing restrictions on imports, but rather by opening up markets to a greater flow of goods. Taken as a whole, these arguments suggest that aggressive pursuit of trade liberalization is the best way to ensure that the U.S. maintains its position as the strongest and most innovative economy in the world.
References
Frankel, Jeffrey, and David Romer, "Trade and Growth: An Empirical Investigation" NBER Working Paper No. 5476, March 1996. Revised, December 1996.
Frankel, Jeffrey, with Ernesto Stein and Shang-Jin Wei, Regional Trading Blocs, Institute for International Economics, Washington DC, 1997.
Haveman, John, 1997, "The Influence of Changing Trade Patterns on Displacements of Labor," Purdue University, January.
Houthakker, Hendrik, and Stephen Magee, "Income and Price Elasticities in World Trade," Review of Economics and Statistics 51, 2, May 1969, 1041-1105.
Krugman, Paul, "Differences in Income Elasticities and Trends in Real Exchange Rates," European Economic Review 33, 5, May 1989, 1031-1046.
Richardson, J.David, and Karin Rindahl, 1996, Why Exports Matter: More! Institute for International Economics and The Manufacturing Institute, Washington, DC.
The White House, 1997, Study on the Operation and Effects of the North American Free Trade Agreement, Washington DC, July.
ALLEN: |
