There is broad agreement between economists and trade negotiators that the multilateral approach to reducing trade barriers--one that is as inclusive as possible--is the best approach. This is somewhat intuitive, in that solutions to problems should be more effective the more broadly they are applied. Thus, the best way to solve world hunger would be with a program that involved every affected country. The same would hold true for illiteracy.
It also holds true for trade, yet there seems to be much disagreement between economists and trade negotiators regarding the potential benefits of regional or bilateral free trade agreements (FTAs). Many economists hold that FTAs can be harmful to the process of liberalization and even result in an economic loss. "[The] present bilateral approach to trade policy is gradually undermining the open multilateral trading system," warns Dr. Anne Krueger, former President of the American Economic Association.1
Criticism of FTAs strikes many trade negotiators and government leaders as a paradox, almost akin to arguing that although a global anti-hunger or anti-illiteracy campaign would be a step forward, the very same campaign adopted by only two countries or a region could represent a step back. Half a loaf, in other words, is worse than none. How can this be so?2
This essay will examine the arguments concerning FTAs and conclude that despite the theoretical possibility of economic hardship, FTAs are generally an economic good, although care must be taken as to how they are constructed.3 Thus, the United States should be open to free trade agreements outside of the World Trade Organization (WTO), be they regional, bilateral, sectoral, or unilateral. The meeting of the WTO in Seattle last November is a step forward for U.S. and global trade interests, but that should not preclude the United States from taking other steps as well.
Economist Jacob Viner first explored the potentially contradictory impact of trade in the 1950s by noting that trade liberalization could result in either "trade creation" or "trade diversion."4 Trade creation is the benefit accrued when a country lowers its trade barriers, such as a tariff, with the most efficient producer of a product as happened, for example when the North American Free Trade Agreement (NAFTA) allowed tomatoes from Mexico to enter the U.S. market. Thus, the most efficient producer of a product would now sell more to the liberalizing country and would do so at lower prices. The consumers of the importing country would enjoy greater quantities of the product at lower prices, and the producers would enjoy greater sales.
But what happens under trade diversion, which is what occurs when trade barriers are lowered with an inefficient producer of a product? For example, NAFTA allows sugar from Mexico to enter the U.S. market.5 Here, the economic results are ambiguous. True, U.S. sugar consumers enjoy the benefits of lower costs and will tend to import more of the product, and Mexican producers enjoy the benefits of greater sales. But the most efficient producer of sugar--Australia, for example--will have been displaced in the U.S. market. Theoretically Australia's loss could be greater than the others' gains. If so, trade has been diverted, and the result is a global economic loss, with the harm to Australia outweighing the gain to the U.S. and Mexico. The Economist notes:
Regional "free-trade areas" need not make trade freer. By liberalizing trade only with their neighbors, countries are by definition discriminating against those not lucky enough to be in the local club. Some goods will be imported from other members of the free-trade area at the expense of producers elsewhere; and members will begin to specialize in industries in which they lack comparative advantage.6
One major defect of the analysis is that it is static. Over time, a new exporting country's productive efficiencies can improve as skills are gained and volumes attained. So the theory of trade diversion would see a problem with NAFTA's allowing the U.S. to import autos from a less efficient producer (Canada) at the expense of the more efficient producer (Japan). But after several years of NAFTA experience, Canadian auto plants can frequently produce at Japanese levels of productivity.
There is not necessarily an intrinsic reason why one country is a more efficient producer to begin with, so to accept the trade diversion argument is to believe that current producers of a product will retain their current competitive advantage in perpetuity. Yet we know from history that competitive advantage can change over time.
Second, the theory as frequently stated misrepresents the effects of trade diversion. The actual losses of the supplanted exporter (its forgone "producer's surplus") are smaller than it would appear, because that country will go on to produce something else, a second-best alternative. Similarly, the new exporting country must forsake the production of something, so it does not gain all of the new producer's surplus.
In other words, the actual shift in fortunes is less than the argument contends because now that Mexico is exporting sugar, it must forgo the production of a different item; now that Australia is no longer exporting sugar, it will produce more of a different item.
Looking solely at production costs has another shortcoming: It ignores the indirect business costs. To take our example, even if tomatoes from China were found to be cheaper than those from Mexico, there are a range of reasons why a U.S. vegetable importer might well prefer to continue doing business with Mexico: less expensive transportation costs, freshness of produce, rapidity of delivery, ease of Spanish-English coordination, ease of same time zone coordination, breadth of dollar-peso instruments, and so on. In other words, the indirect costs of doing business with a seemingly less expensive supplier might in fact be greater.
To the extent that trade diversion does occur, the importing nation has a strong incentive to remedy the matter. If the United States has displaced the most efficient producer of sugar by signing NAFTA, then the United States has a strong incentive to remedy that by allowing Australian sugar to come in tariff-free, at which point U.S. sugar prices would be cheaper still. Why would the U.S. government penalize U.S. consumers just to support inefficient Mexican producers?
Most significantly, the diversion argument also ignores the aggregate effects of liberalization: An FTA might result in some trade creation and some trade diversion, but the former is likely to substantially outweigh the latter. As FTAs are a relatively recent phenomenon, there is not much in the way of scholarly studies to show that trade diversion has exceeded trade creation. Mexican sugar has supplanted Australian sugar, but other aspects of U.S.-Mexican trade could well reflect trade creation.
There are no satisfactory ex post estimates of regional integration on excluded countries' welfare.... We have virtually no hard evidence on how regional integration affects economic welfare in the rest of world.... [We] have largely failed to measure the effects that theory shows to matter and...much of what we have measured hardly matters at all.7
As the economic arguments recede in the debate, two other arguments are advanced against FTAs: tidiness and distraction. The tidiness argument is that it is too messy to have a series of FTAs involving different countries on different issues with different timetables. Trade specialist Claude Barfield states:
[A] trading world dotted with separate bilateral and multilateral FTAs, each with different interim timetables, tariff levels, and nontariff barrier liberalization rules, would become enormously inefficient.... A multinational corporation such as IBM, Siemens, or Samsung would face a daunting task in sorting out trade rules that governed their simultaneous operations in APEC, NAFTA, MERCOSUR, the European Economic Area--not to mention individual countries such as Chile or Turkey, which had separate arrangements with regional groupings and individual nations.8
The distraction argument, on the other hand, is not that FTAs are too messy, but that they are too bothersome, absorbing efforts for small gains when those efforts could be used more effectively elsewhere, such as in multilateral negotiations. "Given limited resources in the office of the USTR and in other federal agencies responsible for formulating and implementing trade policy, it was inevitable that NAFTA would result in some diversion of attention from the multilateral system," states Krueger.9
The tidiness premise is that a multinational corporation would be distraught if one country offered a 10 percent reduction in its tariffs and another country offered a 20 percent reduction, because this would complicate planning. But multinationals successfully accommodate national distinctions in such non-trivial matters as currencies, taxes, labor, health and packaging regulations, and consumer preferences, so why not in trade barriers? A sales manager who complained about an uneven reduction in a tariff because it was too complicated, instead of viewing this as an opportunity to improve sales, is a manager who will soon be leaving his business.
The amount of information to be managed under a plethora of trade regimes would be far less complicated than, say, the amount of information required to assign a special numerical code to every individual. In other words, it would be less complicated than using a phone book. Indeed, there are several Internet sites that offer free calculations of duties once simple data are entered.
But the main rebuttal to the tidiness argument is the matter of practicality. The reason that there are so many peculiar trade initiatives is not because trade negotiators have a penchant for the untidy, or because they enjoy confounding economists; it is because the particular set of trade agreements in place are the ones that are feasible. There is no reason that trade policy formulation and implementation should be any less prone to the slow pace of government, the bureaucratic interplay, and the role of special-interest groups than other aspects of public policy. It should not be surprising that trade policy proceeds unevenly, perhaps even untidily.
But why not take whatever incremental steps one can toward liberalization when it is possible to take them? It is simply easier to get two countries to agree on an initiative than to get the 134 members of the WTO to agree. Why should trade liberalization be a convoy, forced to move only at the speed of the slowest ship? The advantages of the FTA approach are that countries can move at a pace they determine to be the best for themselves; countries can experiment with different arrangements and timetables; and countries are not constrained by the lowest-common-denominator policies of the WTO, in which the need for unanimity allows parochial interest to slow down the entire process.
FTAs are more politically feasible than a multilateral agreement because they involve fewer participants and can be more precisely customized to meet their needs. Sensitive issues can be broached on a smaller scale, which can make the change more politically palatable. Americans can accept the idea of textile imports from Mexico more readily than the idea of textile imports from China. Vietnam will entertain the idea of free trade with its neighbors in the Association of Southeast Asian Nations (ASEAN), but it will take a while for it to be comfortable with the idea of free trade with the largest Pacific economies, all former adversaries: China, Japan, and the United States. Simply put, countries can be more willing to augment their trading relations on an incremental basis than they are to make a bold move through the WTO.
Trade economist Jagdish Bhagwati, who has written elegantly and persuasively on the benefits of trade liberalization, pejoratively labels this uneven approach a "spaghetti bowl" approach to trade.10 If that is what FTAs represent, perhaps more spaghetti is in order.11
Among other weaknesses, the distraction argument is premised on a false alternative: that a global round of liberalization or a regional agreement are both equally likely. If so, a more encompassing multilateral accord is surely the bigger prize. Yet what is the basis for that assumption? The recently completed Uruguay Round required 13 years to bring to a conclusion. The nascent Seattle Round of WTO trade liberalization will involve several dozen more countries and a range of new and potentially sticky issues from genetically modified foods to Internet commerce, all of which implies that this new round could well make the Uruguay Round look speedy by comparison. The protracted contest for the position of WTO Secretary General between former New Zealand Prime Minister Mike Moore and Thailand Deputy Prime Minister Supachai Panitchpakdi should give all onlookers pause for thought about placing all hopes for liberalization on the WTO.12
The unstated premise of pure multilateralism is that the world trading system is one of equal countries of equal economic weight, equally disposed toward trade liberalization. If, on the other hand, most of the world's trade is undertaken by a handful of countries or groupings, could not more economic benefit be achieved more quickly by working with this core group?
The distraction argument is questionable for other reasons. Although it is plausible that FTAs distract from multilateral agreements, it is equally plausible that FTAs build support for such agreements. FTAs demonstrate on a small scale that countries can withstand the problems of dislocation and will benefit from expanded trade. Remember: Even if there is trade diversion, trade liberalization results in benefits for the importing and exporting country. Whatever trade diversion might exist would be suffered by the third country which is not participating in the agreement. The lesson here is: Participate in agreements and you will not suffer losses. Or if you cannot participate in this particular agreement, go and initiate something of your own.
Similarly, FTAs could be viewed as an exercise in constituency-building. They train the government bureaucracies to recognize that liberalization is healthy. They show the general public that increased trade is an economic good. They allow the participating countries to develop the sinews of international commerce, from port improvements to international communication links to financial practices. These all create a predisposition for further liberalization. It creates trade winners to counter the naysayers who believe that trade only creates losers.
The distraction argument also suffers from a confusion of ends and means. Is the goal of international trade to promote the international trading system, or is the goal of the international trading system to promote international trade? A free trade agreement between NAFTA and the European Union (EU) would embody over 50 percent of world gross domestic product (GDP) and could be more important economically than the Uruguay Round.13 If the goal of the WTO could be reached, albeit outside the WTO process, should we not be equally grateful? Rather than FTAs distracting from the WTO, one could better argue in this instance that the real problem with the WTO is that it distracts from the ability to achieve FTAs.
The experience of the United States as both the NAFTA and the Uruguay Round negotiations drew to a close is instructive. The NAFTA discussions were moving along more quickly, and the message to the international community was clear: The United States is moving ahead with trade liberalization, and you are free to stay on the sidelines; but you will not have the ability to block those who want to move ahead. Preserving the ability of the countries that are liberalizing more quickly to operate outside of the WTO will induce the more recalcitrant countries to make the WTO work.
Indeed, a committed protectionist country would pursue two parallel goals in trade policy: First, attempt to ensure that countries have no alternative to the multilateral process. Second, frustrate and delay all multilateral initiatives.
Conversely, FTAs cause the protectionist country to pay a price for its policies. Not every country will participate in every FTA, but the protectionist country will never participate in any. As a result, it will never have the opportunity to offset the trade diversion of one FTA with the trade creation of another.
In the end, the question of whether FTAs help or hinder the multilateral process is unprovable because of the many variables that affect the emergence of FTAs and the conclusion of multilateral rounds. There is not a large body of scholarly analysis on this matter.14 It is worth noting, however, that in other international agreements from the mundane (international postal union and international civil aviation agreements) to the acute (chemical weapons treaties, human rights agreements), plurilateral accords are typically viewed as adding to, not detracting from, multilateral agreements.15
First, FTAs should explicitly adopt a policy of broadening membership on an open-ended basis. No country should be able to pull up the ladder after it enters the clubhouse. In other words, FTAs should be structured so that they represent the first step in liberalization, not the last step.
The reason to structure FTAs to embrace an ongoing augmentation of membership is the sequencing issue. Some might argue that sequencing is not important, that FTAs will naturally be extended beyond the original membership, eventually linking up with other FTAs and leading finally to worldwide free trade. But there could be an inducement for some members not to expand. The least efficient participating countries, having gained preferential access to some markets, might oppose extending the membership of the FTA. For instance, if Chile were to join NAFTA, it would somewhat diminish the value of the preferences won by Mexico. Chile would, to some minor extent, displace Mexico in the U.S. market.16
A second issue involving FTAs is the rules of origin. Because participating countries in the FTA will continue to maintain different external tariff levels, a loophole exists for third parties. Importers from non-member nations could avoid higher tariffs by transshipping through the FTA member with the lowest external tariff. To close this loophole, FTA agreements contain rules of origin clauses. But as more and more bilateral FTAs are formed, "a mass of paperwork is created for customs officials as they try to certify which shipment could benefit from which set of preferences."17 Participating countries should examine mechanisms to mitigate the rules of origin issues, such as establishing low levels of value before a product qualifies as locally produced or eliminating the greatest discrepancies in the participating countries' external tariffs.
Third, countries should not enter into free trade agreements that are hub-and-spoke, which give one country structural privileges over other members. Privileges should be transitive so that any new member of an FTA immediately gains equal privileges with all other members, which will allow the benefits of the FTA to be maximized.
Opposition to FTAs is led by those who favor multilateral mechanisms for trade liberalization; but rather than being called multilateralists, they might more aptly be termed universalists, because by their own logic even the WTO is lacking. Major trading economies like China, Russia, and Taiwan are not in the WTO.18 A pure universalist would disapprove of the Uruguay Round, which will undoubtedly lead, if only in a small way, to some trade diversion.19
For that matter, if you are an all-or-nothing universalist, why should countries have free markets internally? A universalist should be even uncomfortable with the idea of domestic markets, because they are merely miniature FTAs.20
There is an underlying irony in the position of the universalists: If the premise of free trade is that competition works, shouldn't that also be the premise of free trade bodies? Shouldn't countries be able to "shop around" by considering joining NAFTA or the WTO, and even special trade relations with regional groupings such as the EU, AFTA (ASEAN Free Trade Area), Caribbean Common Market, and other organs? Consumers in some countries will respond more readily to the benefits of enhanced trade and should be allowed to move in kind.
There are a range of bilateral and regional trade initiatives under discussion. The European Union is grappling with extending membership to the nations of Central Europe. The United States has been negotiating an FTA with Chile for a number of years, and Chile has had similar discussions with New Zealand. Israel has held discussions off and on with both Canada and Mexico, and an FTA with either would make Israel a member of NAFTA, since it already enjoys an FTA with the United States. Congress recently passed a welcome African trade initiative that will extend unilateral privileges to some of the poorest countries in the world.
Some of the more ambitious initiatives are found in Southeast Asia. At the September 1999 Auckland summit of the Asia-Pacific Economic Cooperation forum (APEC), New Zealand proposed a "P-5" initiative, which would be an FTA among NAFTA, New Zealand, Australia, Chile, and Singapore. Singapore has proposed an FTA among the ASEAN nations and Australia and New Zealand.
On the other side of the globe, several political leaders have proposed a "TAFTA"--a Trans-Atlantic Free Trade Area between NAFTA and the EU. Former Republican presidential candidate Steve Forbes has proposed a NAFTA-UK-Ireland FTA.
Since Viner first noted the potential economic damage that might be done by an FTA, economics literature has at best viewed FTAs with suspicion, and many leading trade economists treat FTAs with outright hostility. But despite the warnings of the critics, there is not a great deal of evidence that FTAs in the aggregate have been harmful, and there are strong arguments that FTAs are not just helpful in themselves, but also useful steps on a path to broader liberalization.
Franklin L. Lavin has served in the U.S. Department of Commerce and on the National Security Council. He currently works for a venture capital firm in Hong Kong.
2. Opponents of regional free trade agreements (FTAs) sometimes label them Preferential Trade Agreements (PTAs) as they convey a preferential treatment to the participants. FTAs should be distinguished from multilateral accords that attempt to encompass all trading nations, typically through the General Agreement on Tariffs and Trade (GATT) or its successor organization, the World Trade Organization.
3. Trade specialists distinguish between four fundamental types of international economic arrangements: Free Trade Areas, in which the participating countries eliminate mutual trade barriers but retain their different external tariffs; Customs Unions, in which the participating countries eliminate mutual barriers and adopt a common external tariff; a Common Market, which is a Customs Union that also allows for free internal movements of labor and investment; and an Economic Union, which is a Common Market whose members also harmonize their monetary policy.
11. FTAs can be more effective than the WTO in reducing participating countries' non-tariff barriers because of the enormous scrutiny they afford the members. NAFTA was a 2,000-page document because it exhaustively reviewed Mexican and American non-tariff trade barriers. How much detailed scrutiny does Mexico get at the WTO?
13. See Thomas J. Duesterberg, "Prospects for an EU-NAFTA Free Trade Agreement," The Washington Quarterly, Vol. 18, No. 2 (Spring 1995), pp. 71-82. See also Robert B. Zoellick, "How to Achieve Trans-Atlantic Free Trade," The Wall Street Journal Europe, June 14, 1995, p. A15.
14. See, for example, Murray C. Kemp and Henry Wan, "An Elementary Proposition Concerning the Formation of Customs Unions," Journal of International Economics, Vol. 6 (February 1976), pp. 95-98, for a positive assessment of the interrelationship. See Pravin Krishna, "Regionalism and Multilateralism: A Political Economy Approach," Department of Economics, Columbia University, 1993, for a negative assessment.
15. A multilateral accord is theoretically one that involves all nations; a plurilateral accord is one that involves only a subset of nations. The actual difference between the two terms is a matter of degree, because even the most inclusive multilateral organizations and agreements do not, in fact, involve all political or economic entities.
16. Of course, there are compensating elements as well. Mexico and Canada gain access to Chile's market and also benefit from what Chile has to supply to their market. Both countries, to their credit, have favored Chile's accession to NAFTA.
19. The universalists are in the awkward position of explaining why the Uruguay Round, which affected about 90 percent of global GDP, is a good but an EU-NAFTA accord, which would affect over 50 percent of global GDP, is to be avoided. They would have a more awkward time still defending the early days of the GATT. When it was established in 1947, the 23 founding countries comprised only 75 percent of global GDP. See Prudence A. Gordon, "The World Trade Organization: Throwing the Baby out with the Bath Water," Australian Journal of International Affairs, Vol. 50, No. 3 (November 1996), p. 247.