The Effects of the WTO on International Trade - Princeton University

26 downloads 50537 Views 298KB Size Report
Woodrow Wilson School of Public and International Affairs ... groups. The paper assesses the effect of the WTO on trade creation, or the degree to which trade ...
The Old and the New: The Effects of the WTO on International Trade (Draft. Comments welcome)

Soo Yeon Kim Niehaus Center for Globalization and Governance Woodrow Wilson School of Public and International Affairs 445 Robertson Hall Princeton University Princeton, NJ 08544-1013 [email protected] & Department of Government and Politics 3140 Tydings Hall University of Maryland College Park, MD 20742 [email protected]

1

Abstract This paper analyzes the effects of the World Trade Organization (WTO) on trade among its members in its first decade of existence, 1995-2004. The analysis focuses on a key variable: the timing of membership. The paper examines how trade among WTO members was affected by the regime’s accession rules and the timing of members’ entry into the WTO. The paper distinguishes between “standing members,” or those who were members of the GATT as well as the WTO, the “early adopters,” those who entered the regime during the Uruguay Round or in the first year of the WTO’s creation, and the “later entrants” who gained membership via the lengthy and complex accession process after 1995. The empirical analysis evaluates how WTO member trade has fared under the regime, both between members of the same group and between members of different groups. The paper assesses the effect of the WTO on trade creation, or the degree to which trade expanded after states entered into the regime. The results of the analysis show positive but divergent effects among the groups, suggesting that old and new members of the WTO have benefited in different ways from participation in the regime.

2

The Old and the New: The Effects of the WTO on International Trade “…the WTO has been termed a Mercedes Benz without gas.” (Ostry 1997, 202)

The World Trade Organization (WTO) was established in 1995, following the Uruguay Round of negotiations under the General Agreement on Tariffs and Trade (GATT). Its “clear” legal status and mandate was regarded as “the crossing of an important threshold in international trade relations” (Gallagher 2005, 2). International trade governance had been, until then, conducted under the auspices of the GATT, a provisional agreement dating back to 1947. The WTO renewed and incorporated the GATT negotiated in 1994, but in itself is a new legal entity that is not ‘provisional’ as the GATT was but a formal intergovernmental agreement with the status of an international treaty. In comparing the WTO to the GATT, the first obvious difference is the size of the agreement itself: the WTO comprises 60 agreements, compared to ten under the GATT. The WTO consists as well of some 60 different formal councils and committees, while the GATT included less than one-third of this number even in its last years. It formalized a great many of the subjects covered in the GATT in separate agreements, thus vastly expanding the scope of the institutional arrangements. 1 The WTO also linked these

1

Subjects such agricultural trade, trade in textiles and clothing, have been transformed into individual agreements under the WTO, each with its own detailed schedules, footnotes, and annexes. GATT’s Article XX(b), for example, which refers to measures ‘necessary to protect human, animal or plant life or health,’ was recast as the Agreement on Sanitary and Phytosanitary Measures (SPS). The Agreement on Subsidies and Countervailing Measures (SCM) addresses shortcomings of the GATT’s vague definition of an “export subsidy,” which had limited their effectiveness in disciplining discriminatory trade in agriculture. The Agreement on Safeguards prohibits ‘voluntary’ export restrictions that were thinly veiled practices of unscheduled import barriers, but also justified the use of safeguards for a legitimate trade adjustment program. The 3

agreements as a “single undertaking,” requiring Members to accept the full set of agreements concluded in the Uruguay Round. 2

This paper analyzes the impact of the WTO on international trade. Whether the WTO has been successful in expanding trade among its members is important for several reasons. One is that there is a strong presumption that increased trade and trade liberalization foster growth and development. This is especially applicable to developing countries, which comprise the vast majority of new members that have joined the trade body since its creation. 3 If, however, the WTO has had little impact on expanding trade among its members, it places at risk the next and perhaps more important outcome in the causal sequence: economic growth and development. It thus presents a pressing case for examining whether or not, and just how much, trade has flourished for WTO members.

Another equally important reason to analyze systematically the impact of the WTO for international trade is distinctly political. The WTO is the “first construct in a new post-Cold War architecture of international cooperation” (Ostry 1997, 238). The end of the Cold War and the systemic change in political tensions between the United States General Agreement on Trade in Services (GATS) extends GATT rules for trade in “intangibles,” to reflect the importance of the fast-growing service sector, which as of 2004, accounted for almost 20% of global trade. And the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) provides for minimum standards of protection for the intellectual property of WTO Members. 2

Reform of the multilateral system also became a major item on the agenda of the newly minted WTO. Although the word is nowhere to be found in the GATT, it is mentioned 10 times in the WTO Agreement on Agriculture alone, including the first line. It suggests that, more than the GATT, the WTO is more strongly directed toward affecting policy choices in member countries (Gallagher 2005, 8) 3

The transition economies of Eastern Europe may also be included in this group. 4

and the former Soviet Union gives rise to questions regarding the “layer cake” (Ikenberry 2001) of institutions designed to solidify the Western alliance against the threat of Soviet expansion. Reform of existing major international institutions such as the United Nations, the International Monetary Fund, and the GATT, all long-standing legacies of the Cold War, has become a major governance issue for this post-Cold War era of globalization. The large-scale protests that have marked WTO meetings since Seattle 1999 have focused attention on new issues such as labor standards, human rights, and the environment. More relevant to this paper, however, is the greater attention brought to the significant divide that exists between the developed and developing countries in the global trade governance system, especially in the inequality of bargaining power that may prevent the latter from more effectively asserting its interests. It has led to calls for “leveling the playing field,” (Kapstein 2006), going so far as to call for an international social compact that extends basic principles of economic justice to relations between developed and developing countries.

This paper is motivated in great part by the criticisms lodged against the global trade regime. Scenes of protest, antagonism, and resistance toward the WTO and the globalization it represents have been regular occurrences at the organization’s meetings since the 1999 WTO ministerial meeting in Seattle, with the exception of the ministerial meeting in Doha, Qatar in 2001 that officially launched Doha Round, the first round of trade negotiations to be launched under the auspices of the WTO. The ministerial meetings in Cancun in 2003, and Hong Kong in 2005, were similarly attended by such

5

protests. 4 The spectacular displays of protest give rise to a profound puzzle: why so much opposition to an organization that is intended to benefit all countries through free trade? Antagonism towards the WTO, and the globalization it represents, made strange bedfellows of environmentalists, consumer advocates, human rights activists, and labor unions. While each had its own set of complaints, these opponents of the WTO were united on one: “The benefits of the global economy are reaped disproportionately by the handful of countries and companies that set rules and shape markets,” according to Jay Mazur writing in Foreign Affairs about the Seattle meeting. 5 Lack of progress in recent trade talks have led to developing countries increasingly criticizing the Doha Round as not a “development round” but a “market access round” benefiting the developed economies. 6

This paper assesses the institutional and empirical bases of these claims, examining the rules of international trade embodied in the WTO and analyzing trade flows among WTO members in its first decade of existence, comparing it chiefly to its predecessor, the GATT. In terms of institutional rules, the paper focuses on the rules for

4

The Ministerial Conference is the highest decision-making body in the WTO. It holds decision-making authority on all matters under any of the multilateral trade agreements. The Ministerial Conference is mandated to meet at least every two years, bringing together all WTO members countries or customs unions. Under the WTO the Ministerial Conference has been held five times: Singapore, 9-13 December 1996; Geneva, 18-20 May 1998; Seattle, November 30 – December 3, 1999; Doha, 9-13 November 2001; Cancún, 10-14 September 2003; and Hong Kong, 13-18 December 2005. 5

President of the Union of Needletrades, Industrial, and Textile Employees (UNITE) and Chair of the AFL-CIO International Affairs Committee, as of 2000. Developing countries are members of groups such as the G20, the G33, the Africa, Caribbean, Pacific (ACP) Group, Least Developed Countries (LDCs), the Africa Group, the Small Vulnerable Economies (SVEs), the NAMA-11, the Cotton-4, and Caricom. 6

6

accession in the regime to distinguish between “standing members,” or those who were members of the GATT as well as the WTO, the “early adopters,” those who entered the regime soon after its creation, and the “later entrants” who gained membership via the lengthy accession process. The empirical analysis evaluates how each group’s trade has fared under the regime, both with members of the same group as well as other groups. The analysis also concentrates on the extent of trade creation, or the degree to which trade expanded after states entered into the regime. The results of the analysis show that membership in the WTO has had divergent effects on international trade, expanding trade for some members but having weak or even negative effects on others.

New Rules, Old Rules The “designers” of the WTO intended the organization to be widely inclusive, in terms of membership, its scope and domain of application. The WTO advocates universal membership, and any state or customs territory that is fully autonomous in the conduct of trade policy may apply to become a member of the WTO. They may do so in one of two ways, one as a Contracting Party to the GATT and subsequently as an original member of the WTO, and the other through the formal accession process following the establishment of the WTO. Those in the latter group may also be distinguished in terms of the timing of their membership, as the “early adopters” and the “later entrants.”

In spite of the inclusiveness of the WTO’s membership rules, they are far more stringent than the accession requirements under the previous GATT system and intrusive in terms of the liberalization programs that must be adopted by applicants. Original

7

members of the GATT had simply to agree not to enact new legislation that would be inconsistent with its principles and provisions, thus grandfathering existing trade restrictions (Barton, Goldstein, Josling, and Steinberg 2006). Subsequent accession cases, many of them newly independent countries, fell under the auspices of Article XXVI: 5(c), which only required that the country have “full autonomy” to conduct its commercial relations. 7

Accession under the GATT It was the founding members of the GATT that set down the rules for accession of new members into the regime. The 23 original members, or Contracting Parties, examined the terms of GATT accession at the Annecy Round in 1949, immediately following the first Geneva Round. Indeed, one of the main objectives of the meeting was to examine more closely the terms of accession for new members. The Annecy Round established the “Procedures Governing Negotiations for Accession,” to apply to accession outside of tariff negotiations, and the “Model Protocol of Accession.” These provisions specified that accession was to be based on negotiations on tariff schedules.

Accession for prospective members often took less time than it does under the WTO, as accession was approved under the GATT solely on the basis of tariff schedules and as a result once negotiations on tariff schedules were complete it was simply

“If any of the customs territories, in respect of which a contracting party has accepted this Agreement, possesses or acquires full autonomy in the conduct of its external commercial relations and of the other matters provided for in this Agreement, such territory shall, upon sponsorship through a declaration by the responsible contracting party establishing the above-mentioned fact, be deemed to be a contracting party.” 7

8

incorporated into a formal document approving the accession of the particular applicant and submitted for approval to existing members. The Working Party that was designated played only a minimal role under the GATT’s accession procedures. 8 It was limited to fostering discussion on the particular accession case and to provide a forum for raising issues related to the applicant. For the applicant country, it satisfied most of the accession requirements through commitments to bind tariffs at a negotiated level. In this way accession was relatively less cumbersome and complicated under the GATT, reflecting the limited scope of the regime.

Nevertheless, it was also the case that industrial countries dominated accession proceedings, especially when it came to negotiating accession for developing countries (Hoda 2001, 74). In a few cases, accession was highly controversial, such as the most famous case of Japan’s accession. When Japan became a GATT member in 1955, 40 percent of contracting parties invoked Article XXXV to withhold most-favored nation treatment, thus making Japan a key exception in the non-discrimination principle and affecting approximately 40 percent of its exports (Patterson 1966, 285-6). This was the largest number of invocations of Article XXXV recorded against Japan, including important trading states such as France and Great Britain, which feared competition from Japan would hurt domestic markets. This exception to the GATT’s non-discrimination principle nullified much of the benefits and preferential treatment that Japan could have gained from accession.

8

The well-known exception is Japan and the non-market economies. 9

On the whole, however, accession was mostly straightforward and unproblematic, especially for the 64 countries or ex-colonies that acceded automatically under Article XXVI:5(c). It provided a means for former colonies to obtain GATT membership that was somewhat outside the usual accession route. Where colonial powers were willing to sponsor an application by a former colony for membership in the GATT, accession was granted on an automatic basis on the “terms and conditions previously accepted by the metropolitan government on behalf of the territory in question.” 9 That is, tariff schedules previously negotiated by the colonizer on behalf of its former colony, before the latter’s formal membership, would continue to apply. Such provisions concerning former colonies reflected the interests mostly of Europe’s colonizers vis-à-vis their respective overseas territories.

Accession under XXVI:5(c) was thus close to automatic. Former colonies had already been applying GATT provisions in their trade practices on an ad hoc basis and hence the costs of de facto membership were low. The former colony merely needed to get their former colonizer to sponsor their application for GATT membership, and consequently, many former colonies were granted de facto membership indefinitely. 10 De facto status was granted on the condition that GATT members would apply its rules and provisions to de facto members on a reciprocal basis. It thus allowed newly independent states with de facto GATT membership “to benefit from and apply on a reciprocal basis, the provisions of the GATT, and, in particular, the rules for most favored 9

BISD 10S/73.

De facto status for many countries was retained until the establishment of the WTO, which eliminated this membership provision. 10

10

nation treatment, even though they are not formal members.” 11 De facto members received, in turn, many of the benefits of the GATT, in spite of fewer obligations and recourse members than formal members. 12

Accession during the Uruguay Round By the 1980s, and in particular the Uruguay Round, launched in September 1996 at Punta del Este in Uruguay, GATT members and especially the powerful developed countries, directed their efforts toward expanding the agenda of trade liberalization. This meant increasing the scope of GATT negotiations, and in this sense the Uruguay Round was unique in that the negotiating agenda covered the widest range of trade issues. 13 Negotiations now extended to textiles and agriculture, as well as new sectors such as services, intellectual property, and investment. Accession during the Uruguay Round was now to be affected by this new commitment to wide-ranging trade liberalization by GATT members.

Variation in accession cases during the Uruguay Round can be largely explained by the unique importance of the Round itself. It was, indeed, the most important Round in the GATT’s history, as the results would substantially increase the scope of issues to be governed by the trade regime, and trade governance was now to be directed by a formal organization. Contracting parties, and especially the United States, “grew increasingly 11

L/2757 March 1967:2.

For example, while de facto membership participated in tariff negotiations and other GATT sessions, they could not vote nor have recourse to the GATT’s dispute settlement mechanism. 12

13

Negotiations involved 15 different issues/sectors. 11

insistent upon the GATT accession process as a means of ensuring that the countries’ trade regime was consistent with the rules and principles of the system” (Van Grasstek 2001, 20). Increasing importance was attached to international trade regulation in applicant countries, and this was reflected in accession cases during this period. Accession under Article XXXIII and XXVI:5(c) came under increasing scrutiny, subjecting accession applicants to a far more cumbersome and complex process than would have otherwise been the case before the Uruguay Round.

Accession under the WTO Under the newly established WTO, existing members, or contracting parties of the GATT, agreed that all members should commit themselves fully to the implementation of all agreements concluded during the Uruguay Round, precluding the possibility of some members selectively opting out or implementing agreements under lax conditions. Thus existing GATT members were subject to the requirement of a “single undertaking” if they were to automatically accede from the GATT to the WTO. This requirement applied to developing and least developed countries as well as to the developed countries. A significant number of developing and least developed countries struggled with the “single undertaking.”

As was the case for existing GATT members, new applicants to the WTO also had to commit to the “single undertaking.” Unlike the GATT members, however, who in their accession process had only to put forward tariff schedules to gain membership, under the WTO applicants negotiated not only tariff concessions but also commitments in

12

many different sectors. The wide scope of issues now covered as a result of the Uruguay Round means that accession under the WTO is far more complex, with far-reaching and long-term consequences for the applicant country. The inherent complexity of the accession process under the WTO makes the experience far more rigorous and demanding. Indeed, applicants are subject to pre-conditions before membership is formally approved. This “WTO conditionality” involves the requirement that members at the time of accession have a national economy that is consistent with WTO rules. Those who have applied for membership but whose policies and regulations are not consistent with WTO rules, are required to implement reforms in order to make them so.

In shifting trade governance from the GATT to the WTO, accessions are governed by Article XII. 14 The provision itself lacks precision (Abbott, Keohane, Moravcsik, Slaughter, and Snidal 2001). It is vague, general, and contains no specific accessions criteria, with the result that it may be subject to differing interpretations. Such imprecision makes problematic the legal side of the accession process, but has also made the accession process uneven, with each case differing from the next. Article XII provides leeway for demands on the part of existing members. At the extreme, accession becomes less about rule compliance than about the representation of existing members’ interests. Applicant countries may well find themselves with little negotiating power and may be pressured simply to accept the terms of accession presented to them. “Any State or separate customs territory possessing full autonomy in the conduct of its external commercial relations and of the other matters provided for in this Agreement and the Multilateral Trade Agreements may accede to this Agreement, on terms to be agreed between it and the WTO. Such accession shall apply to this Agreement and the Multilateral Trade Agreements annexed thereto.” This provision is from the Marrakesh Agreement Establishing the WTO. 14

13

New accession cases under the WTO involves the establishment of a working party consisting of major trading partners that oversees the application process and stipulates the conditions necessary for the applicant to make its policies WTO-consistent. Prospective members must provide information on legislation relevant to issues within the scope of the WTO, including such important sectors as intellectual property and services as well as conventional trade barriers. It is both a lengthy and complex process, in which the applicants engage in numerous trade negotiations to offer concessions to members in return for the presumed benefits of WTO membership. An applicant’s accession also requires approval by vote of a two-thirds majority of existing members. 15

Classification of WTO Members Given significant differences in accession processes and requirements between the GATT and WTO, members of the trade regime may be broadly classified into three main categories: “standing members,” “early adopters,” and “later entrants.” The Appendix lists the countries in each group and their accession dates for the GATT and/or WTO. The first group may be labeled “standing members,” consisting of former members of the GATT, who obtained membership under the old accession rules and were subsequently invited in the Agreement Establishing the WTO signed in Marrakesh in April 1994, to join the new organization, subject to ratification of the Agreement, to which is attached all other agreements, decisions, and understandings. That is, a GATT contracting party could become an “original” member of the WTO on 1 January 1995 when it ratified the WTO agreement and the linked agreements in a “single undertaking,” The voting requirement for approving membership is the same across the GATT and WTO systems: GATT Article XXXV and WTO Agreement Article XII. 15

14

and accepted and annexed to the GATT 1994 and the GATS its commitment to concessions on trade in goods and services. They comprise for the set of accessions taking place under the WTO.

A second group of countries may be classified as “early adopters” of the WTO. This group consists largely of those countries that gained membership during the Uruguay round and the first year of the WTO (1995), and they are notable in two different ways. First, these developing countries, and in particular those that became members toward the end of the Uruguay Round, joined with a view to wielding influence over the final set of agreements leading to the establishment of the WTO (Gallagher 2005, 13). Participation in the Uruguay round allowed them a voice, even not a powerful one, in negotiating for rights and benefits under the new regime. Second, accession to the WTO held definite advantages to being among the “early adopters,” especially for developing countries. The complex accession process was especially challenging for the least-developed country applicants, and they were granted leeway in meeting the conditions for accession, commensurate with their level of development, financial or trade positions, and administrative and institutional capabilities. 16 In 1995, 21 leastdeveloped countries obtained approval for formal membership, as original members, from the WTO’s General Council through this less stringent process.

The Uruguay Round’s Decision on Measures in Favour of Least-Developed Countries exempted least developed countries from commitments and concessions that are “inconsistent with their individual development, financial and trade needs.” (www.wto.org) 16

15

The third group of WTO members, the “later entrants,” was subject to the more complicated and full-fledged accession process, which entailed negotiating the terms with existing WTO members. Though equal with all other Members once membership has been obtained save for any special conditions negotiated during the accession process, later entrants to the WTO were subject to the full acceptance requirement of all WTO agreements and did not automatically qualify for the implementation concessions that developing country members enjoyed at the time of the WTO’s establishment in 1995. The accession process involved bilateral negotiations with WTO Members, which at times have been more demanding, stringent, and intrusive in their requirements of domestic reform. Negotiations dealt with a host of issues, such as an applicant’s trade policies and practices, including issues such as market access concessions and commitments on goods and services, and legislation for the enforcement of intellectual property rights. Individual Working Parties are assigned to each case to evaluate the reports from applicant economies detailing their commercial policies. Prospective members must also negotiate with their most important trading partners among the WTO members the terms and conditions of bound market access rights for goods and services, which are exchanged a the time of formal accession. Such bilateral negotiations may take years to complete. Though a special case whose accession proceedings began in 1987, China’s accession process took 16 years to complete.

The Marrakesh Agreement Establishing the WTO was signed in April 1994 by 128 members. As of January 1, 1995, when the WTO came into existence, 76 economies had qualified for accession to the WTO according to the conditions for membership. An

16

additional 36 members completed the process involving mostly verification and acceptance of national schedules, fulfilling the obligations for membership and allowing them to accede by 13 December 1995. This brought WTO membership to 112 economies by the end of its first year. Of the 51 economies that have applied for accession, 23 have completed their accession procedures and attained membership: Ecuador (1996), Bulgaria (1996), Democratic Republic of Congo (1997), Mongolia (1997), Panama (1997), Kyrgyz Republic (1998), Latvia (1999), Estonia (1999), Jordan (2000), Georgia (2000), Albania (2000), Oman (2000), Croatia (2000), Lithuania (2001), Moldova (2001), China (2001), Chinese Taipei (2002), Armenia (2003), Former Yugoslav Republic of Macedonia (2003), Nepal (2004), Cambodia (2004), Saudi Arabia (2005), and Vietnam (2007). As of this writing, 31 more economies are queued for accession negotiations.17

Analysis This study builds on a set of controversial findings in the literature that challenge the trade-creating effects of the GATT and WTO. Among current studies, Rose (2004) found that overall, trade “cannot be dependably linked” to GATT/WTO membership. Upon closer examination, Rose (2004), Gowa-Kim (2005), and Subramanian and Wei (2005), found that, when GATT members are divided into industrial and non-industrial countries, GATT has had a positive and significant effect only on trade between the former. That is, the GATT expanded trade among industrial countries but had little if any

Afghanistan, Algeria, Andorra, Azerbaijan, Bahamas, Belarus, Bhutan, Bosnia and Herzegovina, Cape Verde, Equatorial Guinea, Ethiopia, Holy See (Vatican), Iran, Iraq, Kazakhstan, Lao People's Democratic Republic, Lebanon, Libya, Montenegro, Russia, Samoa, Sao Tomé and Principe, Serbia, Seychelles, Sudan, Tajikistan, Tonga, Ukraine, Uzbekistan, Vanuatu, and Yemen. 17

17

impact on trade between non-industrial countries or between industrial and non-industrial countries. This is in spite of the “enabling clause,” or the special and differential treatment accorded to the developing countries within the GATT regime. The unevenness in the effect of GATT on trade flows is a product, inter alia, of the exclusion of agriculture from the GATT regime, the creation of special rules for labor-intensive industries such as textiles and apparel and the import-substituting industrialization policies of LDCs. These findings have been challenged by Tomz, Goldstein and Rivers (2007) and Goldstein, Rivers, and Tomz (2007), who argue that the GATT extended rights and obligations to countries who were not formal members of the regime. Their studies find that inclusive of these members with “institutional standing,” the GATT significantly aided trade expansion in the post-World War II period.

The analysis presented here further develops the arguments and findings in the existing literature in two ways. First, it focuses specifically on the impact of the WTO. Given the establishment of a formal organization to govern trade, the greatly expanded scope of issues now covered under the regime, the question arises as to its consequences for trade among its members, both old and new. Second, the analysis is devoted in particular to examining how developing countries have fared under the regime, differentiated by the timing of their membership. The rationale for distinguishing between “standing members,” “early adopters,” and “later entrants,” is due to the increasingly complex and, in particular, intrusive dimensions of accession proceedings. Trade negotiation rounds under the previous GATT regime emphasized the reduction of “barriers at the border,” as members bargained over tariff rates and quantitative

18

restrictions, and less attention was directed to the compatibility of domestic laws with WTO agreements. The more stringent accession requirements of the WTO, on the other hand, call for the active enactment of legislation that is WTO-consistent, reflecting perhaps the changing view that trade liberalization policies legislated in the domestic political arena must precede any trade-creating benefits conferred by inclusion in the regime. From this perspective, then, the expected trade-creating benefits of the WTO should be greater the later one joins the regime.

Data The sample analyzed consists of pairs of countries, or dyads, observed annually between the years 1948 to 2004, covering the GATT era and the first ten years of the WTO. Trade is measured by the sum of the exports and imports from country a to country b. The trade data, calculated in constant U.S. dollars (2000), were obtained from IMF’s Direction of Trade Statistics (DOTs). 18 Countries comprising the dyads include members of the interstate system, as identified by the Correlates of War 2 Project (COW2) and subject to data availability. 19 The sample is limited to system members as data on the political variables described below exist only for these states. The variable of

18

Trade values were converted to constant U.S. dollars (2000) using the U.S. GDP (chained) price index, obtained from the GPO: http://www.gpoaccess.gov/usbudget/fy05/sheets/hist10z1.xls) 19

The COW2 Project defines a system member as a state that is i) a UN member; or ii) its population exceeds five hundred thousand and it receives diplomatic missions from at least two major powers. Correlates of War Project. 2005. State System Membership List, v2004.1. Online, http://correlatesofwar.org. 19

interest, WTO membership, is defined in terms of formal membership where countries have completed the process of accession. 20 The model is as follows: Log(Tradeijt) = α + β1 Log(GDPi * GDPj)t + β2 Log(GDPPCi * GDPPCj)t + β3 Log(Currency Union)ijt + β4 Log (PTA)ijt + β5 (GSP)ijt + β6 (Democracy)ijt+ β7 (Alliance)ijt + β8 (WTO Years)ijt + β9 (Colonial History *WTO Years)ijt + β10(Standing Members *WTO Years)ijt + β11(Early Adopters *WTO Years)ijt + β12(Later Entrants*WTO Years)ijt + ∑ t δ t Yeart + εijt The dependent variable is the (natural) log of bilateral trade volume: exports from and imports between country i and country j in a given year t. Among the explanatory variables on the right-hand side are the logged products of the two countries’ GDP and per capita GDP (GDPPC), both in constant U.S. dollars (2000). 21 PTA and Currency Union equal one if the two countries belong to the same preferential trade agreement or

20

Differences exist in the way studies have coded for membership. Gowa and Kim’s (2005) GATT membership variable makes reference to Tomz, Goldstein, and Rivers (2007), which argued that institutions such as the GATT extended rights and obligations broadly to nonmembers, creating “institutional standing” for current and former colonies of the contracting parties and some countries in the process of accession. Gowa and Kim code these as GATT members as long as they are system members, that is, as of their year of independence, and states in the process of becoming GATT members. Goldstein et al. constructed their membership variable from GATT archival material, and they have kindly made their data available for this paper. The roster of informal members identified in Gowa and Kim, therefore, may not exactly replicate those in Goldstein et al. The WTO eliminated this provision of de facto membership and required countries either to submit to accession proceedings or lose their benefits.

21

Data on GDP and GDP per capita were obtained from the Penn World Tables, 6.2. 20

share the same currency in a given year, and zero otherwise.22 GSP equals one if one country is a beneficiary of trade preferences given by the other under the Generalized System of Preferences, which were special agreements negotiated under the auspices of the United Nations Conference on Trade and Development (UNCTAD) to aid developing countries expand their trade vis-à-vis the developed world. Democracy equals one if both countries in the dyad score 6 or greater in Polity IV’s democracy index, and Alliance equals one if both countries are members of the same alliance. 23

The variables of interest include WTO Years, which is a dummy variable that equals one for the years since the establishment of the WTO (1995) to 2004, and several interaction variables that combine WTO Years with the three member groups. Standing Members equals one if both countries in the dyad were GATT members who joined before the start of the Uruguay Round (1986); Early Adopters equals one if both countries joined during the Uruguay Round and were members in 1995; and Later Entrants equals one if both countries joined the WTO after 1995 and thus subject to the full-fledged accession process. Each of these member group variables is combined with WTO Years in interaction terms, to capture the effect of the WTO on their respective trade. The variables represent not only each group on its own before and after the WTO 22

Data on PTAs were obtained from Mansfield and Pevehouse (2000) and updated using information from the WTO and Arashiro (2005). Data for Currency Unions were obtained from Rose (2004) and extended using Cohen (2004). Formal alliances include mutual defense pacts, non-aggression treaties, and ententes. Data for democracy were obtained from the Polity IV database: http://www.cidcm.umd.edu/polity/. Data on alliances were obtained from the COW2 Project: http://cow2.la.psu.edu/, version 3.03 (Gibler and Sarkees). Alliance data from 2000 were extended to 2004 for this analysis. 23

21

was created, thus making the WTO a “treatment” for these groups. In addition, the model specifies effects of the WTO on trade between Standing Members & Early Adopters, Standing Members & Later Entrants, and Early Adopters & Later Entrants between 1995 and 2004 (WTO Years). The model includes as well year fixed effects using dummy variables for each year (except one) in the sample. These account for unmeasured factors that are year-specific, such as the oil price shocks of the 1970s and 1980s, and the years toward the end of the Cold War (1989-1991), that are expected to affect the trade of all countries.

In constructing the membership variables, this paper utilizes four different classifications of membership that are found in the literature. The first classification employs the official GATT/WTO date of membership. The second employs Rose’s (2004) classification, which includes colonies paired with their colonizers as part of the GATT’s founding member group. The third classification is that of Tomz, Rivers, and Goldstein (2007), which includes de facto and provisional members as well as formal membership. Finally, the Gowa-Kim (2005) classification utilizes information from both Rose and from Tomz et al. to construct its set of informal and formal members.

The analysis employs fixed effects regression analysis to examine changes in bilateral trade before and after the WTO’s establishment. This is appropriate for two reasons. First, it accounts for heterogeneity among the units, thus controlling for idiosyncratic features of a country-pair’s relations that are not captured by more objective and generalizable factors such as economic size. Second, the comparisons that the

22

analysis seeks to make are longitudinal rather than cross-sectional. That is, the paper focuses on the extent to which trade increases, decreases, or remains unaffected when countries become members of the WTO. It is less concerned with whether WTO members trade more with each other than with non-members, since by 1995 and certainly with the inclusion of more member countries, the vast majority of countries in the international system. This makes comparisons between members and non-members less meaningful, and hence the use of fixed effects analysis is more appropriate to assess the longitudinal impact of trade within each dyad. 24

Results Table 1 presents the results of the analysis. The four columns represent the different membership classifications: 1) official GATT/WTO membership; 2) Rose (2004); 3) Tomz, Rivers, and Goldstein (2007); and 4) Gowa-Kim (2005). Each column reports the effect of WTO membership among members of the main groups--Standing Members, Early Adopters, and Later Entrants and for mixed pairs of various combinations of these groups. By and large, trade among members of each group varied following the creation of the WTO. Pairs consisting of Standing Members increased anywhere from about 7.5 percent to about 12.8 percent depending on the member classification; Early Adopters saw more than a 60 increase in their trade when using the TRG and Gowa-Kim classifications but no significant effect is shown for the less

The use of fixed effects regression analysis, however, precludes inclusion of timeinvariant variables, such as distance, land area, island status, landlocked status, contiguity, common language, and colonial history, which are standard control variables in models of trade. Also excluded from the analysis are the dummy variables for the three member groups’ main effects, since these also do not vary over time. 24

23

inclusive classifications; trade among Later Entrants increased by more than 47 percent. In terms of trade with members of other groups trade between a pair of countries consisting of a Standing Member and an Early Adopter increased anywhere from about 11 percent to almost 32 percent depending on membership classification. Trade among Standing Members and Later Entrants also increased, varying between about 28 to 30 percent, and trade between Early Adopters and Later Entrants increased by 36 to 66 percent depending on membership classification. These estimates are statistically significant, suggesting the membership in the WTO have had a varied but positive impact on trade creation.

Results for the control variables are consistent across the two models and with findings in current studies. Estimates for GDP and GDP per capita are positive and statistically significant, demonstrating the importance of the size of the economy on trade flows. Country pairs sharing the same currency increase their trade by about 57 percent, and this result is statistically significant. Common membership in a preferential trading arrangement also increases trade between two countries, by almost 20 percent. Country pairs in which one country enjoys preferential treatment for its goods under the GSP system has no significant impact on trade. Among the political variables included in this analysis, democracy has a positive effect, increasing trade by 26-28 percent when both countries in a dyad are democracies. Common membership in a formal alliance, however, is inconclusive, as the results show a significant and positive effect under only one classification--the formal membership classification of the GATT/WTO.

24

Table 2 presents the results of the analysis from additionally classifying countries according to level of development, contrasting the effect of WTO membership on various combinations of industrialized and non-industrialized countries within the member groups. 25 They illustrate the distribute consequences of the WTO on trade among its members and focus specifically on how the WTO has affected trade among industrialized and non-industrialized countries. For these mixed pairs of countries, only the group consisting of Standing Members experienced an increase in their trade, ranging from about 14 to 18 percent. Mixed pairs consisting of Standing Members and Later Entrants show no statistically significant change in trade during the WTO period. For industrialized and non-industrialized country pairs consisting of Standing Members and Early Adopters, the WTO years show a decrease in their trade, ranging from 22 to 32 percent depending on membership classification.

Separating out industrialized and non-industrialized country pairs affected changes as well in the effects of other member groups. The remaining Standing Members show a decrease in trade of about 8 percent that is statistically significant; Early Adopters also show a decrease of about 47-50 percent of trade that is statistically significant; and there is no significant effect on trade for Later Entrants. In contrast, pairs consisting of Later Entrants and Early Adopters see an increase in trade of 43-50 percent that is statistically significant for the more inclusive TRG and Gowa-Kim classifications.

The list of industrial countries includes the following 21 countries: Australia, Austria, Canada, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States. 25

25

Conclusion This paper presents an analysis of the effects of the World Trade Organization (WTO) on trade among its members in its first decade of existence, 1995-2004. The analysis focuses on a key variable: the timing of membership. The paper examines how trade among WTO members was affected by the regime’s accession rules and the timing of members’ entry into the WTO. The paper distinguishes between “standing members,” or those who were members of the GATT as well as the WTO, the “early adopters,” those who entered the regime during the Uruguay Round or in the first year of the WTO’s creation, and the “later entrants” who gained membership via the lengthy and complex accession process after 1995. The empirical analysis evaluates how WTO member trade has fared under the regime, both between members of the same group and between members of different groups. The paper assesses the effect of the WTO on trade creation, or the degree to which trade expanded after states entered into the regime. The results of the analysis show positive but divergent effects among the groups, suggesting that old and new members of the WTO have benefited in different ways from participation in the regime. They also indicate that there may be strong distributive consequences of membership in the WTO, as the effect on trade varies as well between mixed pairs of industrialized and non-industrialized country pairs.

26

Table 1. The Effect of the WTO on International Trade : By Group Dependent Variable: Bilateral Trade

(1)

(2)

(3)

(4)

Membership Classification:

WTO

Rose

TRG

Gowa-Kim

GDP

0.886** (0.049) 0.256** (0.048) 0.571** (0.118) 0.197** (0.028) 0.028 (0.031) 0.277** (0.023) 0.117* (0.060) -1.659** (0.133)

0.873** (0.049) 0.269** (0.048) 0.571** (0.117) 0.196** (0.028) 0.027 (0.031) 0.274** (0.023) 0.116 (0.060) -1.634** (0.133)

0.857** (0.049) 0.292** (0.048) 0.574** (0.117) 0.197** (0.028) 0.027 (0.031) 0.263** (0.023) 0.117 (0.060) -1.617** (0.133)

0.856** (0.049) 0.294** (0.048) 0.575** (0.117) 0.197** (0.028) 0.027 (0.031) 0.258** (0.023) 0.116 (0.060) -1.619** (0.133)

Colonial History* WTO Years

-0.499** (0.074)

-0.497** (0.073)

-0.490** (0.071)

-0.490** (0.071)

Standing Members* WTO Years

0.128** (0.038)

0.104** (0.038)

0.077* (0.037)

0.075* (0.037)

Early Adopters* WTO Years

-0.010 (0.130)

0.106 (0.168)

0.618** (0.127)

0.653** (0.140)

Later Entrants*WTO Years

0.476* (0.242)

0.478* (0.242)

0.478* (0.242)

0.479* (0.242)

Standing Members & Later Entrants

0.301** (0.059)

0.296** (0.059)

0.280** (0.058)

0.276** (0.058)

Early Adopters & Later Entrants

0.357** (0.132)

0.409** (0.141)

0.577** (0.153)

0.658** (0.158)

Standing Members & Early Adopters

0.114* (0.046)

0.169** (0.050)

0.281** (0.054)

0.317** (0.055)

Constant

-45.016** (1.722)

-44.626** (1.724)

-44.199** (1.720)

-44.194** (1.718)

Observations

223610

223610

223610

223610

R-squared

0.88

0.88

0.88

0.88

GDP per capita Currency Union PTA GSP Democracy Alliance WTO Years

Robust standard errors in parentheses; * significant at 5%; **significant at 1%; Fixed effects regression analysis using areg in Stata 9.2. Year effects included but not reported.

27

Table 2. Effects of the WTO on Trade: Industrial and Non-Industrial Country Pairs Dependent Variable: Bilateral Trade Membership Classification: GDP GDP per capita Currency Union PTA GSP Democracy Alliance WTO Years Colonial History*WTO Years Standing Members* WTO Years Industrial & Non-industrial Country Pair: Standing Members Early Adopters*WTO Years: Later Entrants*WTO Years Industrial and Non-industrial Country Pair: Standing Members & Later Entrant * WTO Years Later Entrants & Early * WTO Years

Adopters

Industrial and Non-industrial Country Pair: Standing Member & Early Adopter* WTO Years Constant Observations R-squared

(1)

(2)

(3)

(4)

WTO 0.886** (0.051) 0.272** (0.050) 0.534** (0.121) 0.186** (0.028) 0.045 (0.031) 0.299** (0.023) 0.123* (0.059) -1.557** (0.135) -0.467** (0.077) -0.029

Rose 0.888** (0.051) 0.272** (0.050) 0.533** (0.121) 0.182** (0.028) 0.048 (0.031) 0.303** (0.023) 0.124* (0.059) -1.555** (0.136) -0.476** (0.077) -0.057

TRG 0.881** (0.051) 0.281** (0.050) 0.532** (0.120) 0.183** (0.028) 0.049 (0.031) 0.298** (0.023) 0.126* (0.060) -1.552** (0.136) -0.484** (0.076) -0.083*

Gowa-Kim 0.881** (0.051) 0.281** (0.050) 0.532** (0.120) 0.182** (0.028) 0.049 (0.031) 0.297** (0.023) 0.126* (0.060) -1.548** (0.135) -0.485** (0.076) -0.088**

(0.034) 0.138**

(0.033) 0.161**

(0.033) 0.183**

(0.033) 0.184**

(0.044) -0.147 (0.128) 0.333 (0.241) 0.020

(0.043) -0.041 (0.167) 0.328 (0.241) 0.014

(0.043) 0.466** (0.125) 0.330 (0.241) 0.015

(0.043) 0.493** (0.139) 0.326 (0.241) 0.011

(0.064) 0.217

(0.064) 0.261

(0.064) 0.428**

(0.064) 0.503**

(0.131) -0.314**

(0.139) -0.319**

(0.152) -0.221**

(0.157) -0.226**

(0.048) -45.248** (1.776) 223610 0.88

(0.054) -45.327** (1.775) 223610 0.88

(0.060) -45.165** (1.773) 223610 0.88

(0.062) -45.177** (1.772) 223610 0.88

28

Bibliography Abbott, Kenneth W., Robert O. Keohane, Andrew Moravcsik, Anne-Marie Slaughter, and Duncan Snidal. 2001. The Concept of Legalization. In Judith L. Goldstein, Miles Kahler, Robert O. Keohane, and Anne-Marie Slaughter, eds. Legalization and World Politics. Cambridge: The MIT Press, pp. 17-35. Arashiro, Z., C. Marin, and A. Chacoff. 2005. Challenges to Multilateralism: The Explosion of PTA’s. Sao Paulo: Institute for International Trade Negotiations. Barton, John H., Judith L. Goldstein, Timothy E. Josling, and Richard H. Steinberg. 2006. The Evolution of the Trade Regime : Politics, Law, and Economics of the GATT and the WTO. Princeton: Princeton University Press. Cohen, Benjamin J. 2004. The Future of Money. Princeton, N.J.: Princeton University Press. Gallagher, Peter. 2005. The First Ten Years of the WTO: 1995-2005. Cambridge: Cambridge University Press. Gibler, Douglas M., and Meredith Sarkees. Measuring Alliances: the Correlates of War Formal Interstate Alliance Data set, 1816-2000. Journal of Peace Research, forthcoming. Goldstein, Judith L., Douglas Rivers and Michael Tomz. 2007. Institutions in International Relations: Understanding the Effects of the GATT and the WTO on World Trade. International Organization 61, 1 (January): 37-67. Gowa, J., and S. Y. Kim. 2005. An Exclusive Country Club - the Effects of the Gatt on Trade, 1950-94. World Politics 57 (4):453-78. Hoda, Anwarul. 2001. Tariff Negotiations and Renegotiations under the GATT and the WTO: Procedures and Practices. Cambridge: Cambridge University Press. Ikenberry, G. John. 2001. After Victory : Institutions, Strategic Restraint, and the Rebuilding of Order after Major Wars. Princeton, N.J. ; Oxford: Princeton University Press. Kapstein, Ethan B. 2006. Economic Justice in an Unfair World: Toward a Level Playing Field. Princeton: Princeton University Press. Mansfield, Edward D., and Jon C. Pevehouse. 2000. Trade Blocs, Trade Flows, and International Conflict. International Organization 54 (4):775-808. Mazur, Jay. 2000. Labor's New Institutionalism. Foreign Affairs January/February.

29

Ostry, Sylvia. 1997. The Post-Cold War Trading System: Who's on First? Chicago: University of Chicago Press. Patterson, Gardner. 1966. Discrimination in International Trade: The Policy Issues, 1945-1965. Princeton: Princeton University Press. Rose, A. K. 2004. Do We Really Know That the WTO Increases Trade? American Economic Review 94 (1):98-114. Subramanian, Arvind and Shang-Jin Wei. 1995. The WTO Promotes Trade, Strongly but Enevenly. CEPR Discussion Paper Series No. 5122. Tomz, Michael, Douglas Rivers, and Judith Goldstein. 2007. Membership Has Its Privileges: The Impact of GATT on International Trade. American Economic Review, forthcoming.

30

Appendix 1. WTO Member Groups Standing Members (GATT & WTO Entry Dates) Australia Belgium Canada France Luxembourg Netherlands United Kingdom United States of America Cuba South Africa India Norway Zimbabwe Myanmar Sri Lanka Brazil New Zealand Pakistan Chile Haiti Indonesia Greece Sweden Dominican Republic Finland Denmark Nicaragua Italy Germany Peru Turkey Austria Uruguay Japan Ghana Malaysia Nigeria Sierra Leone Tanzania Portugal Israel Uganda Trinidad and Tobago Gabon Kuwait

1/1/1948 1/1/1948 1/1/1948 1/1/1948 1/1/1948 1/1/1948 1/1/1948

1/1/1995 1/1/1995 1/1/1995 1/1/1995 1/1/1995 1/1/1995 1/1/1995

1/1/1948 1/1/1948 6/13/1948 7/8/1948 7/10/1948 7/11/1948 7/29/1948 7/29/1948 7/30/1948 7/30/1948 7/30/1948 3/16/1949 1/1/1950 2/24/1950 3/1/1950 4/30/1950 5/19/1950 5/25/1950 5/28/1950 5/28/1950 5/30/1950 10/1/1951 10/7/1951 10/17/1951 10/19/1951 12/6/1953 9/10/1955 10/17/1957 10/24/1957 11/18/1960 5/19/1961 12/9/1961 5/6/1962 7/5/1962 10/23/1962

1/1/1995 4/20/1995 1/1/1995 1/1/1995 1/1/1995 3/5/1995 1/1/1995 1/1/1995 1/1/1995 1/1/1995 1/1/1995 1/1/1995 1/30/1996 1/1/1995 1/1/1995 1/1/1995 3/9/1995 1/1/1995 1/1/1995 9/3/1995 1/1/1995 1/1/1995 1/1/1995 3/26/1995 1/1/1995 1/1/1995 1/1/1995 1/1/1995 1/1/1995 1/1/1995 7/23/1995 1/1/1995 1/1/1995 4/21/1995 1/1/1995

10/23/1962 5/3/1963 5/3/1963

3/1/1995 1/1/1995 1/1/1995

Central African Republic Burkina Faso Cameroon Congo Chad Cyprus Spain Benin Senegal Mauritania Madagascar Côte d'Ivoire Jamaica Niger Kenya Togo Malawi Malta The Gambia Burundi Rwanda Guyana Switzerland Yugoslavia Barbados Korea, Republic of Argentina Poland Ireland Iceland Egypt Mauritius Zaire Romania Bangladesh Singapore Hungary Suriname Philippines Colombia Zambia Thailand Maldives Belize

5/3/1963 5/3/1963 5/3/1963 5/3/1963 7/12/1963 7/15/1963 8/29/1963 9/12/1963 9/27/1963 9/30/1963 9/30/1963 12/31/1963 12/31/1963 12/31/1963 2/5/1964 3/20/1964 8/28/1964 11/17/1964 2/22/1965 3/13/1965 1/1/1966 7/5/1966 8/1/1966 8/25/1966 2/15/1967 4/14/1967 10/11/1967 10/18/1967 12/22/1967 4/21/1968 5/9/1970 9/2/1970 9/11/1971 11/14/1971 12/16/1972 8/20/1973 9/9/1973 3/22/1978 12/27/1979 10/3/1981 2/10/1982 11/20/1982 4/19/1983 10/7/1983

5/31/1995 6/3/1995 12/13/1995 3/27/1997 10/19/1996 7/30/1995 1/1/1995 2/22/1996 1/1/1995 5/31/1995 11/17/1995 1/1/1995 3/9/1995 12/13/1996 1/1/1995 5/31/1995 5/31/1995 1/1/1995 10/23/1996 7/23/1995 5/22/1996 1/1/1995 7/1/1995 1/1/1995 1/1/1995 1/1/1995 7/1/1995 1/1/1995 1/1/1995 6/30/1995 1/1/1995 1/1/1995 1/1/1995 1/1/1995 1/1/1995 1/1/1995 1/1/1995 4/30/1995 1/1/1995 1/1/1995 5/31/1995 1/1/1995

31

Early Adopters (GATT & WTO Entry Dates) Hong Kong Mexico Antigua and Barbuda Morocco Botswana Lesotho Tunisia Venezuela Bolivia Costa Rica Macao El Salvador Guatemala Mozambique Namibia Mali Swaziland Saint Lucia Czech Republic Slovak Republic Dominica

4/23/1986 8/24/1986

1/1/1995 1/1/1995

3/30/1987 6/17/1987 8/28/1987 1/8/1988 8/29/1990 8/31/1990 9/8/1990 11/24/1990 1/11/1991 5/22/1991 10/10/1991 7/27/1992 9/15/1992 1/11/1993 2/8/1993 4/13/1993 4/15/1993 4/15/1993 4/20/1993

1/1/1995 1/1/1995 5/31/1995 5/31/1995 3/29/1995 1/1/1995 9/12/1995 1/1/1995 1/1/1995 5/7/1995 7/21/1995 8/26/1995 1/1/1995 5/31/1995 1/1/1995 1/1/1995 1/1/1995 1/1/1995 1/1/1995

Saint Vincent & the Grenadines Fiji Brunei Darussalam Bahrain Paraguay Grenada United Arab Emirates Guinea Bissau Saint Kitts and Nevis Liechtenstein Qatar Angola Honduras Slovenia Guinea Djibouti Papua New Guinea Solomon Islands

5/18/1993 11/16/1993 12/9/1993 12/13/1993 1/6/1994 2/9/1994

1/1/1995 1/14/1996 1/1/1995 1/1/1995 1/1/1995 2/22/1996

3/8/1994 3/17/1994

4/10/1996 5/31/1995

3/24/1994 3/29/1994 4/7/1994 4/8/1994 4/10/1994 10/30/1994 12/8/1994 12/16/1994 12/16/1994 12/28/1994

2/21/1996 9/1/1995 1/13/1996 11/23/1996 1/1/1995 7/30/1995 10/25/1995 5/31/1995 6/9/1996 7/26/1996

Later Entrants Ecuador Bulgaria Democratic Republic of the Congo Mongolia Panama Kyrgyz Republic Latvia Estonia Jordan Georgia Albania Oman

1/21/1996 12/1/1996 1/1/1997 1/29/1997 9/6/1997 12/20/1998 2/10/1999 11/13/1999 4/11/2000 6/14/2000 9/8/2000 11/9/2000

Croatia Lithuania Moldova China Chinese Taipei Armenia Former Yugoslav Republic of Macedonia (FYROM) Nepal Cambodia Saudi Arabia Viet Nam

11/30/2000 5/31/2001 7/26/2001 12/11/2001 1/1/2002 2/5/2003 4/4/2003 4/23/2004 10/13/2004 12/11/2005 1/11/2007

32