I thank Robert Keohane and Anne-Marie Slaughter as well as the editors of International Organization for their helpful comments and suggestions.

1. Canada–Mexico–United States: North American Free Trade Agreement, done 8–17 December 1992, entered into force on 1 January 1994, 32 I.L.M. 289 and 605.

2. General Agreement on Tariffs and Trade, opened for signature on 30 October 1947, 61 Stat. A3, T.I.A.S. No. 1700, 55 U.N.T.S. 187.

3. Agreement Establishing the World Trade Organization, done 15 April 1994, entered into force 1 January 1995. World Trade Organization 1995. The WTO agreement incorporates by reference a number of multilateral trade agreements, including GATT 1994, as well as agreements and understandings ancillary to GATT 1994. I use the term WTO agreement to refer to the Agreement Establishing the WTO plus the multilateral trade agreements and ancillary texts, unless otherwise indicated expressly or by context.

4. For a definitive description of GATT as a soft-law system, see Long 1985. I traced the evolution from soft to hard legalization in the GATT/WTO context and undertook a detailed analysis of that trend in Abbott 1997a.

5. Second-generation trade barriers refer to nontariff governmental measures directed toward trade, such as quotas, export subsidies, antidumping measures, and voluntary restraints. Third-generation trade barriers refer to internal government regulatory measures not specifically directed toward trade, but which may distort international competition, such as services regulatory measures, intellectual property norms (over- or under-protection), and competition policies (over- or under-regulatory). See, for example, Cottier 1992; and Hillman 1997.

6. On the Tokyo Round negotiations, see Glick 1984.

7. For background regarding the objectives of the Uruguay Round negotiations, see Jackson 1989; Petersmann and Hilf 1988; and Hudec 1993. For details concerning the negotiating history, see Stewart 1993–99.

8. Free Trade Agreement Between Canada and the United States of America, signed on 2 January 1988, entered into force on 1 January 1989, 1989 Can. T.S. No. 3.

9. For detailed comparison of underlying social and economic policies, see, for example, Abbott 1992a, b and 1997b. For general background regarding the EU legal and social framework, see Weiler 1991. For a review of EC governance preceding the Single European Act and subsequent reforms, see Riesenfeld 1974.

10. Moreover, an empirical comparison of the two arrangements would require a large-scale investment of resources well beyond the scope of the present project.

11. See Fried 1998; and Schoenbaum 1998.

12. Kahler, this issue.

13. Other counter-examples to the NAFTA model of legalization are the NAFTA supplemental agreements on the environment and labor. In preliminary drafts of this article I suggested that these soft agreements have had a very modest effect on the conduct of the parties and that their major impact has been in promoting transnational cooperation among NGOs. To keep this article to a manageable length, I have excluded discussion of these supplemental agreements.

14. Keohane, Moravcsik, and Slaughter, this issue.

15. See Nadal 1997; and Ramos Uriarte 1995.

16. Abbott 1993a.

17. Johnson 1994.

18. Kennish 1995.

19. Hurtig 1992.

20. Ramos Uriarte 1995, 90–91.

21. Keohane, Moravcsik, and Slaughter, this issue.

22. Abbott 1992a.

23. Abbott 1997b, 910–12.

24. Barlow and Campbell 1993.

25. Although both Canada and Mexico might see advantages to having regional regulatory institutions in which each country was a co-equal with the United States, this prospect was most unlikely, and thus there was little incentive for establishing a regional regulatory/bureaucratic framework.

26. Haas 1958.

27. Abbott 1995a.

28. See interview observations of the Japanese business sector reported in Abbott 1995a, chap. 8.

29. As noted infra tan 49, the very high level of precision regarding rules of origin for automobiles and parts addressed a major Canadian government concern over alleged prior misuse of CUSFTA rules of origin by the United States.

30. Abbott 1997a, 45–46.

31. The annexes to NAFTA include a somewhat higher level of detail in relation to party legislative measures than set forth in CUSFTA.

32. As is well known, GATT 1947 entered into force as a provisional agreement in contemplation of a more extensive set of trade rules that would have been established by the Havana Charter for the International Trade Organization. GATT 1947 was framed largely as a broad set of negative rules, prohibiting governments from engaging in certain types of discriminatory and trade restrictive conduct. The history of GATT 1947 through to the Uruguay Round was a process of refining broad rules into more precise obligations. This process took place most particularly in the Tokyo Round, which culminated in 1979 and resulted in a series of codes (or supplemental agreements) covering a number of subject matter areas already addressed in general terms by GATT 1947. The GATT contracting parties negotiated more precise rules because the existing general provisions of GATT 1947 did not provide an adequate basis on which governments could demand compliance and were not capable of sufficiently precise application by dispute-settlement panels. For example, the GATT contracting parties were considerably dissatisfied with the application of antidumping and countervailing duty measures by the United States (and the EU); there was considerable concern with the EC's flexible interpretation of GATT export subsidy provisions; and GATT panels had considerable difficulty interpreting the safeguards provisions of GATT 1947 insofar as they related to health and safety measures. See texts cited in note 7.

33. Regarding the structure of the U.S. trade policy apparatus that, because of the small number of USTR negotiators, virtually assures that the same individuals will participate in important negotiations, see Abbott 1997b, 941–42. It was accepted that NAFTA negotiations diverted attention from the Uruguay Round negotiations because the same individuals at USTR could not be negotiating in two fora at the same time. Ibid. The Canadian trade policy apparatus operates along much the same lines.

34. Abbott forthcoming.

35. The Tokyo Round Antidumping Code, the Subsidies and Countervailing Duties Code, and the Technical Standards Code each represented the efforts of the GATT contracting parties to clarify obligations that were generally embodied in provisions of GATT 1947. The Uruguay Round negotiations were in substantial measure devoted to further clarifying the general rules in the aforementioned areas, as well in the areas of import licensing, rules of origin, sanitary and phytosanitary measures, agricultural subsidies, and trade-related investment preferences. The Uruguay Round also added precise new rules regarding third-generation trade barriers in the areas of services and trade-related aspects of intellectual property rights.

36. Coase suggested that transaction costs associated with the bargaining process may have a significant impact on economic efficiency. Coase 1960. Robert Hudec observes that, “A rule-based system is the most resource-efficient way to resolve conflicts with other countries.” Hudec 1999. 10.

37. If the NAFTA parties agree that tariffs will be phased out between them, but do not indicate which tariffs will be phased out over what period of time, the chances are great that they will need to bargain ex post over the specific mechanics of tariff reduction. If the parties agree that imports from third countries will not enjoy tariff-free transit between the parties, but do not agree on the conditions under which third-party goods will be deemed to have been transformed into regionally originating goods, then the chances are great that the parties will need to bargain ex post over rules of origin and criteria for transformation.

38. If an increase should occur in the frequency of disputes because governments are in a better position to assert claims concerning their entitlements, then the precise terms on which such claims are based should reduce the duration of these disputes. Precise terms should provide better guidance to parties as to the extent of their obligations, and this should decrease the frequency of disputes over the nature of implementation obligations. Paradoxically, however, precise terms may also encourage a greater frequency of ex post dispute-settlement actions as parties become more certain about their capacity to vindicate perceived rights. Yet, if the nature of obligations is precisely defined, it would also be expected that parties would more frequently settle disputes in the course of consultations. While more precise rules might at least initially increase the frequency of dispute-settlement actions, it is also foreseeable that such actions would be of shorter duration (and thus result in relatively low transaction costs). As governments became more certain that the outcome of disputes would be strongly correlated to precise legal rules, the necessity for pursuing claims to a litigious end would likely diminish. Abbott 1997a, 43–44. Empirical support for the proposition that more precise rules may lead to an increased frequency of disputes, but that these disputes may be substantially more prone to settlement, can be found in Hudec 1999, 25–27. Hudec found that in the first three years of operation of the WTO, the frequency with which consultations had been initiated was greater than in respect to the GATT 1947 dispute-settlement system, but that a significantly lower percentage of disputes resulted in the issuance of panel reports.

39. The effect of uncertainty on private operators is considered in the next section.

40. We might assume that the ex ante costs of negotiating an imprecise trade agreement would be lower than the ex ante costs of negotiating a precise agreement. The government negotiators involved would have fewer tasks to accomplish. Agreement would be easier as each government considered itself to retain flexibility in implementation. There are, however, caveats to this premise.

41. Recall that bargaining costs not only relate to costs imposed on governments to conduct negotiations but also include delays in the implementation of commitments assumed to have beneficial economic welfare effects.

42. Hudec 1999, 11 (“A rule-based system is … the most effective way to negotiate and capture desired policy changes in achievable incremental steps”).

43. This would include briefings and study on the issues, arranging negotiating schedules and venues, and so forth.

44. In the context of the WTO negotiating round, this is generally referred to as the possibility of providing cross-concessions. There is a recognized risk that a single set of trade negotiations may become overburdened by the incorporation of disparate subject matter, and trade negotiators must take this risk into account when establishing a negotiating agenda.

45. There are, of course, exceptions to this general rule. A small proportion of business operators prefer to operate in high-risk environments that may present higher profit opportunities. Such operators may forgo purchasing risk insurance that would reduce the potential for profit. Abbott 1997b, 912–15.

46. Hudec 1999, 10 (“A rule-based system creates the most predictable conditions for business decisions”).

47. The availability of enforceable legal rules is a point repeatedly made by Alan Greenspan in distinguishing successful and unsuccessful economies.

48. Coase postulated that uncertainty over the meaning of legal rules would increase private transaction costs as enterprises were forced to bargain under conditions of uncertainty. Coase 1960.

49. Economists generally refer to these additional steps under the rubric of increasing risk premiums.

50. The United States interpreted the parts roll-up rules in a way that increased the local content requirement for transformation to duty-free status, making it less attractive for Japanese automobile companies to locate in Canada and export to the United States. See Customs Rules that Canadian Honda Civics Failed to Meet Content Standards Under FTA, BNA International Trade Reporter, 4 March 1992, 384Google Scholar.

51. Hudec notes in the context of the WTO that a rule-based system helps to cement one's own liberal trade policies against the internal political pressures of protectionism. Hudec 1999, 10. Internal political pressures provide one of the main reasons why governments pursue strategic behaviors, though external political pressures might also underlay such behaviors.

52. Ibid.

53. If the terms of the arrangement are antithetical to third countries, the parties to the arrangement might have a preference for nontransparency. Transparency may be more beneficial when the interests of third countries are adequately addressed by a trade arrangement.

54. Abbott 1995a, chap. 3.

55. Bankers, for example, might prefer a situation in which they could continuously lobby for rules expanding their scope of action, even if this required tolerating uncertainty concerning other rules. Oil refiners might prefer more precise ex ante rules that do not leave open the prospects for subsequent legislation (that might, for example, require major additional investments in environmental control equipment).

56. Weiler 1991, referenced extensively.

57. Much of this enhanced transparency has resulted from the development and use of Internet resources.

58. Abbott 1990. The long-running banana dispute at the GATT/WTO involved a nontransparent decision-making process and regime and is but one reflection of a difficult relationship between the EU and the GATT/WTO that has persisted since the formation of the Community.

59. The APEC Secretariat views the dissemination of information concerning the national laws of APEC members as an important part of its role. This, however, does not constitute providing information about the terms and operation of a regional arrangement; instead it is more in the nature of performing ordinary functions that some APEC members have failed to perform.

60. Kahler, this issue.

61. NAFTA, art. 105.

62. See, for example, NAFTA, art. 301, “Each party shall accord national treatment.”

63. See, for example, NAFTA, art. 712(1), Right to Take Sanitary and Phytosanitary Measures.

64. NAFTA, chap. 8.

65. NAFTA, art. 2104.

66. Some reduce the discretion of the parties, as in the balance-of-payments safeguard. NAFTA, art. 2104.

67. These withdrawn concessions are expected to be equivalent to the level of concessions lost in consequence of the breach. NAFTA, art. 2019(1).

68. See Weiler 1991, 2419 et seq., and Riesenfeld 1974.

69. Compare Jackson 1997 with Bello 1996.

70. Art. 2205.

71. WTO agreement, art. XV. International agreements sometimes include provisions authorizing withdrawal, and a right to withdraw may otherwise be inferred. If not expressly stated, an international agreement is considered not to permit withdrawal, unless this right may be inferred from the context. See Vienna Convention on Law of Treaties, art. 56.

72. See discussion in Weiler 1991, 2412.

73. Nadal 1997.

74. Ramos Uriarte 1995.

75. Abbott 1997a. If NAFTA were to incorporate a higher level of compulsion, a similar appellate mechanism or other arrangement might be required.

76. Whether Canadian or Mexican negotiators would have agreed to a more strictly binding disputesettlement arrangement is superfluous under these conditions.

77. See Kahler, this issue; and Fried 1998.

78. Abbott 1997a, 46–47.

79. See, for example, U.S. withdrawal from the compulsory jurisdiction of the International Court of Justice during the pendency of the Nicaragua proceedings, and the text of Section 301 of the Trade Act of 1974, as amended, which specifically authorizes the USTR to take actions inconsistent with U.S. obligations under international trade agreements.

80. See the introduction to this issue. A second article by these authors refers to delegation in more limited terms, that is, to the delegation of authority for the settlement of disputes. Keohane, Moravcsik, and Slaughter, this issue.

81. See Weiler 1991, 2466.

82. NAFTA, art. 2008.

83. Ibid., art. 2009 and 2011. A panelist who is not on the roster may be proposed but is subject to peremptory challenge. Art. 2011(3).

84. Ibid., art. 2018(1).

85. Ibid., art. 2019.

86. Ibid., art. 2018(2). “Whenever possible, the resolution shall be non-implementation or removal of a measure not conforming with this Agreement or causing nullification or impairment… or, failing such a resolution, compensation.”

87. Chapter 19 contains a weak “stabilization” clause, largely directed at transparency.

88. NAFTA, art. 1904(5).

89. Ibid., art. 1904(9).

90. Ibid., arts. 1115, et seq. See Abbott 1995a, 102.

91. Ibid., art. 1136(4).

92. Ibid., art. 1136(6), referring to the ICSID Convention, the New York Convention on the Recognition and Enforcement of Foreign Arbitral Awards, and the Inter-American Convention on International Commercial Arbitration.

93. See, for example, New York Convention on the Recognition and Enforcement of Foreign Arbitral Awards, art. V.

94. Ibid., art. 1135.

95. Ibid., art. 1136(5). This may result in a recommendation that a party comply with the award.

96. A treaty may have direct effect in the law of a state that is party to it. See Riesenfeld and Abbott 1994a. If a treaty (or part of a treaty) is directly effective, it may be relied on in the courts of a state as a source of rights. The question whether a treaty has direct effect has both an international law and national law component. A treaty may be intended by its parties to be capable of direct effect, as determined by an examination of its terms and context under international law. However, whether a treaty that is capable of direct effect under international law will be given that effect in national law is determined by the constitution of the state in which the treaty is invoked. Some national constitutions permit treaties to be directly effective and others do not. Both the U.S. and Mexican constitutional systems permit treaties to have direct effect, whereas the Canadian constitution (modeled on the British constitutional system) does not. See notes 102–103. In some states in which a treaty may be given direct effect, the national parliament or legislature may deny direct effect to it in its approving or implementing legislation.

97. It may have been understood by U.S. trade negotiators that NAFTA implementing legislation would deny the agreement direct effect in U.S. courts, and this may argue that neither Canada nor the United States intended the agreement to be directly effective (since Canadian law bars direct effect). The U.S. Supreme Court has held (as has the European Court of Justice) that whether another state permits an agreement to be directly effective does not determine whether it will be directly effective under the U.S. Constitution (i.e. there is no test of reciprocity). Ware v. Hylton, 3 U.S. (3 Dall.) 199 (1796). The failure of the Parties to proscribe direct effect under national law in the body of the NAFTA, the possibility of direct effect in Mexico, and the U.S. decision to expressly legislate against direct effect each suggest that the agreement might be directly effective absent specific national legislative action.

98. NAFTA, art. 2021.

99. The North American Agreement for Environmental Cooperation (NAAEC) Secretariat factual report procedure provides an alternative outlet for private actors but does not result in enforceable decisions.

100. Ibid. 102(a) provides that “no provision of the Agreement, nor the application of any such provision to any person or circumstance, which is inconsistent with the law of the United States shall have effect.” Though this appears to leave room for the direct application of provisions of NAFTA that would not amend existing federal legislation, the Implementation Act at 102(c) provides that “no person other than the United States (1) shall have any cause of action or defense under (A) the Agreement or by virtue of Congressional approval thereof, or (B) the North American Agreement on Environmental Cooperation or the North American Agreement on Labor Cooperation; or (2) may challenge, in any action brought under any provision of law, any action or inaction by any department, agency, or other instrumentality of the United States, any State, or any political subdivision of a State on the ground that such action or inaction is inconsistent with the Agreement, the North American Agreement on Environmental Cooperation, or the North American Agreement on Labor Cooperation.” See Abbott 1993b (written prior to publication of the NAFTA Implementation Act).

101. In addition to its general statement on direct effect, the NAFTA Implementation Act states that ”nothing in this Act shall be construed—(A) to amend or modify any law of the United States, including any law regarding—(i) the protection of human, animal, or plant life or health, (ii) the protection of the environment.” Ibid., 102(a)(2).

102. Gal-Or 1998.

103. Ibid. In approving and implementing NAFTA Parliament indicated that Canadian courts should interpret domestic implementing legislation consistently with NAFTA. Fried 1994, n.78.

104. See Weiler 1991, 2413 et seq.; and Riesenfeld 1974. In relation to secondary legislation, for example, a directive that obligates the member states to bring national law into conformity with EU rules generally may only be directly relied upon if the national legislature has failed to implement the directive within the prescribed time period.

105. The potential constraints that the U.S. constitution imposes upon the formation of regional political institutions and a regional judicial institution are analyzed in Abbott 1995b.

106. Regarding Canada, see Gesser 1998. Regarding Mexico, see, for example, Ramos Uriarte 1995, 90–92.

107. U.S. Constitution, art. II, sec. 2.

108. Ibid., art. VI, cl. 2.

109. Riesenfeld and Abbott 1994b. Perhaps the widest delegation of decision-making power by the United States to an international institution is the delegation of power to the UN Security Council to make decisions regarding threats to the peace, including the power to authorize the use of force to address such threats. The United States maintains a veto at the Security Council, and it does not face the possibility of decisions being made and carried out without its consent. U.S. participation in the Security Council was approved by the Senate in ratifying the UN Charter and by the Congress in the UN Participation Act (UNPA). There is a potential conflict between the power of the president to vote in favor of the use of force in the Security Council and the residual constitutional power of the U.S. Congress to declare war. This conflict surfaced during the Gulf War when President Bush asserted the right to use U.S. military force based on Security Council action, which right was challenged by members of Congress. Without conceding the point, the president sought and received consent from Congress to the use of force against Iraq. For details, see Abbott 1995b, 148–49.

110. Reid v. Covert, 354 U.S. 1223 (1957)Google Scholar.

111. INSv. Chadha, 462 U.S. 817 (1983)Google Scholar.

112. The question of whether the procedure used by the president and Congress in approving NAFTA comported with the requirements of the Constitution was discussed at length by a federal district court that rejected a constitutional challenge in Made in the USA v. United States, 56 F. Supp. 2d 1226 (N.D. Ala. 1999)Google Scholar.

113. Bermann 1998.

114. See, for example, Carman 1997; and Gesser 1998.

115. Carman 1997, 7.

116. NAFTA modestly expanded the scope of the extraordinary challenge procedure that permits a party to challenge a panel determination when an abuse of process is alleged to have occurred. Fried 1994, 48.

117. CUSFTA, arts. 1608, 1806–1807.

118. See CUSFTA, art. 1608; and Kennish 1995, 16–17.

119. Ramos Uriarte 1995, 88–92.

120. Ibid., 88–90.

121. Ibid., 98.

122. Ibid., 98–99, 107–109. Note that the Mexican government faced constitutional obstacles to concluding the investor-to-state arbitration provisions of NAFTA and may yet face constitutional challenges to this aspect of the agreement. Ibid., 90–91.

123. Fried 1994, 48–49.

124. See, for example, Treaty Between the United States of America and the Republic of Senegal Concerning the Reciprocal Encouragement and Protection of Investment, done December 1983, entered into force 25 October 1990, at art. VII, <http://www.mac.doc.gov:80/tcc/data/commerce-html/tcc-2/Senegal.html>.

125. Kennish 1995, 7–8.

126. Ibid., 1–3.

127. Judges may invoke “political question” doctrines to avoid deciding cases.

128. Abbott 1993b.

129. Keohane, Moravcsik, and Slaughter, this issue.

130. See, generally, U.S. Executive Branch 1997.

131. For example, following the collapse of the peso on international currency markets in 1982, Mexico nationalized the Mexican banking system and imposed a two-tier system of exchange controls designed to curtail capital flight and restrict imports. See Buira 1983; Hobbs 1990; Philip 1985; and Skiles 1991. The 1982 peso crisis resulted in a substantial enhancement of government intervention in the domestic marketplace, at least for the short term. See, for example, Buira 1983, 54–60. State control over private-sector activity was intensified as the means to bring the domestic economic situation into equilibrium. In 1982 the Mexican government was not without international legal constraints responding to the peso crisis. Specifically, in order to satisfy the International Monetary Fund in respect to emergency borrowing, the Mexican government undertook to meet criteria regarding budget deficit control and agreed to other fiscal and monetary constraints. See Hobbs 1990, 187–89. In 1982 the Reagan administration made a $600 million loan to Mexico from the Exchange Stabilization Fund, the same fund that was used to extend loans to Mexico in 1994. Covey 1996.

132. Mexico announced its request for the formation of a panel regarding its cross-border trucking dispute with the United States on 23 September 1998. Mexico Seeks NAFTA Arbitration Panel on U.S. Resistance to Mexican Truckers, BNA International Trade Reporter, 30 September 1998, 1645Google Scholar. As of this date, the panel has not yet been constituted.

133. Lowenfeld 1996. This legislation was challenged by the EU in the WTO, and a temporary negotiated settlement was reached between the EU and the United States.

134. Clinton Mexico Trip to Include Little Discussion of Trade Issues, Americas Trade, 11 February 1999, 1Google ScholarPubMed.

135. A split-run periodical is one that is produced with different content in two markets. Canada sought to effectively prohibit U.S. magazine publishers from selling U.S.-generated works with Canadian advertisement in Canada. Canada's theory was that U.S. publishers were able to sell advertising space in Canada at unfairly low cost because U.S. advertisers had in essence already paid for the production of magazine content. The WTO Appellate Body found these Canadian restrictions to violate GATT 1994. WTO Appellate Body, Canada—Certain Measures Concerning Periodicals, 1994. WT/DS31/AB, 30 June 1997 <http:///www.wto.org>.

136. Final Report of the Panel in the Matter of Tariffs Applied by Canada to Certain U.S.-Origin Agricultural Products, 1997 BDIEL AD LEXIS 24 (2 December 1996).

137. Final Panel Report, in the Matter of the U.S. Safeguard Action Taken on Broom Corn Brooms from Mexico (USA-97–2008–01), 30 January 1998.

138. See index of proceedings and reports available at <http://www.nafta-sec.org>.

139. ICSID Additional Facility, Robert Azinian et al. v. United Mexican States, Case No. ARB (AF)/97/2, 1 November 1999 <http://www.worldbank.org/icsid/cases>.

140. Lutz and Trice 1998.

141. Peter Menyasz, NAFTA Chapter 11 Provisions Said to Threaten Environmental Protection Rights, BNA International Trade Reporter, 7 July 1999, 1146Google Scholar.

142. NAFTA Ministers Lock Horns on Investor, Environment/Labor Issues, Americas Trade, 6 May 1999. 1Google Scholar.

143. Menyasz, Chapter 11 Provisions.

144. See Lutz and Trice 1998; and NAFTA Ministers Lock Horns.

145. Keohane, Moravcsik, and Slaughter, this issue.

146. In the period from 1994 through 1996 covered by the U.S. Executive Branch Study, there was a substantial increase in two-way trade between the United States and Mexico. The share of U.S. imports in the Mexican market also substantially increased as did the share of Mexican imports in the U.S. market. Economic movements in NAFTA services markets were not significant during the same three-year measurement period. U.S. Executive Branch 1997, 8–11, 29–44.

147. The U.S. Executive Branch Study reported only a modest increase in the level of U.S. foreign direct investment in Mexico in the 1994–96 period. U.S. Executive Branch 1997, 22–24. A later Mexican government report indicates a significant acceleration of foreign direct investment inflows. Mexico Economic Update, NAFTA Works, March 1999, 1Google Scholar.

148. The U.S. Executive Branch Study confirms that it is difficult to isolate the effects of NAFTA on regional trade or investment flows during the first three years of its operation because the peso crisis overwhelms NAFTA-specified changes. The U.S. economy was exceptionally strong during the measurement period, and this alone might account for changes in import-export trends. U.S. Executive Branch 1997, 12–14.

149. Ernst Haas observed that the progress of regional integration is rendered unpredictable by turbulence and by the close linkage between regional integration mechanisms and external global economic and political factors. Haas 1975. The peso crisis was a turbulent event that NAFTA survived intact. This demonstrates some measure of durability, even if factors other than legalization (for example, U.S. interest in preserving Mexican political stability) affected developments.