We thank Frank Alcock, Jeffrey Berejikian, Thomas Bernauer, Liliana Botcheva-Andonova, Robert Darst, Walter P. Falcon, James Fearon, Robert O. Keohane, Barbara Koremenos, Charles Lipson, Lisa Martin, James Morrow, Rosamond Naylor, Thomas Oatley, Kenneth Oye, John Richards, Duncan Snidal, Alexander Thompson, Michael Zürn, and the other contributors to this volume for helpful comments on earlier drafts of this article. Insightful suggestions by David Lake, Peter Gourevitch, and two anonymous reviewers proved particularly helpful in refining the argument. Valuable research assistance was provided by Hannah Fairbank, Elyce Hues, Aaron Knott, and Jonah Spiegelman. Ronald Mitchell thanks the University of Oregon's Department of Political Science and Stanford University's Center for Environmental Science and Policy for research support that contributed to completion of this article. Patricia Keilbach thanks the University of Oregon's Department of Political Science for research support that contributed to completion of this article.

1. Milner 1997, 44.

2. See the discussion of the number of actors variable, number, in Koremenos, Lipson, and Snidal, this volume, 777–78.

3. Milner 1997, 8–9.

4. Asymmetries are discussed as part of the number variable in Koremenos, Lipson, and Snidal, this volume, 778.

5. Koremenos, Lipson, and Snidal, this volume.

6. On interactions among variables, see the comments by Koremenos, Lipson, and Snidal, this volume, 779–80.

7. Ibid., 796.

8. For a similar simplification of international economic interactions, see Oye 1992.

9. Limiting negotiations to a single issue helps states avoid the potentially debilitating complexity of linkages to other issues. See Sebenius 1983; and McGinnis 1986.

10. Thus states have created separate regimes for acid precipitation, climate change, and stratospheric ozone loss and for biodiversity, endangered species, deforestation, and desertification rather than broader ones covering atmospheric or wildlife issues, respectively. Many environmentalists criticize this tendency to compartmentalize as ineffective in resolving environmental problems that are fundamentally integrated. Esty 1994.

11. This insight has not been fully incorporated in the debate between the “managerial” and “enforcement” schools. See Chayes and Chayes 1995; and Downs, Rocke, and Barsoom 1996.

12. See Oye 1992, 17; and Milner 1997, 8–9.

13. Mitchell 1999.

14. See Stein 1983; Krasner 1991; and Martin 1992a.

15. Waltz uses a similar analogy. Waltz 1979, 196.

16. The quote is from Stein 1983, 120. On such externalities, see Coase 1960; Conybeare 1980; Oye 1992; and Bernauer and Ruloff 1999.

17. Koremenos, Lipson, and Snidal, this volume.

18. Asymmetric externalities have received scholarly attention only recently but do not appear particularly rare empirically. See Rittberger and Zürn 1990, 31–32; and Martin 1992a,b.

19. We are indebted to James Morrow for clarifying this point.

20. See Knorr 1975, 310–19; Baldwin 1979, 184; and Morgenthau 1993, 31.

21. Keohane and Nye 1989.

22. This distinction coincides with the “paradox of unrealized power,” that whether a state can convert its control over resources into influence over outcomes depends on how it deploys those resources. Baldwin 1979.

23. In contrast to Oye, we seek to explain only cases in which the institution changes, rather than clarifies, the contingent response of the dissatisfied state, thus excluding his category of “explanation.” Oye 1992.

24. Kydd, this volume.

25. Koremenos, Lipson, and Snidal, this volume, 777–78.

26. For other views on coercion, contracts, and extortion, see Oye 1992, 35; and Krasner 1999, 26.

27. See Hasenclever, Mayer, and Rittberger 1997, 106; Krasner 1991, 340f; and Amini 1997, 7.

28. X can be either a positive or negative action, that is, A can be attempting to get B to commence a new activity or to halt or change an existing activity.

29. Although many regimes replace or reinforce issue-specific reciprocity with reciprocity based on positive and negative linkage to other issue-areas, the core of Axelrod's argument lay in demonstrating that Tit-for-Tat fosters cooperation without such linkage. Axelrod 1984.

30. Michael Zürn and Thomas Bernauer helped us clarify this point. Thus states can threaten to punish their political dissidents unless other states stop punishing theirs but do not do so because the nature of the situation makes it obviously ineffective.

31. See Axelrod and Keohane 1986; and Oye 1992, 17–33.

32. See McGinnis 1986; Sebenius 1983; and Haas 1980, 371–72.

33. Krasner 1999, 26.

34. Baldwin 1971, 24.

35. See Fearon 1998; and Morrow 1994c.

36. Hirschman 1970.

37. Axelrod and Keohane 1986.

38. See Snidal 1985; and Kindleberger 1981.

39. Keohane and Nye 1989, 36, 235–36. It remains unclear how often hegemonic states view imposed agreements as cheaper ways to coerce cooperation than traditional threats exercised outside of an international institution. Young 1989, 84–89.

40. See Young 1989, 84–89; Martin 1992a,b; and Rittberger and Zürn 1990.

41. Drezner 2000.

42. See Baldwin 1971; and Schelling 1960, 177.

43. Thucydides 1954.

44. See Hasenclever, Mayer, and Rittberger 1997, 106; Keohane and Nye 1989, 52–53, 122; and Krasner 1991, 340f.

45. This is evident, for example, in the agreement to induce North Korea to forgo its nuclear weapons ambitions. Dorn and Fulton 1997.

46. Oatley, this volume.

47. See John Richards' example of airline regulation and Milner and Rosendorff's discussion of escape clauses, in this volume.

48. Koremenos, Lipson, and Snidal, this volume, 766–67 and fn. 19.

49. We are indebted to Michael Zürn for this insight.

50. Mitchell 2001.

51. Peterson 1992, 158.

52. Levy 1988, 5.

53. See Stoett 1997, 57; and Tonnessen and Johnsen 1982, 509.

54. See Stoett 1997, 57; Peterson 1992, 158, 160; and Levy 1988, 17.

55. See ICRW 1946, Art. IX; and Walsh 1999.

56. On escape clauses, see Milner and Rosendorff 2001.

57. Walsh 1999, 313. In 1959 Norway and the Netherlands withdrew altogether in protest of quotas they considered too restrictive. In subsequent cases opting out created the awkward but unsurprising situation of species-specific quotas that bound only states that did not hunt those species.

58. Even development of a system of independent inspectors to verify compliance failed to address the question of “after detection, what?” Ikle 1961.

59. Earlier Soviet violations seem quite distinct from more recent Japanese violations that have been smaller in magnitude and coupled with quite public denunciations of the IWC. See Yablokov 1994; Baker and Palumbi 1994; and Peterson 1992.

60. In terms of problem structure, a “conflict over means” had become a “conflict over values.” Rittberger and Zürn 1990.

61. D'Amato and Chopra 1991.

62. Stedman 1990, 168.

63. See Levy 1988, 29; and Andresen 1989, 109, 116.

64. See Birnie 1985, 616; Sigvaldsson 1996, 330 citing Holt 1985, 192; and Stedman 1990, 168.

65. See Andresen 1989, 111; Wilkinson 1989; and Martin and Brennan 1989.

66. DeSombre 2000.

67. See Hoel 1993; and Caron 1995.

68. If the strategy is eventually judged ineffective, the United States, and even nongovernmental organizations, may yet decide that paying states not to whale is a more effective strategy, however morally repugnant.

69. Sell 1996, 100.

70. Ibid., 99–100.

71. Ibid., 102.

72. Three years later, that body adopted a weak “list of measures that might be taken” (including “issuing cautions” and suspending treaty rights and privileges) if industrialized states failed to meet deadlines for phasing out ozone-depleting substances or to fund the financial mechanism. UNEP 1991.

73. Sell 1996, 100. Restrictions on trade in ozone-depleting substances “would have no inhibiting effect on China and India because of their huge potential domestic markets.” Benedick 1991, 100.

74. Benedick 1991, 151; and ratification list compiled by authors.

75. Montreal Protocol 1987/1990, Art. 5(5–7).

76. Weiss and Jacobson 1998.

77. Clapp 1997. Victor considers the threat of cutting off multilateral funds as a sanction, but it is one that would have been unavailable without initial adoption of the reward-based strategy. Victor 1998, 165–66.

78. The section that follows builds extensively on the excellent analyses of the Rhine River case by Bernauer 1995 and 1996.

79. See LeMarquand 1977, 125; and Mingst 1981, 164.

80. See Kamminga 1978, 66; and LeMarquand 1977, 119.

81. Bernauer 1995, 372.

82. Bernauer 1996, 220–21.

83. Ibid., 209.

84. Although not specified in the convention and often misinterpreted as kilograms per second, the unit of measurement is kilograms per stere, with 1 stere equal to 1 cubic meter of water.

85. Bernauer 1996, 210.

86. LeMarquand 1977, 118.

87. Mingst 1981, 168.

88. Bernauer 1995, 377.

89. Bernauer 1996, 216.

90. See LeMarquand 1977, 119; and Bernauer 1996, 205.

91. Indeed, Bernauer argues that the agreement had only a small effect on French behavior. Bernauer 1996. As noted earlier, however, we are explaining regime design, not regime effectiveness.

92. Bernauer 1996, 209–10.

93. See LeMarquand 1977, 124; and Bernauer 1995, 372.

94. See LeMarquand 1977, 119; and Bernauer 1996, 221.

95. Bernauer 1995, 378.

96. Ibid., 378–79. See also the failure of efforts to induce Eastern European states to decommission unsafe nuclear reactors detailed in Connolly and List 1996.

97. Bernauer 1996, 225.

98. See Hasenclever, Mayer, and Rittberger 1997, 33, 42–43; and Keohane 1983, 153.

99. Hardin 1968.

100. Koremenos, Lipson, and Snidal, this volume, 782.

101. See Yablokov 1994; Clapp 1997; Biermann 1997; Victor 1998; and Bernauer and Moser 1996.

102. We are indebted to an anonymous reviewer for clarifying these points.

103. On the distinction between a logic of appropriateness and a logic of consequences, see March and Olsen 1989; and Finnemore 1996.

104. Young 1979.

105. See Fearon 1998; and Morrow 1994c.

106. See Martin 1992b; and Rittberger and Zürn 1990.

107. On imposed regimes, see Young 1989; Martin 1992a; and Gruber 2000.

108. Coase 1960.

109. See Chayes and Chayes 1995; and Downs, Rocke, and Barsoom 1996.