1. Putnam, Robert D., “Diplomacy and Domestic Politics: The Logic of Two-Level Games,” International Organization 42 (Summer 1988), pp. 427–60CrossRefGoogle Scholar.

2. For the two-level game analysis, see Eichenberg, Richard C., “Dual Track and Double Trouble: The Two-Level Politics of INF,” in Evans, Peter B., Jacobson, Harold K., and Putnam, Robert D., eds., Double-edged Diplomacy: International Bargaining and Domestic Politics (Berkeley: University of California Press, 1993), pp. 45–76Google Scholar.

3. Putnam, , “Diplomacy and Domestic Politics,” pp. 437–38 and 440Google Scholar.

4. The main developers of this approach recognize that it incorporates many previous observations and so do not claim its main contribution lies in introducing new ideas. I show here how extensively the two-level game model relies on previous observations, not to criticize its developers but to set the stage for a comparison that shows the three-and-three approach better captures the best insights of earlier scholars.

5. These are drawn from Putnam, , “Diplomacy and Domestic Politics,” pp. 441–52Google Scholar.

6. See, for example, Gowa, Joanne, “Anarchy, Egoism, and Third Images: The Evolution of Cooperation and international relations,” International Organization 40 (Winter 1986), pp. 167–86CrossRefGoogle Scholar, and in particular p. 181; and Evangelista, Matthew, “Cooperation Theory and Disarmament Negotiations in the 1950s,” World Politics 42 (07 1990), pp. 502–28CrossRefGoogle Scholar and in particular p. 526.

7. For more recent and theoretical analyses of this case that cite the earlier studies, see Caldwell, Dan, The Dynamics of Domestic Politics and Arms Control (Columbia: University of South Carolina Press, 1991)Google Scholar; or Skidmore, David, “The Politics of National Security Policy: Interest Groups, Coalitions, and the SALT II Debate,” in Skidmore, David and Hudson, Valerie M., eds., The Limits of State Autonomy (Boulder, Colo.: Westview Press, 1993), pp. 205–33Google Scholar.

8. See Putnam, , “Diplomacy and Domestic Politics”; and Andrew Moravcsik “Introduction: Integrating International and Domestic Explanations of World Politics,” in Jacobson, Evans, and Putnam, , Double-edged Diplomacy, pp. 3–42Google Scholar.

9. See Putnam, , “Diplomacy and Domestic Politics,” pp. 452–53Google Scholar; Moravcsik, , “Introduction: Integrating International and Domestic Explanations of World Politics,” p. 28Google Scholar; Jensen, Lloyd, Bargaining for National Security (Columbia: University of South Carolina Press, 1988), pp. 38–39Google Scholar; and Raiffa, Howard, The Art and Science of Negotiation (Cambridge, Mass.: Harvard University Press, 1983), pp. 166–86Google Scholar.

10. On reverberation and targeting, see Putnam, , “Diplomacy and Domestic Politics,” p. 454Google Scholar; and Moravcsik, , “Introduction: Integrating International and Domestic Explanations,” p. 29Google Scholar. For Griffiths's hypothesis, see Griffiths, Franklyn, “The Political Side of ‘Disarmament,’” International Journal 22 (Spring 1967), pp. 293–305Google Scholar and especially pp. 294 and 302.

11. See Moravcsik, , “Introduction: Integrating International and Domestic Explanations,” p. 26Google Scholar; Mastanduno, Michael, Lake, David A., and Ikenberry, G. John, “Toward a Realist Theory of State Action,” International Studies Quarterly 33 (12 1989), pp. 457–74CrossRefGoogle Scholar and especially p. 464; Lowi, Theodore, The End of Liberalism, 2d ed. (New York: W.W. Norton, 1979), p. 146Google Scholar; Adelman, Kenneth L., The Great Universal Embrace (New York: Simon and Schuster, 1989), p. 42Google Scholar; and Marra, Robin F., Ostrom, Charles W. Jr, and Simon, Dennis M., “Foreign Policy and Presidential Popularity,” Journal of Conflict Resolution 34 (12 1990), pp. 588–623CrossRefGoogle Scholar and especially pp. 600–601.

12. I am commenting here only on the originality of the essays that first laid out the theory of two-level games. As one would expect, the conclusions that emerge from the eleven case studies in the collaborative project mentioned above include some genuinely new insights. (These are summarized in Evans, Peter B., “Building an Integrative Approach to International and Domestic Politics: Reflections and Projections,” in Jacobson, Evans, and Putnam, , Double-edged Diplomacy, pp. 397–430.)Google Scholar This does not change the argument here, however. Those new observations, while certainly welcome, were not produced directly through the process of elaborating the two-level game concept itself. Indeed, Evans explictly emphasizes that his conclusions represent results “which grew out of the project itself” and were not “fully anticipated in the original Putnam article” (see ibid, p. 399). Thus, the new findings generated by the collaborative project illustrate the third type of contribution I discussed above—the value of two-level games as a framework for guiding empirical inquiry—since these insights were ascertained only inductively, through case study research. I argue below however that even greater insight into the implications of domestic–international interaction could be gained by using an alternative to the two-level game framework to guide case research.

13. Moravcsik contends that one of the two-level game's most valuable features is its emphasis on domestic–international interaction. He argues persuasively that this makes it better than what he calls additive approaches that simply explain some portion of the variance with international variables and a further portion with domestic variables. Such additive approaches do not recognize the possibility that processes at one level might alter the impact of variables at the other level. In contrast, the two-level game and three-and-three approach I propose emphasize precisely this possibility. See Moravscik, , “Introduction: Integrating International and Domestic Explanations,” p. 17Google Scholar.

14. I am aware of only two exceptions. The first, Mayer, Frederick W., ”Managing Domestic Differences in International Negotiations: The Strategic Use of Internal Side-payments,” International Organization 46 (Autumn 1992), pp. 793–818CrossRefGoogle Scholar, is concerned with only one narrow issue, identifying the conditions under which internal side-payments will influence the outcome of international negotiations. Mayer does not seek to develop a formal model from which to deduce a range of propositions about domestic–international interaction in general. The other exception, McGinnis, Michael D. and Williams, John T., “Policy Uncertainty in Two-level Games: Examples of Correlated Equilibria,” International Studies Quarterly 37 (03 1993), pp. 29–54CrossRefGoogle Scholar, is more comprehensive. However, the model they develop involves some radically different assumptions from Putnam's and hence cannot be considered a formalization of two-level games as Putnam and his associates have used that term.

15. See Putnam, , “Diplomacy and Domestic Politics,” pp. 433–34Google Scholar; Moravcsik, , “Introduction: Integrating International and Domestic Explanations,” p. 23Google Scholar; and Evans, , “Building an Integrative Approach to International and Domestic Politics,” p. 398Google Scholar.

16. Moravcsik, , “Introduction: Integrating International and Domestic Explanations,” p. 33Google Scholar.

17. Hoffmann, Stanley, Contemporary Theory in International Relations (Englewood Cliffs, N.J.: Prentice-Hall, 1960), pp. 7–8 and 40Google Scholar. George's method of “structured, focused comparison” can be seen as one attempt to systematize this idea of theory as a set of questions. See George, Alexander, “Case Studies and Theory Development,” in Lauren, Paul Gordon, ed., Diplomacy: New Approaches in History, Theory, and Policy (New York: Free Press, 1979), pp. 43–68Google Scholar. This is the method employed by Evans, Jacobson, and Putnam in their Double-edged Diplomacy.

18. Putnam's, quotation is from “Diplomacy and Domestic Politics,” p. 429Google Scholar; for Keohane and Nye's definition, see Keohane, Robert O. and Nye, Joseph S., “Transgovernmental Relations and International Organizations,” World Politics 27 (10 1974), pp. 39–62CrossRefGoogle Scholar; and for other, similar propositions, see Allison, Graham T. and Halperin, Morton H., “Bureaucratic Politics: A Paradigm and Some Policy Implications,” in Ikenberry, G. John, ed., American Foreign Policy: Theoretical Essays (Glenview, Ill.: Scott, Foresman, 1989), pp. 378–409Google Scholar and especially p. 394.

19. The quotation is from Putnam, , “Diplomacy and Domestic Politics,” p. 444Google Scholar; see also p. 459. Compare Keohane, Robert O. and Nye, Joseph, eds., Transnational Relations and World Politics (Cambridge, Mass.: Harvard University Press, 1972)CrossRefGoogle Scholar with their article cited in the preceding note, or see their Power and Interdependence (Boston: Little, Brown, 1977), pp. 24–25 and 33–35Google Scholar.

20. See the discussion in Moravcsik, , “Introduction: Integrating International and Domestic Explanations,” p. 32Google Scholar. A figure illustrating Moravcsik's overview can be found at the bottom of that page. The figure lists five discrete strategies that COGs can employ, each of which is explored in the subsequent case studies, but it does not designate a single specific strategy arising from the ability of non-COGs to initiate contacts abroad. Hence, Moravscik provides no basis to explore the impact of the three forms of interaction I am discussing here. In his conclusion, Evans discusses some conditions under which transnational alliances had an impact in the case studies (see Evans, , “Building an Integrative Approach to Domestic and International Politics,” pp. 418–23Google Scholar). But Evans explicitly excludes transnational alliances from a role in security issues, which the case study below suggests is incorrect. Moreover, he does not bring up the distinction I have made here between transnational and other forms of connection across state borders, so he cannot compare the impact of transnational alliances with that of either transgovernmental alliances or connections between domestic players and government officials abroad. Thus, possible differences in these types of domestic–international interaction remain unexplored.

21. This definition is slightly broader than the traditional conception of transgovernmental relations. The classic use of the term suggests comparable bureaus in two internally divided governments working together. But this symmetry is not necessary. I believe the same process is at work when the COG of a relatively unified government colludes with a bureaucratic faction in another, divided government. If these external signals affect the debate in the divided government, the interaction would have changed the outcome from what bargaining between two unitary governments or what bureaucratic politics occurring on one side in isolation would have produced. Yet this process involves interaction only between governmental officials and not private actors. Cases that do not exactly match the ideal-typical case still need to be classified, and I believe the gray-area case of interaction between a unified and a divided government better fits the transgovernmental category than the intergovernmental category or one of the other two transboundary classifications.

22. Cross-level strategies are one species of what Rosenau once labeled “linkage politics”; see Rosenau, James, Linkage Politics (New York: Free Press, 1969)Google Scholar. Although I have retained existing labels for the other two forms of transboundary connection, I do not use Rosenau's term because he used linkage politics to connote connections across any two levels of analysis. I want to restrict my purview purely to interactions between the international and domestic levels and hence have created the new term “cross-level.”

23. For evidence that this is true in domestic politics, see Light, Paul C., The President's Agenda: Domestic Policy Choice from Kennedy to Reagan, rev. ed. (Baltimore: Johns Hopkins University Press, 1991), pp. 69–70Google Scholar.

24. Thomas Risse-Kappen, Cooperation Among Democracies: Norms, Transnational Relations, and the European Influence on U.S. Foreign Policy, typescript, chap. 4.

25. Thornton, Robert L., “Governments and Airlines,” in Keohane, and Nye, , eds., Transnational Relations, p. 201Google Scholar.

26. For a discussion of reasons why states care about relative gains (i.e., who gets more), see Grieco, Joseph, Cooperation Among Nations (Ithaca, N.Y.: Cornell University Press, 1990)Google Scholar.

27. Knopf, Jeffrey W., “Domestic Politics, Citizen Activism, and U.S. Nuclear Arms Control Policy,” Ph.D. diss., Stanford University, Stanford, Calif., 1991Google Scholar, chap. 5.

^28. Jönsson, Christer, Soviet Bargaining Behavior (New York: Columbia University Press, 1979)Google Scholar.

29. Putnam considers the possibility that some issues may require analysis at more than two levels. But his discussion is connected to the possibility of multiple steps in domestic ratification. He does not consider the issue in conjunction with possible alliance dynamics. See Putnam, , “Diplomacy and Domestic Politics,” pp. 449–50Google Scholar.

30. Gowa, Joanne and Mansfield, Edward D., “Power Politics and International Trade,” American Political Science Review 87 (06 1993), pp. 408–20CrossRefGoogle Scholar.

31. Kahneman, Daniel and Tversky, Amos, “Prospect Theory: An Analysis of Decision Under Risk,” Econometrica 47 (03 1979), pp. 263–91CrossRefGoogle Scholar. For discussion of prospect theory's applicability to international relations, see Levy, Jack S., “Prospect Theory and International Relations: Theoretical Applications and Analytical Problems,” Political Psychology 13 (06 1992), pp. 283–310CrossRefGoogle Scholar.

32. Rather than seek to develop something entirely new, my goal is more to preserve the valid insights of earlier approaches that have not been incorporated into the two-level game framework.

33. Eichenberg, “Dual Track and Double Trouble.” At points that further illuminate advantages of the three-and-three approach, I also note contrasts between my analysis and those of other recent studies.

34. I interviewed officials from various agencies as part of a project on the impact of domestic forces on U.S. arms control policy. Unsolicited, seventeen of the twenty-three interview subjects brought up the issue of INF and/or European pressures on arms control. My interpretation of the case is based mostly on previously available information, with the interviews used only to supplement the existing data, but they nonetheless add useful information to the data base on INF. Because I guaranteed most interview subjects anonymity, I identify only the agency and not the name of the officials I quote.

35. Eichenberg, , “Dual Track and Double Trouble,” pp. 46 and 72Google Scholar. Recall that one normally expects agreement to be more likely when win-sets expand—hence, the anomaly.

36. Risse-Kappen, Thomas, “Did ‘Peace Through Strength’ End the Cold War? Lessons from INF,” International Security 16 (Summer 1991), pp. 162–88CrossRefGoogle Scholar and in particular pp. 173–75 and 182–85.

37. See Schwartz, David N., NATO's Nuclear Dilemmas (Washington, D.C.: Brookings Institution, 1983), pp. 219–27Google Scholar; Garthoff, Raymond L., Detente and Confrontation: American–Soviet Relations from Nixon to Reagan (Washington, D.C.: Brookings Institution, 1985), pp. 860, 1024, and 1031–32Google Scholar; and Risse-Kappen, Thomas, The Zero Option: INF, West Germany, and Arms Control (Boulder, Colo.: Westview Press, 1988), pp. 42–43, 54–58, and 79–85Google Scholar.

38. Risse-Kappen, , “Did ‘Peace Through Strength’ End the Cold War?” p. 181Google Scholar.

39. Richard Allen is quoted in Talbott, Strobe, Deadly Gambits (New York: Vintage Books, 1985), p. 57Google Scholar.

40. See Risse-Kappen, , The Zero Option, pp. 79–83Google Scholar; Sigal, Leon V., Nuclear Forces in Europe (Washington, D.C.: Brookings Institution, 1984), pp. 81–83Google Scholar; and Talbott, , Deadly Gambits, pp. 59–62Google Scholar.

41. This is where a three-and-three approach adds to what is otherwise a fairly similar analysis by Risse-Kappen. Though not in the terminology employed here, Risse-Kappen discusses in The Zero Option all three of the elements I discuss as important. However, his purpose is to explain the outcome of the INF case itself. Thus, he does not seek to identify the separate impact of transgovernmental and cross-level connections and the interalliance context or to draw more general conclusions from their respective roles in this case. While mostly consistent with Risse-Kappen's analysis, the analysis here goes one step further by trying to draw potentially generalizable hypotheses from the INF experience.

42. See Risse-Kappen, , “Did ‘Peace Through Strength’ End the Cold War?” p. 169Google Scholar, footnote 15, and p. 178; Haslam, Jonathan, The Soviet Union and the Politics of Nuclear Weapons in Europe, 1969–1987 (Ithaca, N.Y.: Cornell University Press, 1990), pp. 111–112, 115–17, 125, and 139Google Scholar; and Goldberg, Andrew C., “Moscow's INF Experience,” in Mandelbaum, Michael, ed., The Other Side of the Table (New York: Council on Foreign Relations, 1990), pp. 89–119Google Scholar and in particular pp. 106–8.

43. Defense Secretary Weinberger cites in his memoirs “arrangements made by the Soviet Union” to aid antinuclear demonstrations and “manipulation of Western public opinion by the Soviets” as major causes of U.S. concern in the INF issue. See Weinberger, Caspar W., Fighting for Peace (New York: Warner Books, 1990), pp. 338–39Google Scholar. This perspective was not unique to Pentagon hard-liners. A senior State Department official likewise stated, “The Soviet game was to see if they could … stop the deployments”; personal interview, 10 August 1989, not for attribution.

44. Personal interview, 22 May 1989, not for attribution.

45. See Richard Perle's statements as quoted in Talbott, , Deadly Gambits, pp. 82–84Google Scholar.

46. Personal interviews, 16 May 1989, and 10 May 1989, respectively, not for attribution. Later in the same interview, the Arms Control and Disarmament Agency official cited above gave the figure of 75 percent instead.

47. Talbott, , Deadly Gambits, pp. 80–81Google Scholar.

48. The quotations are from Moravcsik, , “Introduction: Integrating International and Domestic Explanations,” p. 16Google Scholar; and Evans, , “Building an Integrative Approach to International and Domestic Politics,” p. 401Google Scholar. Moravcsik does note that the transboundary forms of interaction I have described may give domestic actors strategic options of their own (pp. 31–32), but this observation was not made the basis of any hypotheses for exploration in the case studies in that volume and does not receive any attention in Evans's conclusion. For two other examples of the near-exclusive focus on how chief negotiators can strategically manipulate two-level games, see Lehman, Howard P. and McCoy, Jennifer L., “The Dynamics of the Two-level Bargaining Game: The 1988 Brazilian Debt Negotiations,” World Politics 44 (07 1992), pp. 600–644CrossRefGoogle Scholar; and Mayer, “Managing Domestic Differences in International Negotiations.”

49. Eichenberg, , “Dual Track and Double Trouble,” p. 66Google Scholar.

50. For an outline of the peace activists’ reservations by one of the antinuclear campaign's founders, see Thompson, E.P., Beyond the Cold War (New York: Pantheon Books, 1982), pp. 123–32Google Scholar. Thompson's, response to the zero-option offer was, “yes, but MORE” (p. 125Google Scholar, emphasis original).

51. Personal interviews, 15 June 1989, and 22 May 1989, not for attribution. Further confirmation that transgovemmental and cross-level pressures constrained the United States to put forward a zero option can be found in the memoirs of Reagan's Secretary of Defense. See Weinberger, , Fighting for Peace, pp. 336–39Google Scholar.

52. See Moravcsik, , “Integrating International and Domestic Explanations of World Politics,” p. 27Google Scholar; and Evans, , “Building an Integrative Approach to International and Domestic Politics,” pp. 399 and 403Google Scholar.

53. See Talbott, , Deadly Gambits, pp. 171–200Google Scholar; and Dean, Jonathan, Watershed in Europe (Lexington, Mass.: Lexington Books, 1987), pp. 133–34Google Scholar.

54. Goldmann, Kjell, “International Opinion and World Politics: The Case of the INF Treaty,” Political Studies 41 (03 1993), pp. 41–57CrossRefGoogle Scholar; the quotations are from p. 56. Goldmann ascribes the unpredictability of the INF case to the impact of European protests on their own governments, which in turn pressured Washington. The analysis here agrees with Goldmann's arguments but also points out how European protests had in addition a direct cross-level effect in advantaging certain positions in the U.S. administration's internal debates.

55. Risse-Kappen, , “Did ‘Peace Through Strength’ End the Cold War?” pp. 173–75 and 182–85Google Scholar.

56. See, for example, Herf, Jeffrey, War by Other Means: Soviet Power, West German Resistance, and the Battle of the Euromissiles (New York: Free Press, 1991), pp. 219–23Google Scholar.

57. See Dean, , Watershed in Europe, pp. 146–47Google Scholar; Eichenberg, Richard C., Public Opinion and National Security in Western Europe (Ithaca, N.Y.: Cornell University Press, 1989), p. 107CrossRefGoogle Scholar; and Risse-Kappen, , The Zero Option, pp. 132–35Google Scholar.

58. Eichenberg, , “Dual Track and Double Trouble,” pp. 65–66 and 68Google Scholar.

59. The first incident is discussed in Evangelista, “Cooperation Theory and Disarmament Negotiations”; the second, in Divine, Robert A., Blowing on the Wind: The Nuclear Test Ban Debate, 1954–1960 (New York: Oxford University Press, 1978), pp. 245–51Google Scholar; and the third, in Talbott, Strobe, Endgame: The Inside Story of SALT II (New York: Harper & Row, 1979), pp. 38–78Google Scholar.

60. In the end, NATO canceled Lance deployments anyway. However, the United States was able to retain a unified Germany as a member of the Western alliance and has been able to maintain a presence in Europe even after the collapse of the Soviet Union. These developments might not have been possible had the United States reneged on its willingness to accept an agreement welcomed by most of European public opinion.

61. Eichenberg, , Public Opinion and National Security, pp. 29, 94–95, 108–11, 198–99, and 230Google Scholar.

62. Lyne, Roderic, “Making Waves: Gorbachev's Public Diplomacy, 1985–86,” in Laird, Robbin F., ed., Soviet Foreign Policy, Proceedings of the Academy of Political Science, vol. 36, 1987, pp. 235–53Google Scholar.

63. In Germany, for example, the opposition Social Democrats warmly welcomed Gorbachev's embrace of the zero option: see Herf, , War by Other Means, pp. 220–22Google Scholar.

64. In his memoirs, for example, Reagan claims he sought “a world free of nuclear weapons.” See Reagan, Ronald, An American Life (New York: Simon and Schuster, 1990), p. 550Google Scholar. The language of Reagan's “Star Wars” speech and his reported eagerness at Reykjavik to accept a proposal to eliminate all ballistic missiles make assertions of his antinuclear sentiment appear credible.

65. According to then Secretary of State Shultz, “President Reagan was consistently committed to his personal vision of a world without nuclear weapons; his advisors were determined to turn him away from that course,” as were “the British, French, Dutch, [and] Belgians”; see Shultz, George P., Turmoil and Triumph (New York: Charles Scribners Sons, 1993), pp. 360 and 701Google Scholar. While one might not expect such differences between a President and his aides, they were possible because there had been nothing to make them salient before Gorbachev took office. Before that, the whole Reagan team had been in agreement that a major military buildup should take precedence over arms control. Differences in opinion became important only when, following several years of increased defense spending, a possible Soviet partner for Reagan's disarmament schemes suddenly appeared. The President was unlikely to dismiss advisers who had supported the rest of his program simply because they did not share his views on that question, especially because he was “terribly reluctant to discipline or fire anyone”; see ibid, p. 691.

66. Newhouse, John, War and Peace in the Nuclear Age (New York: Vintage Books, 1988), pp. 394–99Google Scholar; Talbott, Strobe, The Master of the Game: Paul Nitze and the Nuclear Peace (New York: Alfred A. Knopf, 1988), pp. 314–29Google Scholar.

67. Eichenberg, , “Dual Track and Double Trouble,” p. 73Google Scholar.

68. Shultz, , Turmoil and Triumph, pp. 701 and 777–78Google Scholar.

69. Shultz, , Turmoil and Triumph, pp. 708 and 899–900Google Scholar.