Public Opinion, Domestic Structure, and Foreign Policy in Liberal Democracies Thomas Risse-Kappen World Politics, Vol. 43, No. 4. (Jul., 1991), pp. 479-512. Stable URL: http://links.jstor.org/sici?sici=0043-8871%28l99107%2943 %3A4%3C479%3APODSAF%3E2.0.CO%3B2-6 World Politics is currently published by The Johns Hopkins University Press. Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/jhup.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is an independent not-for-profit organization dedicated to creating and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact support@jstor.org. http: // w w w j s tor. org/ TueNov 28 07:22:42 2006 PUBLIC OPINION, DOMESTIC STRUCTURE, AND FOREIGN POLICY IN LIBERAL DEMOCRACIES By THOMAS RISSE-KAPPEN* Introduction HOW is it to be explained that similarly powerful states more often than not respond differently to the same international conditions and constraints? Efforts to answer this question lead inevitably to the study of the domestic sources of foreign policy and international politics. One issue, though, is rarely addressed by the available literature: Who is in charge of the foreign policy-making process in liberal democratic states? Elites or masses? Who influences whom? What is the policy impact of societal actors and public opinion? Are public attitudes on foreign affairs manipulated by the elites? Finally, if public attitudes follow similar patterns., what accounts for differences between various countries in the policy impact of public opinion? This paper tries to shed light on the relationship between public opinion and the foreign policy of liberal democracies.1 T will argue that the policy impact of public opinion does not depend so much on the specific * Drafts of this paper were presented at the European Consortium af Political Research (ECPR) Workshop on "Public Opinion, Foreign Policy, and the Democratic Process," Bachum, West Germany, April 2-7, 1990, and at the Annual Convention of the International Studies Association, Washington D.C., April 10-14, '99ö- For very valuable comments I thank Richard Eichenberg, Matthew Evangelista, Philip Everts, Kjell Galdmann, Paul Joseph, Claudia Kappen, Peter Katzenstein, Richard N- Lebow, David Meyer, Nobuo Oka-wara, Hans Rattinger, Bruce Russett, and the participants in the ECPR workshop. For their help and advice regarding public opinion data I thank Lisa Brandes and Richard Eichenberg. 1 The paper is part of a growing body of literature. See, for example, Michael Barnett, "High Politics Is Low Politics: The Domestic and Systemic Sources of Israeli, Security Policy, 1967-1977," World Polities 42 (July 1990), 510-62; Richard Eichenberg, Public Opinion and National Security in Weitem Europe (Ithaca, N.Y.r Cornell University Press, [989); Matthew Evangelista, Innovation and the Amu Race (Ithaca, N.Y.: Cornell University Press, 1988); Gert Krell, Rüstungsdynamti^ and Rüstungskontrolle- Die gesellschaftlichen Auseinandersetzungen um SALT in den USA '969-/975 (Frankfurt am Main: Haag und Herchen, 1976); Richard N. Lebow, Between Peace and Wat (Baltimore, Md.: Johns Hopkins University Press, 198t); Jack S. Levy, "Domestic Politics and War," Journal of Interdisciplinary History 18 (Spring 1988), 653-73; Erwin Müller, Rüstungipoliti\ und Rüstungsifpn trolle: Fall USA (Baden-Baden: No-mos, 1985); Bruce Russett, Controlling the Sword: The Democratic Governance of National Security (Cambridge: Harvard University Press, 1990); and Eugene Wittkopf, Faces of Internationalism: Public Opinion and American Foreign Policy (Durham, N.C; Duke University Press, 1990). World Politics 43 (July 1991), 479-512 480 WORLD POLITICS issues involved or on the particular pattern of public attitudes as on the domestic structure and the coalition-building processes in the respective country. The paper analyzes the impact of public opinion on the foreign policy-making process in four liberal democracies with distinct domestic structures: the United States, France, the Federal Republic of Germany, and Japan. I will show that differences in political institutions, policy networks, and societal structures account for different foreign policy outcomes when the impact of the international environment is controlled for and public attitudes follow similar patterns across countries. Thus, the four countries responded differently to Soviet policies during the 1980s despite more or less comparable trends in mass public opinion. The analysis of the interaction between public opinion and elite coalition-building processes in the four countries reveals that the policy outcomes differ according to variances in domestic structures and not in the international status of the states. Public Opinion and Foreign Policy; Theoretical Concepts and Methodological Problems Who Follows Whom? Elites and Masses Most of the available literature on the interaction between mass public opinion and elites in the foreign policy-making process of liberal democracies can be categorized according to two broad concepts.1 In accordance with the pluralist theory of democracy, a "bottom-up" approach assumes that the general public has a measurable and distinct impact on the foreign policy—making process.3 In sum, leaders follow masses. However, the bottom-up concept is difficult to reconcile with the following empirical counterevidence: —In Western Europe, the U.S., and Japan, elites and masses show similar support for basic foreign policy goals and institutions. Prevailing political, religious, and ideological cleavages structure elite as well as mass public opinion.4 1 For excellent reviews of the debate, see Philip Everts and Arthur Faber, "Public Opinion, Foreign. Policy, and Democracy" (Paper presented at the ECPR Workshop "Public Opinion, Foreign Policy, and the Democratic Process," Boehum, April 2-7, 1990); and Russett (fn. r). 3 See, e.g., Benjamin I. Page and Robert Y. Shapiro, "Effects of Public Opinion on Policy," American Political Science Review 77, no. 1 (1983), 175-90. 4 For Western Europe, see Eichenberg (fn- t), pt, 5. For the U.S., see Ole R. Holsti and James N. Rosenau, "The Domestic and Foreign Policy Beliefs of American Leaders," Journal of Conflict Resolution 32 (June rg88), 24^-94; and Wittkopf (fn. 1). For Japan, see Davis Bobrow, "Japan in the World: Opinion from Defeat to Success" Journal of Conflict Resolution 33 (December 1989), 571-603. LIBERAL DEMOCRACIES 481 —There are many cases in which crucial foreign policy decisions have been taken in the absence of mass public consensus. Examples are the U.S. decisions in favor of an active international role in the postwar world and of becoming permanently involved in European security affairs; the West German decisions to rearm and join nato in the early 1950s and to pursue an active Oitpolttik^ in the late 1960s; the French decisions to build an independent nuclear force in the 1950s and to leave nato's military institutions in the mid-1960s. Therefore, a second approach representing the conventional wisdom in the literature suggests a "top-down" process, according to which popular consensus is a function of the elite consensus and elite cleavages trickle down to mass public opinion. This viewpoint concurs with either a "power elite" (C. Wright Mills) or a state^centered and realist approach to foreign policy. It is assumed that the public is easily manipulated by political leaders, because of (r) the low salience, or significance, of foreign and security policy issues as compared with economic policies, (2,) the low degree of knowledge about the issues involved, and {3) the volatility of public opinion.5 Again, empirical evidence suggests that these assumptions are questionable: —While only a minority can be regarded as politically active, large portions of the public seem regularly to follow news about foreign policy in the media.* While domestic problems usually outweigh foreign and security issues in public salience, data reveal that substantial minorities consider foreign affairs among the most important problems facing their respective countries. On the average, 20-30 percent of the public indicate serious concern about foreign affairs.' ' See, e.g., Noam Chomsky and Edward Herman, Manufacturing Consent (New York*. Pantheon., 1988); Benjamin Ginsberg, The Captive Public: How Mass Opinion Promotes State Power (New York: Basic Books, rrj86); Barry Hughes, The Domestic Context of American Foreign Policy (San Francisco: Freeman, 1979); Michael Margolis and Gary Mauser, eds., Manipulating Public Opinion: Essays on Public Opinion as a Dependent Variable (Belmont, Calif: Wads worth, eg 89); and James N. Rosenau, Public Opinion and Foreign Policy (New York: Random House, 1961). * During election campaigns in the U.S., for example, the majority of the public was aware of the different foreign policy choices under debate and could identify the positions of the presidential candidates. See John H. Aldrich et al., "Foreign Affairs and Issue Voting: Do Presidential Candidates 'Waltz before a Blind Audience'? "American Political Science Review 83, no. 1 (1989), 113-41', Hans Rattinger, "Foreign Policy and Security Issues as Determinants of Voting Behavior in the 1988 U.S. Presidential Election" (Paper presented at the ECPR Workshop "Public Opinion, Foreign Policy, and the Democratic Process," Bochum, April 2-7, 1990). See also Thomas Graham, "The Pattern and Importance of Public Knowledge in the Nuclear Age." Journal of Conflict Resolution 32, no, 1 (1988), 319-34- ? For the U.S., see Tom W. Smith, "The Polls: America's Most Important Problem, Part I: National and International," Public Opinion Quarterly 49, no. 1 (1985), 264-74; f°r France, West Germany, and Japan, see Hastings and Hastings, eds., Index to International Public Opinion, tySi-iqSs (New York: Greenwood, 1984), 563; and igS^-ro8^, 689, 482 WORLD POLITICS —Public attitudes3 at least on basic foreign policy issue sa seem to be more stable than is usually assumed, A large percentage of the public does nor change attitudes frequently. In sum, the public seems to be more rational and less open to elite manipulation than the top-down thesis asserts/ —That leaders try to manipulate masses and that governments launch sophisticated propaganda campaigns only suggest that the elites take the power of the uneducated masses seriously and feel vulnerable to it. The very existence of state propaganda and efforts at "spin control" belies the "power elite" hypothesis. In addition to these empirical problems, the bottom-up and the top-down models both suffer from conceptual shortcomings. First, they treat masses and elites as unitary actors. While certain segments of the public may be manipulated by government propaganda, others may resist efforts to influence them.9 One should at least distinguish between (i) mass public opinion, (2) the attentive public, which has a general interest in politics, and (3) issue publics^ which are particularly attentive to specific questions. Moreover, elites are themselves frequently divided, and different segments of the elite try to convince the general public of their respective viewpoints.10 Second, a simplistic view of either bottom-up or top-down processes tends to ignore that public opinion and societal groups may influence the policy-making process in several ways and at different stages. They can directly affect the choices of top decision makers by changing policy goals or how those goals are prioritized, by narrowing the range of options and/or means to implement goals, or by winning symbolic concessions in the sense of changed rhetoric rather than policy reforms. Moreover, the public may also indirectly affect policies by influencing the coalition-building processes among the elites. It can strengthen or weaken the positions of bureaucracies or single actors within the government. Public 4 Robert Shapiro and Benjamin I. Page, "Foreign Policy and the Rational Public," Journal of Conflict Resolution 32, no. 2 (1988), 211-47. F°f data, see Eichenherg (fn, 1); Gregory Flynn and Hans Rattinger, eds., The Public and Atlantic Defence (London: Croom Helm, 1985); Thorn a. j Graham, American Public Opinion, ón NATO, Extended Deterrence, and Use of Nuclear Weapons (Cambridge: Center for Science and International Affairs, Harvard University, 1989); and Rus sett (fn. i), chap. 4. * For example, the well-documented "rally around the flag" effect, according to which presidents can use military interventions to raise their popular support, applies only to a minority of the U.S. population and does not last very long. See Russett (fn. 1), 34-51. Far the original study on the effect, see John Mueller, War, Presidenti, and. Public Opinion (New York: Wiley, 1973). For a general argument regarding elite-public interactions, see Paul Joseph, Realizing the Peace Dividend: The End of the Cold War and the Emergence of Progressive Politics (forthcoming), chap. 4. 10 For a discussion of the cleavages in U.S. public and elite opinion, see Michael Mandel-baum and William Schneider, "The New Internationalism: Public Opinion and American Foreign Policy," in Kenneth Oyc et al., eds,, Eagle Entangled: U.S. Foreign Policy in a Complex World (New York: Longman, 1979), 34-90; and Wictkopf (fn. 1). LIBERAL DEMOCRACIES 4S3 opinion and the activities of public interest groups may lead to changes and/or realignments within or between political organizations such as parties. These indirect effects on policy are difficult to trace and are therefore easily overlooked. However, their impact may be as important as the above-mentioned direct influences. Finally, it cannot be assumed that public and elite opinion interact with each other and are transformed into policy decisions in the same way across different countries. In other words, similarities in public attitudes across various countries do not necessarily lead to similar policies. Public attitudes and policies toward nuclear weapons provide a good example. The patterns of mass public opinion in the U.S., Western Europe, and Japan are more or less similar. Nuclear weapons are mostly tolerated for their deterrent value, provided that arms control efforts are underway. Attitudes become increasingly hostile if nuclear use in wartime situations is contemplated. Throughout the cold war, "first use" of nuclear weapons had only minority support in the U.S., France, West Germany, and Japan." However, for all these similarities in public attitudes, the nuclear policies of the four countries have differed from one another. Japan has had the strongest antinuclear policy in place since the late 1960s. The Federal Republic of Germany changed from an outspoken supporter of U.S. extended deterrence and "coupling" to a strong advocate of nuclear arms control. The United States emphasized arms control in the 1970s, turned to a massive arms buildup during Ronald Reagan's first term, and returned to arms control from the mid-1980s on. Finally, France has pursued the most pronuclear policy of the four countries and has refused so far to take part in nuclear arms control. The variations in policies are in part a result of different positions and capabilities in the international environment (for example, the nuclear status of France and the U.S. as compared with that of Japan and Germany). However, while the frequent changes in United States arms-control policies might be considered to be reactions to changes in the "threat environment" (Soviet arms buildup in the 1970s; Soviet intervention in Afghanistan; the "Gorbachev revolution"), French policy seems to remain largely unaffected by either external or internal factors. The change in West German policies even runs counter to what one would expect from looking solely at the "threat environment" such as the Soviet " For data regarding the U.S., see Graham (fn. 8); and Russett (fn, 1), 57-65; regarding Western Europe, see Eichenberg (fn, c), chap, 4; and Stephen Szabo, "West European Public Perceptions of Security Issues" (Manuscript, prepared for the US I A, July 1988); for Japan, see Bobrow (fn. 4). 484 WORLD POLITICS arms buildup of the late 1970s. The stability of Japanese nuclear policies, while in line with public attitudes, is also not explicable in terms of changes in the country's international environment. The Missing Link: Domestic Structures and Coalition-building Processes [f, first, public attitudes as such cannot account for differences in policies and, second, variations in the international environment do not explain them either, what is the missing link between mass public opinion and elite decisions in foreign and security policy? It is suggested here that one has to look at domestic structures and coalition-building processes to understand the impact of public opinion on the foreign policy of liberal democracies.17 Domestic structure approaches deal with tine nature of the political institutions (tine "state"), basic features of the society, and the institutional and organizational arrangements linking state and society and channeling societal demands into the political system. In other words, domestic structures determine how political systems respond to societal demands. Social movement research talks about "political opportunity structures. Several approaches can be distinguished. The first focuses on state institutions and has found its most prominent expression in the concept of "strong" and "weak" states. It emphasizes the degree of centralization of state institutions and the ability of political systems to control society and to overcome domestic resistance."4 Weak states have fragmented political institutions and are open to pressures by societal interest groups and political parties. Their ability to impose policies on society and to extract resources from it is fairly limited. Strong states, by contrast, con- 11 For a similar argument focusing on the effects of party systems and electoral rules, see Eichenberg (fn. 1), 235-41. 'J See, for example, Herbert P. Kitschelt, "Political Opportunity Structures and Political Protest: Anti-Nuclear Movements in Four Democracies," British Journal of Political Science 16 (January 1986), 57-85; and Sidney Tar row, Struggle, Politics, and Reform: Collective Action, Social Movements, and Cycles of Protest (Ithaca, N.Y.: Center for International Studies, Cornell University, 1989). For an overview, see David Meyer, "Peace Movements and National Security Policy: An Agenda for Study" (Paper prepared for the Annual Meeting of the American Political Science Association, Atlanta, Ga., August 31-September 3, 1989). I+ The concept was originally developed by Peter Katzenstein, "International Relations and Domestic Structures: Foreign Economic Policies of Advanced Industrial States," International Organization 30 (Winter 1976), 1-45; Stephen Krasner, Defending the National Interest: Raw Materials, Investments, and U.S. Foreign Policy (Princeton: Princeton University Press, 1978). For more recent discussions, see Michael Mastanduno et al., "Toward a Realist Theory of State Action," International Studies Quarterly 33 (December 1989), 457-74; and G. John Ikenberry, "Conclusion: An Institutional Approach to American Foreign Economic Policy," International Organization 42 (Winter 1988), 219-43. LIBERAL DEMOCRACIES 485 sist of centralized political institutions with strong bureaucracies; they are able to resist public demands and to preserve a high degree of autonomy vis-a-vis society. However, the "strong versus weak states" distinction, while parsimonious, is too simplistic to account for the variations between domestic structures. Weak states like the U.S. are sometimes able to conduct highly efficient policies, whereas strong systems might not always pursue forceful and energetic foreign policies.'5 Moreover, institutionalist approaches have been challenged as apolitical and therefore unable to explain specific policy outcomes. Rather than emphasizing state structures, one should analyze the coalition-building processes within societies and political systems/* These approaches focus on the "policy networks,11 that is, the mechanisms and processes of interest representation by political parties and interest groups that link the societal environment to the political systems, This concept emphasizes the ability of political actors to build consensus among the relevant elite groups in support of their policies.'7 Simply to set off the coalition argument against the institutionalist approach seems to be inappropriate, though. On the one hand, state structures do not determine the specific content or direction of policies. On the other hand, coalition building takes place in the framework of political and societal institutions. The structures of the society and the political system determine the size and strength of policy coalitions needed to create the support basis for specific policies. Domestic structures also seem to account for general features of foreign policies, the degree of stability as well as the level of activity and commitment.18 In sum, a "mixed" approach encompassing both institutional structures and coalition-building processes seems to be more appropriate. In [s See, for example, G. John Ikcnberry, "The Irony of State Strength: Comparative Responses to the Oil Shocks in the 1970s," International Organization 40 (Winter 1986), 105-37. See also the special edition of International Organization 42 (Winter 1988), "Approaches to Explaining American Foreign Economic Policy." 15 Sec Page and Shapiro (fn. 3). Note, however, that the study does not prove the bottom-up model. It correlates public opinion data with policy outcomes and does not control for the impact of elite opinion- It would be consistent with the data to assume, for example, that elite opinion changed first and then affected the general public, which in turn led to policy changes. LIBERAL DEMOCRACIES 501 most prominently expressed by the Committee on the Present Danger.& As the data presented above reveal, the change in public opinion toward "peace through strength" policies took place in late 1979/early 1980. The Soviet intervention in Afghanistan and not just the conservative part of the U.S. elites tipped the balance in favor of the antid^tente mood among the public. This contributed to the election victory of Ronald Reagan, and, as a result, to the anti-sALT coalition,57 In other words, while the events of the late 1970s and early 1980s do not confirm a simplistic version of the bottom-up model, they nevertheless show that U.S. public opinion played a role in bringing the anti-sALT coalition to power. A similar process can be observed in 1982-84. Reagan's military buildup and his cold war rhetoric quickly eroded the public support for his defense policy. The "politics of opposites" (Nincic) worked (see data above). This change as well as the stable public consensus against the first use of nuclear weapons provided issue publics such as the freeze campaign with a window of opportunity. The nuclear protest was a bottom-up movement that originated outside of Washington's arms-control elites8 It quickly gained ground, penetrated the more skeptical arms-control community, and found support among numerous social organizations such as the Catholic church. Given the openness of the U.S. political system, the movement found easy access to the political elites. In 1983 the House of Representatives adopted a freeze resolution. Congress became increasingly active on issues of nuclear arms control. As a result of these coalition-building processes in the policy network, the power balance in the Reagan administration between hardliners in the Pentagon and more pragmatic conservatives in the State Department slowly shifted in favor of the latter.^ By 1984 and continuing throughout 1985, that is, before Mikhail Gorbachev entered upon the scene, Ronald Reagan had softened his rhetoric and adopted a more compromising stance on arms control. The superpower relationship improved, nuclear- 5* On the latter, see Mas Rampelmann, ed., Alerting Amertca: The Papers of the Committee on the Present Danger (Washington, D.C: Pergamon-Brasscy's, 1984); and Jerry W. Sanders, Peddlers of Crisis: The Committee on the Present Danger {Boston: South End Press, 1983). w For details, see Bernd W. Kuh big, Amerikanische Rüstungskpntrollpolitik_. Die innergesell* schaftlichen Kräfteverhältnisse in der ersten Amtszeit Reagans (1981-1985) (Frankfurt am Main: Campus, 1988); and Barry Posen and Stephen Van Evera, "Defense Policy and the Reagan Administration: Departure from Containment," International Security 8 (Summer 1983), 3-45. The best account of U.S. nuclear arms control policy during Reagan's first term is Strobe Talbot, Deadly Gambits (New York: Knopf, 1984). 3s The most comprehensive account of the freeze movement is Meyer (fn. 35). 59 For details, see Strobe Talbot, The Master of the Game: Paul Nitze and tke Nuclear Arms Race (New York: Knopf, 1988). See also Ernst-Otto C2empiel, "U.S. Policy towards the Soviet Union under Carter, Reagan, and Bush" (Paper presented to the Annual Convention of the International Studies Association, Washington, D.C, April 10-14, 1990); and Joseph (fn. 9), chap. 5. 502 WORLD POLITICS arms-control talks resumed in early 1985, and the U.S. defense budget began to decline by 1986. However, the freeze campaign failed to achieve its immediate goals, and the shift in Reagan's policy did not produce substantial arms-control results until the late 1980s (inf Treaty 1987). But public opinion clearly affected the ability of the issue public and interest groups to promote their cause and thereby influence the coalition-building process among the elites, which in turn produced tangible, albeit limited, results in U.S. policies. The impact of public opinion on the conduct of U.S. policy toward the Soviet Union was also noticeable in the spring of 1989, When the Bush administration came into power, government officials conducted a comprehensive review of Reagan's policies and apparently concluded that a more cautious approach should be adopted toward Moscow. However, as discussed above, by that time the Soviet threat had already withered away for the American public, so much so that the public reacted negatively to Bush's early foreign policy statements. The administration quickly adjusted to the public mood, however, and in June 1989 the president assured the public and the allies that the U.S. was firmly supportive of Gorbachev and the politics of perestroika.^0 While public opinion did not determine U.S. policy toward the Soviet Union during the 1980s, it nevertheless left a discernible mark on the decision-making process. The analysis confirms that the U.S. domestic structure is a comparatively open system that allows societal actors to mobilize support and to affect the balance of forces within the policy network. Nevertheless, the fragmentation and decentralization of the political system work against the stabilization and institutionalization of policies. Policymakers were able to decrease the salience of the issues in question by symbolically adjusting to the public mood, as was the case in 1984-85. The public impact was more limited regarding the substance of policies, mainly because the domestic structure does not provide institutional support for lasting consensus among the elites (as, for example, a strong party system would). In sum, the very openness of the U.S. system to societal pressures also limits their impact. weaknesses of a "nuclear monarchy": france The French domestic structure, as compared with that of the U.S., provides public opinion with only limited influence on foreign policy deci- *° This account is based largely on newspaper articles. See, e.g., "Cheney Remarks on Soviet Future Ruffle the White House's Feathers," New YorJ{ Timet, May 1, "Bush Asks an End to Divided Europe," New Yorfy Ttmes, [une i, 1989; and "Bush Policy Makers Reach Uneasy Balance on an Approach to the Soviets," New Yůr\ Times, July 1, 19Š9. LIBERAL DEMOCRACIES 503 sions. With the possible exception of the ending of the Algerian war,*51 there is indeed not much evidence that any of the major foreign policy decisions by the presidents of the Fifth Republic were taken in response to public opinion. This holds true for the building of an independent nuclear force, the withdrawal from nato's military command, the policy of detente toward the Soviet Union in the 1960s and 1970s, and the gradual shift toward more pro-European attitudes under de Gaulle's successors. In most cases public opinion was deeply split on the issues. Moreover, the French presidents were usually able to create a public consensus supporting their foreign policies after the event.*3 A similar pattern prevailed during the 1980s. As shown above, French public opinion did not develop in a markedly more hawkish way than did public opinion in Germany or Japan. Nevertheless, the new Socialist president, Mitterrand, adopted a more hostile approach toward the Soviet Union than that held by his predecessors, who had basically continued the legacy of de Gaulle's independent policy of detente. While Mitterrand's policy change did not reflect trends in mass public opinion, it was in line with parts of the issue public and the political elites. French left-wing intellectuals who had been fairly pro-Soviet during the 1950s and 1960s "discovered" repression in the USSR in the early 1970s and turned overwhelmingly anticommunist as a result. When President Gis-card D'Estaing tried to continue detente in the aftermath of the Soviet intervention in Afghanistan, he was chastised by the intellectuals and the media. His successor, Mitterrand, changed the policy, not least of all for domestic reasons, in an effort to isolate and marginalize the French Communist Party (PCF) on the Left.*11 The limited role of French society and public opinion in influencing policy decisions was also apparent during the early 1980s when France was the only major Western European country that did not face a significant public security debate and protest movements against nuclear weapons. This may be explained in part by the fact that the new U.S. inf missiles were not about to be deployed in France. Still, the French favored nuclear arms control and were as opposed to Reagan's policy as was any other European public. However, the French peace movement never developed into a mass movement; on the contrary, it was not only 61 Sec Grosser (fn. 22), 161-71. 61 For details, see Cohen (fn. 12); Grosser (fn. 21); Ho worth (fn. 3 r); and Aid rich and ConncLl (fn. 21). *3 For details, sec Cohen (fn. 2a); Julius W. Friend, Seven Years in France (Boulder, Colo.: Wcstvitw Press, 1989), 75-81; and Michel Ta.tu, "Val^ry Giscard d'Estaing ct U detente," in Samy Cohen and Marie Claude Smouts, eds., La politique exiA-ieure de VaUry Giscard d'Esta-ing (Paris: Presses de la fondation nationals des sciences politiques, 1985), 196-317. 504 WORLD POLITICS marginalized by the mass media but also internally split along ideological lines from the very beginning. In that sense, it reflected the fragmentation of French society/54 While the French centralized political institutions and the state -dominated policy network seem to constrain the public impact on foreign policy, there are a few instances in which public opinion apparently did play a role. One recent example was Mitterrand's decision in 1987 to endorse the inf treaty, despite the objections of his conservative foreign and defense ministers and of the majority of the French security community, who thought that the zero option was detrimental to Western deterrence policy, One year before the presidential elections, Mitterrand was apparently aware that French public attitudes overwhelmingly endorsed the U.S.-Soviet agreement. The decision is an example of presidential dominance in foreign policy matters even in times of "cohabitation"; it also suggests that public opinion sometimes affects French foreign policy directly, through presidential decisions rather than through coalition-building processes among the elites.^ While French public opinion only marginally influenced foreign policy decisions during the 1980s, the process did not resemble the top-down model. The French public did not just follow the policies of the elites or of Mitterrand. A comparison between public opinion and elite opinion shows that the attitude gap between elites and masses in France is one of the largest in Europe.^ French public opinion seems simply to be disconnected from the policy-making process. The limited role of French society in the country's foreign policymaking process does not, however, result in a highly active policy unhampered by domestic constraints and able to concentrate solely on external factors, as the concept of "state strength" would assume. The powerful, centralized state institutions of the Fifth Republic were designed to integrate a heavily divided and fragmented society. The foreign policy of de Gaulle and, albeit to a lesser extent, of his successors was intimately linked to the efforts to strengthen the societal support basis for the insti- *+ In addition, there were historical reasons for the isolation of the French peace movement. The notion of pacifism was identified with the politics of appeasement in the 1930s, and the largest French peace movement, the Mouvement de la Paix, was never able to distance itself convincingly from the Communist Party. For details, see Howorth and Chilton (fn.39). 65 In this particular case, Mitterrand was apparently influenced by public opinion not only in France but also in West Germany and by the fear that rejection of zero inf would lead to German neutralism. Cf. folyon Howorth, "Consensus and Mythology: Security Alternatives in. Post-Gaullist France," in Aid rich and Connell (fn. 21), 16-34; an<^ LePrestre (fn. 38). ^ See Eichenberg (fn. 1), 222-24. UBERAL DEMOCRACIES 505 tutions of the Fifth Republic.ä? To this day the official ideology emphasizes French unity,grandeur ^ and independence from external influences and alliances, notions that cut across ideological cleavages.*8 As a result, the famous French consensus on foreign and defense affairs remains fragile. The French state pays a price for excluding society from the foreign policy-making process. Rather than being able to pursue an autonomous and active policy. French decision makers seem to be constantly concerned with stabilizing a vulnerable elite consensus. from top-down to bottom-up: the federal republic of germany A review of West German foreign policy reveals that the impact of public opinion increased over time and that the decision-making process became more democratic. In the 1950s, when Chancellor Adenauer integrated the Federal Republic into the Western alliance and the European Community, general public opinion was as deeply divided on the issue as were the political elites and the parties. However, Adenauer went ahead, and by about i960 the public as well as the opposition had accepted his security policies.69 This top-down pattern gradually changed over the next two decades. In 1968-70, when Chancellor Brandt began his Ostpolitik and concluded the treaties with the Soviet Union and Eastern Europe, public opinion as well as the political elites were again divided on the policy. The Christian Democratic opposition (CDU) launched a vehement campaign against Ostpolitik and was even successful in depriving the government of its majority in Parliament. This time, however, and unlike the 1950s, the populace decided the issue: Brandt won the 1972 national elections primarily on the basis of Ostpolitik Throughout the 1970s the public consensus in favor of detente stabilized, and by 1980, despite the Soviet arms 6-1 Sec Philip G. Cerny, "Gaullism, Nuclear Weapons, and the State," in Howorth and Chilton (fn. 39), 46-74; and David Hanlcy, "The Parties and the Nuclear Consensus," ibid., 75-93. On the institutional uncertainty of the French system as a dominant feature ai the country's policy network, see Ashford (fn. 28). 46 "Independence," for example, was understood in an anti-Soviet sense by the French conservatives, in an anti-American way by the Left, and in a neutralist sense by nationalists at both ends of the political spectrum. 69 For details, see Arnulf Baring, Am Anfang war Adenauer: Die Entstehung der Kanzler-demokratie (Munich: Deutscher Taschenbuch-Verlag, 1971)', Cioc (fn. 36); and Hans-Karl Rupp, Außerparlamentarische Opposition in der Am Adenauer (Cologne: Pa hL Rügen stein, 1970). 7° For details, see Helga Haftendorn, Sicherheit und Entspannung: Zur Außenpolitik der Bundesrepublik Deutsehland, 1955-1982 (Baden-Baden: Nomos, 1983); Wolfram F. Han-rieder, Germany, America, Europe (New Haven: Yale University Press, 1989), 170-219; Clay Clemens, Reluctant Realists: The Christian Democrats and West German Ostpolitik (Durham, N.C: Duke University Press, 1989); and Christian Hacke, Die Ost- and Deutschlandpolitik^ der CDU/CSU (Cologne: Wissenschaft und Politik, 1975). 506 WORLD POLITICS buildup and the intervention in Afghanistan, it was firmly in place, as shown in the data above. The opposition Christian Democrats, although internally divided, continued to oppose detente and in 1980 conducted another anti-Ostpoliti^ election campaign, hoping it would succeed because of the deterioration in East-West relations, Unlike the anti-salt coalition in the U.S., however, they failed. In November 1980, that is, two years before he became chancellor, Helmut Kohl finally declared the battle over, and the CDU accepted Ostpoliti\?1 In sum, general public opinion was crucial in affecting the coalition-building process within the party system and, thus, in bringing about the German detente consensus. Unlike the French case, the domestic structure of the Federal Republic not only proved to be more open to societal influences, but it also institutionalized the consensus in the policy network, thus allowing for a highly active Ostpolitikj. General public opinion and societal actors also made a difference during the nuclear debate of the 1980s.^ The nuclear controversy originally began at the elite level—mainly within the SPD—in 1977-78, in the debate over the neutron bomb. NATO's inf decision of 1979, the new "cold war" between the U.S. and the USSR, and the election of Ronald Reagan then triggered mass opposition. New peace movements launched their protests against the deployment of inf missiles. As with the freeze campaign in the U.S., the protesters in Germany relied on general public opinion as a support basis regarding two issues: (1) opposition to a belligerent U.S. policy and (2) opposition to nuclear weapons as instruments of war fighting and to their deployment on German soil. While the peace movements failed to prevent the inf deployment in 1983, they had a profound impact on the coalition-building process among the elites. First, support for vigorous arms-control efforts increased within all parties, in particular within the CDU. Second, the "disarmers" entered the party system. Since 1983 policy decisions within the SPD elite have resulted from policy compromises between disarmers and the traditional arms-control establishment, but excluding members of the right wing of the party, such as former Chancellor Helmut Schmidt. Moreover, a disarmament party, the Greens, entered the political scene. Third, the nuclear elite consensus broke down as a result. By ?' For details, see Risse-Kappen (fh. 17), pt. B. 71 For the following, see details ibid. On the peace movements in particular, see Josef Jan-ning et ah, eds., Fritaensbewegungen (Cologne: Wissenschaft und Politik, 1987); and Thomas Rochon, The Politics of the Peace Movement in Western Europe (Princeton; Princeton University Press, 1988). LIBERAL DEMOCRACIES 5Ö7 1984 the SPD had adopted an antinuclear stance. In sum, the coalitions within the policy network gradually moved to the Left—in line with trends in general public opinion and in response to the protests of the issue publics. The consequences of both the detente debate and the antinuclear protests were seen in the mid- to late 1980s. Germans were the first to embrace Gorbachev's new policy, because it responded favorably to what had previously emerged as a consensus on Ostpolitii{ and common security. Moreover, the opposition to new nuclear weapons increasingly included the CDU, the consequences of which became apparent during nato's debate on the modernization of battlefield nuclear weapons in the spring of 1989. There was a coincidence of public opinion, left-wing an-tinuclearism, and conservative fears that short-range systems would "sin-gularize" Germany in the aftermath of the inf treaty. As a result, Chancellor Kohl had no choice but to resist a modernization decision by nato. To conclude, the role of public opinion in West German security policy changed over time and moved closer to the bottom-up model during the 1980s. The general and the issue publics decisively affected the coalition-building processes within the party system, which in turn considerably narrowed the range of options available to policymakers. However, the change has nothing to do with public opinion as such. For example, the public had been as ambiguous about the deployment of nuclear weapons in West Germany in the 1950s as it was in the 1980s, although the policy impact was markedly different in the latter case, Rather, the difference in influence seems to reflect a change in the domestic structure that opened up German politics to societal pressures. Adenauer's "chancellor democracy" changed into a "party democracy" during the 1960s, thereby establishing a democratic corporatist policy network. On top of that, new social movements—from the student protests of the late 1960s to the environmentalists of the 1970s and the peace groups of the 1980s—affected the political culture and were crucial in opening the party system to societal influences.^ In contrast to France the domestic structure of the Federal Republic was open enough to allow for such changes in the first place. And unlike the United States, the strength of the party system in the policy network allowed for the institutionalization of the societal consensus. '! For analyses of these changes, see Kendall Baker et al., Germany Transformed: Political Culture and the New Politics (Cambridge: Harvard University Press, 1981); Samuel Barnes et al., Political Action: Mass Partietpation in Five Western Democracies (Beverly Hills, Calif.: Sage, 1579); Karl-Werner Brand et al., Aufbruch in erne andere Gesellschaft: Neue soziale Bewegun-gen tn der Bundesrepttblik^ Deutschland (Opladen: Westdeutscher Verlag, 1983); and Richard Stdss, ed., Paruien-Handbuch, 2 vols, (Opladen: Westdeutscher Verlag, 1983-84). 508 WORLD POLITICS conservative leaders versus public opinion! japan Given the Japanese domestic structure and the stability of Japanese public opinion, the case is expected to resemble Germany. If one considers the low-key approach of Japanese military security policy over the last forty years, the self-imposed arms-control measures such as the non-nuclear principles and the limits on defense spending, and the reluctance to play a more global role in world politics, Japanese public opinion and foreign policy seem to be more or less in line. Indeed, almost every analysis of Japanese defense policy points to the constraining forces of public opinion on the margins of maneuver of the policymakers.^ At first glance, Japanese foreign policy during the 1980s seems to confirm the analysis. Prime Minister Nakasone met with firm domestic opposition when he talked about making Japan an "unsinkable aircraft carrier"75 and tried to move Japanese security policy toward a more pro-defense stance and to initiate a major military buildup program, While all Japanese governments of the 1980s declared that they wanted to abandon the principle of not exceeding 1 percent of the gnp for defense expenditures, they were only able to do so from 1986 to 1989. The principle was reinstated in December 1989. Finally, Japanese policy toward the Soviet Union remained in line with the comparatively high degree of public threat perception (see data above). Japan was reluctant to react to Gorbachev's peace initiatives and to the changes in Soviet foreign policy. In contrast to the Federal Republic of Germany, Tokyo only grudgingly accepted the inf treaty after having insisted that the Asian-based Soviet SS-20S be included in the zero option.76 However, this description gives only half of the picture. The very fact that Nakasone and his successors were able to change Japanese defense posture as well as to implement a more globally oriented foreign policy belies the notion that Japanese policy-making is dominated by trends in public opinion. Moreover, the 1 percent gnp limit on defense procure- ?4See, for example, J, W. M. Chapman, R. Drifte, and I- T. M. Gow, japans Quest for Comprehensive Security (New York; St, Martin's Press, 1982); Scalapino (fn. 23); and Wclfield (fn. 24). 75 In March 1983, 72% of Japanese were worried about Nakasone's defense policy; 48% disagreed with his remarks about the "unstnkable aircraft carrier." While public opposition to Nakasone's policies declined somewhat over the years, he was nevertheless unable to change the prevailing trends- For data, see Hastings and Hastings, tds^ Index to International Public Opinion, igSz-tgSj (New York: Greenwood, 1984), 316. Far analyses of Japanese foreign policies during the 1980s, see, e.g., Drifte (fn. 23); idem, Japan's Rise to International Responsibilities (London: Athlone Press, 1990); and Malcolm Mcintosh, Japan Re-Armed (London: Frances Pinter, i^H6). 1s See details in Drifte (fn. 75, 1990), 48-58. LIBERAL DEMOC R ACT ES 509 ment seems to be largely symbolic and primarily meant for domestic consumption. If one uses nato criteria to measure military expenditures, Japan had the third highest military budget worldwide in 1988. Finally, Tokyo's tough-minded policy vis-a-vis Gorbachev's Soviet Union reflected public opinion on only a superficial level. As the above-quoted data suggest, most Japanese would have been perfectly willing to support a policy that tried to resolve the grievances with the USSR—particularly the dispute over the northern territories—by means of diplomacy. The contradictory evidence is explainable: simply put, the LDP's conservative leaders did not share the general public's outlook on Japan's role in the world and moreover felt under constant U.S. pressure to build up the military. In essence, security policy seems to have remained the one issue-area in Japanese politics in which the beliefs of the leaders of the conservative coalition did not accord with public attitudes. As a result, Nakasone and his successors had to conduct their defense policy against the prevailing trends in public opinion; unable to change it and given Japan's domestic structure, however, they had to acquiesce to it. They thus muddled through while stretching public tolerance to its limits. If one accepts the notion that the Japanese governments indeed felt vulnerable to public opinion, the strange coincidence of low-key rhetoric (except for Nakasone's early years) and considerably more active security policies becomes explicable. Thus, in the 1980s Japanese public opinion placed broad, but nevertheless effective, constraints on the actions of policymakers. The impact of the public was reinforced by the peculiarities of the Japanese domestic structure, in particular, the emphasis on consensus building in the policy network. Unlike France and more comparable to Germany's corporatist process, issue publics and societal actors were able to affect the coalition-building processes within the LDP by forcing it to take account of the viewpoints of the opposition parties. As a long-term consequence of the i960 crisis, the conservative coalition tried to include the opposition parties in major decisions concerning foreign and security pohcy, The institutionalization within the LDP of "policy tribes" corresponding to the particular ministries and Diet committees also enhanced the consensus-building capacity within the LDP and reduced the impact of factionalism.77 " I thank Nobuo Okawara for pointing this out to me. Sec also Calder (fn. 41), 463. The most recent domestic turmoil in Japan on how to deal with the crisis and the war in the Persian Gulf confirms the analysis that mass public opinion severely contrains. the country's ruling conservative coalition. 510 WORLD POLITICS Conclusions The study Leads to three major conclusions. First, mass public opinion mattered in each of the four countries, albeit to very different degrees. Policymakers in liberal democracies do not decide against an overwhelming public consensus. In most cases, mass public opinion set broad and unspecified limits to the foreign policy choices- In the U.S. and— more recently—Germany it also defined the range of options available for implementing policy goals. Additionally, general issue salience seems to be of minor importance, once there is a clear majority in favor of certain policies. In the Japanese case the issue salience of foreign policy was comparatively low; decision makers nevertheless felt the pressure of mass public opinion. In sum, public consensus may substitute for a lack of issue salience, whereas highly mobilized societal actors may be able to influence policies even in the absence of such consensus. Neither was the case in France. However, there are discernible limits to the impact of the general public on foreign and security policies. Rarely does general public opinion directly affect policy decisions or the implementation of specific policies. In sum, the bottom-up and the top-down models are too simplistic to fit the reality in the four countries. Second, most of the previous research on the policy impact of public opinion treats the domestic decision-making process as a black box and directly compares opinion polls with policy outcomes. This study suggests, however, that the indirect effects ofpublic opinion are far more important. To the extent that the empirical material was available,?8 it could be shown that the main role of the public in liberal democracies is to influence the coalition-building processes among elite groups. In the U.S. and the Federal Republic of Germany public opinion had a crucial impact on elites as they reconfigured their coalitions; groups in line with public preferences were able to carry the day. In Japan the public forced the ruling conservative coalition to integrate the opposition into the decision-making process, at least to some extent. Moreover, support by mass public opinion seems to be essential for issue publics, public interest groups, and other societal actors tc influence policy decisions. This accounts for the partial success of the peace movements in Germany and the U.S. and of the otherwise marginalized opposition parties in Japan in affecting the debates in the policy network. '* This proved to be a major limitation of this study, particularly in the French case. The prevailing' realist paradigm in international relations and the focus of most studies on elites do not encourage scholars to trace the impact of public opinion and societal actors on decisions in detail. LIBERAL DEMOCRACIES 511 For both the political elites and societal actors, mass public opinion proves to be a resource for strengthening one's position, in the coalition-building process. The third conclusion concerns the usefulness of the domestic structure approach. Domestic structures seem indeed to account for the differences between the four countries. The degree of mass public consensus on foreign policy correlates strongly with the degree of societal fragmentation, in particular, the ideological divisions between Left/liberals and Right/ conservatives. Japan as the most homogeneous society of the four countries also enjoys the most stable public foreign policy consensus. By contrast, the divisions in French public opinion on security policy reflect the country's ideological and class cleavages. The U.S. and Germany are cases in between. Furthermore, there is strong empirical evidence that domestic structures are the intervening variable between public opinion and foreign policy. Under given international conditions and despite relatively similar public attitudes across countries, variances in the interaction between the general public and elites in the foreign policy-making process can be explained by differences in domestic structures. The degree to which political institutions are centralized and the degree to which the state dominates the policy networks seem to be the determining factors. In both categories, France and the U.S. are located at opposite ends of the spectrum. The two countries also form the two cases in which public opinion seems to have the greatest (U.S.) and the least (France) impact on policy decisions. The Federal Republic of Germany changed over time from top-down to a more interactive or even bottom-up process, which in turn seems to reflect changes in the overall domestic structure. In the Japanese case the stable public consensus in favor of a nonmilitary security policy severely constrained a political elite that tried to pursue a more active foreign policy. Moreover, societal protest such as was experienced during the i960 crisis left a discernible mark on the consensus-building process in the conservative coalition and, as a result, on Japanese foreign policy. At first glance, the empirical findings appear to confirm the distinction between strong and weak states. The stronger the state institutions and the greater their control of the policy networks, the less access the public has to the policy-making process. However, while the strong versus weak state distinction seems to make sense regarding the input side of the political system, it is flawed if one considers the relationship between societal demands and policy outcomes. The strong French state and cor-poratist Japan both pursued comparatively low-key and noncommittal 512 WORLD POLITICS policies with respect to the Western alliance system, arms control, and the Soviet Union during the 1980s, but for very different reasons. French foreign policy reflected not so much state strength as the fragility of an elite consensus and a general lack of policy legitimacy in the absence of societal input into the decision-making process. But Japanese foreign policy was constrained by the institutionalization of the domestic consensus—elite and public—in the policy network. The comparison between the weak U.S. state and corporatist Germany is also revealing. Both countries pursued comparatively active and highly visible foreign policies (Germany at least in the East-West context). But there are important differences that reflect the respective domestic structures. The frequent volatility and unpredictability of American policies seems to result from constantly shifting coalitions in Washington's domestic process. In comparison, the Federal Republic has pursued a more stable foreign policy over the last twenty years. In contrast to the U.S., but similar to Japan, the German policy network and the political structure emphasize consensus building and the institutionalization of coalitions in support of specific policies. As a result, the Germans have less direct input into the policy-making process than the Americans have, but their impact on policy outcomes seems to last longer. I am aware that this paper suffers from a number of limitations and that further comparative research is indicated. First, while this article more or less covers the spectrum of domestic structures in liberal democracies, it deals neither with authoritarian regimes nor with one-party systems, Does the empirical evidence confirm the conventional wisdom that public opinion does not count in those countries, or do systems that are not democratic in the Western sense provide the masses with different ways to influence policies? Second, the role of cultural factors to facilitate or to complicate the public's impact on foreign policy has to be examined systematically. The emphasis on consensus building as a social value in both Japan and Germany suggests, for example, that the domestic structure approach as presented here might have to be expanded. Finally, this paper does not deal with the role of the media, though it is fair to assume that the media play a critical role both in the process of opinion formation and in conveying public attitudes to the policymakers.