View the h-diplo Discussion Logs by month
View the Prior Message in h-diplo's July 2011 logs by: [date] [author] [thread] View the Next Message in h-diplo's July 2011 logs by: [date] [author] [thread] Visit the h-diplo home page.
The exchange among Reiter, Desch, and Downes raises a whole host of important questions. In the spirit of continuing the conversation, I would like to push a bit on the alternative views of democracy and foreign policy, some of them extending beyond the boundaries of what they said. Of course the framing issue is the relative impact of domestic and international factors: the level of analysis question or, in the older German formulation, Primat der Innenpolitik versus Primat der Aussenpolitik. Realists minimize internal factors because they see the external environment as highly compelling, although neo-classical realism seeks to restore domestic politics to a seat at the table, even if usually below the salt. To continue with this metaphor, arguments for the greater effectiveness of democracies also set Realists' teeth on edge because they can quickly bleed over into triumphalism and the claim for the moral superiority of democratic foreign policy. The connection here is not absolute because effective policies need not be moral. Indeed, one of the striking findings of the Reiter and Stam book that set off this discussion is that most of the wars that democracies initiate end within 6 months,[1] which suggests predatory rather than self-defensive behavior. But nevertheless it is easy—too easy?—to move from claims for the effective policy to claims for wise ones, and from there to ethically warranted ones. Reiter goes at least some of this distance in his last paragraph.[2] Realists object not only because this implies a greater freedom from external restraints than their world-view allows, but because even without the recent experience of the Bush administration this faith in a democratic foreign policy can lead to over-extension, self-righteousness, and efforts to install democracies abroad. Indeed, as Reiter also notes in his last paragraph, these are the sins that Realists such as George Kennan and Hans Morgenthau associated with democracy. They were building on a tradition of skepticism about the ability of democracies to conduct foreign policy that goes back to the birth of the modern form of this kind of government. Public opinion was seen as fickle and prone to demagogic appeals. In his least-known book, Kenneth Waltz, ever the contrarian, made a good case for democracies' abilities, but it is interesting, if a digression here, to note that the current proponents of this view rarely acknowledge and let alone seek to directly rebut the earlier conventional wisdom.[3] Reiter's own arguments, although fascinating and important, do not go to these broader questions, and I raise them here both to indicate the framing and to encourage others to take them up. The argument that democracies are particularly likely to initiate wars that they will win brings out a possible conflict with effectiveness, at least as measured by maximizing the country's subjective expected utility (SEU). Leaving aside the now-familiar question of whether bargaining and anticipations of what others will do should lead states to be able to avoid war and focusing only on one side's choice, it can make sense for a state to start a war that it expects to lose. Utilities as well as probabilities are relevant, and knowing who is likely to win speaks only to the latter. A state that expects to lose can rationally start a war if the gains from winning are extremely large and/or the costs of losing are relatively low. Furthermore, even a negative expected utility does not mean that the decision was ill-founded. The crucial calculation involves comparisons not with the current status quo, but between the expected world in which the state starts the war with the expected one in which it does not. The expected utility of the former could be negative and yet starting the war would be appropriate if the expected utility of following alternative policies was lower still. In this reading, a democracy could initiate a war it expected to lose, which violates Reiter's theory, but would do so as a reasonable path to the national interest, which fits with the spirit of Reiter's and related theories of the advantages of democratic foreign policies. To turn this around, Reiter's argument implies that democratic leaders over-weight the probabilities attached to various outcomes and under-weight if not ignore the utilities. This makes sense in terms of electoral politics if not the good of the country because the public sees only the outcome that is achieved and is not likely to appreciate that a gamble that failed may have been worth taking or that the bad outcome might have been better than what was likely to have happened had the state followed a different course of action. This may help explain the case of Vietnam that Reiter and Downes debate (and to which I will return in another context below). Even if Johnson believed that the chances of winning were slight, fighting could have been necessary if the expected costs of not doing so were extremely high (as indicated by the infamous Domino Theory) and the costs of losing after making a major effort were not much greater, or even were less (what Assistant Secretary of Defense John McNaughton articulated as the "good doctor theory"—i.e., dominos would be less likely to fall if others believed that the US did all it could to try to save South Vietnam).[4] Narrowing the focus a bit, Reiter and others stress the importance of democratic responsiveness and accountability. The exchange does a service in showing that these mechanisms are different, and indeed can conflict. They also can be rooted in long-standing normative arguments about democracy. Should democratic leaders seek to represent the views of their constituents or to do what they think is in the national interest? The counterpart empirical question is when leaders are willing to sacrifice power to do what they think is best, which, contrary to some overly simple interpretations of rational choice theory, clearly is sometimes the case. Like Downes, I had read Reiter and Stam as having stressed the retrospective or accountability argument rather than the popularity or representing one. I agree with Reiter that the two are not contradictory, but this does not mean that they cannot come into conflict. Indeed the desire to curry current favor by following a policy that meets with popular approval would not lead democracies to win the wars that they initiate unless one also believes that public opinion will not only shun losing wars, but has enough information to judge whether this is likely. Even if the "marketplace of ideas" works well, this latter condition seems to me a stretch. One could perhaps argue that popular wars will generate greater enthusiasm and that the resulting greater effort will increase the chance of victory, but this too is a stretch. So it may be possible to link popularity and war-initiation, but the extra step to victory is more problematic. Putting this aside, we can make one of our traditional four-bold tables here, with one dimension being the state of current public opinion and the other being estimates of whether the war will succeed or fail. Since there is every reason to expect political leaders to be influenced by both incentives, the selection effect should operate most strongly in the upper left-hand corner. Here war gives leaders a boost in popularity, producing immediate benefits, and also increases the chances of reelection. Indeed, I think an implication of this is that under these circumstances leaders should be willing to opt for war even if they believe the foreign policy gains are negligible, or perhaps even negative. The opposite diagonal is equally clear. Democratic leaders should not go to war when public opinion is against it and the chances of victory are not great. At first glance, this seems trivial, but other theories could generate contrary predictions. One rooted in the military-industrial complex could argue that these special interests could push for even a losing war, at least if it was a limited one. Proponents of the diversionary theory of war could argue that domestic unity might be gained by fighting, again as long as the loss was not too severe. (This example raises the broader question of whether Reiter thinks what is important is pre-war opinion or expectations of what it will be once the war has started.) A strict Realist would argue that public opinion would be ignored, and as noted above, even a long-shot might be taken if it were in the national interest. As Reiter points out (pp.103-04) the off-diagonal cells yield competing predictions from the two mechanisms that can be at work, and I would expect leaders to feel cross-pressure in these situations. Evidence for theory would be provided if the record of deliberations contain concern with these conflicting imperatives, although official records rarely focus on public opinion—the belief that this is not a legitimate foreign policy concern inhibits it from being raised in official meetings and I suspect discourages note-takers from recording it when it does come up. Alternatively, we could develop propositions about the conditions under which each of the mechanisms was likely exert most influence. For example, the need for current popularity is likely to weigh most heavily on governments with low approval ratings; considerations of the final outcome are likely to be most potent when the results of the conflict are expected to be in by the time of the next election. More intriguingly, the two impulses can interact rather than simply add or subtract.[5] Covert action may be most appealing when the public does not favor the state using force to replace another regime, but this outcome would be seen as benefitting the state and a success for the administration. In some cases, perhaps, the combination of current unpopularity and expectations of success (at a low cost) could lead not to hesitation but to enhanced incentives to fight. The domestically oriented calculations here would be that if the opposition party follows public opinion and opposes the war, it will not only bring acclaim to the administration but will discredit its opponents.[6] The other off-diagonal contains cases in which public opinion supports going to war but the leaders doubt that it will be successful. In the extreme, leaders could be pushed into war, as in one interpretation origins of Spanish-American War. When the pressures are less great, they still could be sufficient to push over the brink a leader who was otherwise ambivalent about whether to fight. Furthermore, to the extent that motivated biases and the psychological need to avoid painful value trade-offs operate, the perception that the public wants war could lead decision-makers to believe that this course of action is in national interest. Here too interaction effects might be possible as popular pressures could have limited impact when a government has a great deal of support on other grounds, or, conversely, could dominate when an election is in the immediate offing and would be conducted while fighting was going on but before the results could be predicted. These cases would not fit Reiter's basic argument, however, because the desire for popularity leads democracies to initiate war despite the fact that the chances of victory are not good. Reiter and Downes disagree about the status of American opinion on the eve of Vietnam War. I believe that it was confused, volatile, and subject to question-wording effects. But perhaps more important is the general problem, one that bedevils politicians as well as scholars. This is that the former must be concerned not only with whether opinion now seems to favor going to war, but whether it will remain steadfast if the war goes on any length of time (and, as noted, Reiter and Stam show that the tendency for democracies to win sharply diminishes as wars last more than six months). They also have to estimate how the public will react if the state does not fight and things develop badly, and this is the dilemma that Johnson faced. The choice was not between going to war and maintaining the status quo, but between fighting and being pushed out of Vietnam. He was well aware of how much the Democrats had been damaged by "losing China" fifteen years earlier and had reason to believe that even if the public did not support fighting in Vietnam, it would turn against him even more sharply were he to allow Saigon to fall. A final methodological point is the need for relevant comparisons. Some but I think not enough discussions of whether and how democracies are different have made the necessary comparisons to the behavior of non-democracies. But we may also want comparisons to cases and circumstances in which the dogs do not bark—instances in which democracies do not fight despite the fact that the theory leads us to expect that they will. Of course, much depends on what we are claiming for the theory: if the argument is that popularity and/or perceptions of likely victory are "only" necessary conditions for the democracy to initiate a war, then cases in which they are present but the democracy stays at peace do not disconfirm the theory. But if the necessary conditions are met with great frequency but war is quite rare, the significance of the finding would be significantly reduced. If more is claimed for the theory, then the negative cases are even more important. In any event, looking at them would help us think through the boundary conditions for the theory and what else we think is operating. Reiter, Desch, and Downes have done a great service in digging deeply into the subject of democracies and war. I hope these observations will join with their research in stimulating further thought. Robert Jervis Columbia University 1 Reiter and Allen Stam, _Democracies at War_ (Princeton University Press, 2002), pp. 169-70. 2 "Democracy and Victory: Final Thoughts," in H-Diplo|ISSF "Exchange on Democracy and Victory," https://www.h-net.org/~diplo/** ISSF/PDF/ISSF-Roundtable-2-12.**pdf<https://www.h-net.org/%7Ediplo/ISSF/PDF/ISSF-Roundtable-2-12.pdf>p.113. 3 Kenneth M. Waltz, _Foreign Policy and Democratic Politics: The American and British Experience_ (Boston: Little, Brown, 1967). 4 _The Pentagon Papers_ (Senator Gravel Edition), vol 3 (Boston: Beacon Press, 1971), p. 582. 5 For a general discussion of interaction effects see Robert Jervis, _System Effects: Complexity in Political and Social Life_ (Princeton: Princeton University Press, 1997); for an application, see Jervis, "Correspondence: Thinking Systemically About China," _International Security_ 31 (Fall 2006), pp. 206-08. 6 Of course it might not do so: see Kenneth Schultz, _Democracy and Coercive Diplomacy_ (New York: Cambridge University Press, 2001).
|