Adam Berinsky
Department of Political Science, MIT
February 18, 2004
I want to talk about how to generate public support for a war, and under what
situations will that support disappear. To look at a current war, Gallup asked
in December 2003, "Is the Iraq War worthwhile?" 63% supported the
war, 33% opposed. More recently (February 2004), that support had dropped to
49%, with 49% opposed. The Christian Science Monitor (February 14) offered 4
reasons: a rise in casualties, lack of evidence of WMD, an "expectation-reality
mismatch," and David Kay's recent public comments. However a letter from
Tom Daschle and Nancy Pelosi expressed dissent at the Congressional level, and
on the other side, Bill Reilly expressed similar sentiment. This dissent at
the elite level I believe offers an important clue.
So what drives public opposition to war? The Christian Science Monitor's explanation is emblematic of John Mueller's theory of wartime opposition, which points to external events, in particular casualties. So the theory goes, as casualties rise, so do the costs of war. The public will not tolerate high costs. The thoeory is not explicitly political, resting on the notions of collective rationality, which may be problematic. In 1996 Eric Larson modified the theory to include costs and benefits. John Zaller, by contrast, points to elite discourse as the driving factor. If elite discourse is unified, which he calls the mainstream pattern, more informed citizens will be more supportive of war. He contrasts this to a polarization pattern, arguing that as discourse becomes polarized, support among citizens will dwindle.
To support this theory, Zaller looked at Vietnam. According to CBS News-New York Times surveys, in 1964 those members of the mass public that paid attention to politics tended to support war, irrespective of political ideology (ie. conservative versus liberal). By contrast, in 1970, as political awareness increased in members of the mass public, those that called themselves liberals were increasingly anti-war, while conservatives were increasingly pro-war. He also looked at the first Gulf War. In October 1990 surveys showed a mainstream pattern, similar to 1964, at work with respect to the Gulf War. In January 1991, on the other hand, surveys indicated that people who identified themselves as Democrats or Republicans showed opposite tendencies to back a war given their political awareness - trending upward for Republicans and trending downward for Democrats. Both sets of polls took place before the war began; all that had changed was the unity of elite discourse.
So we have different explanations with different explanatory mechanisms. The problem, at least in Korea and Vietnam, is that the mechanisms of both theories are collinear with time. Over time cumulative casualties increased, and also over time elite discourse split. I believe that World War II is therefore the critical test. Over time in WWII cumulative casualties increased, while at the time elite discourse remained unified.
To test WWII, I have looked at a series of polls and compared, for those that backed FDR versus those that opposed him, people's support for the war given how informed they were. A poll from 1941, before the U.S. joined the war, showed that people who approved of FDR exhibited a steady increase in support for helping England, at the risk of getting involved in the war, given their level of information. By contrast, those opposed to FDR were significantly less supportive of helping England given their level of information, though the slope of this curve is not negative. I speculate that in 1940 this curve might have been negative, given the greater prominence of isolationism as an opposition platform.
Another poll from 1942 asked whether or not the US Army should be sent abroad and whether or not America should take an active part in world affairs after the war is over. Both FDR and Wilkie supporters showed an increasingly positive response for given levels of information. Another poll from 1943, which asked whether the US should play a larger role in world affairs and take an active part in an international organization after the war, shows a similar trend. And importantly a Roper poll from February 1944, after US had started taking significant casualties, again showed that people with different politics continued to remain on the same page given their level of information. To control these tests by looking at a domestic issue, the same poll asked, "should the next administration work with businessmen or take care of the people," and supporters of FDR showed sharply differing opinions given information level than did Republicans.
This affirmation of Zaller's theory begs a number of questions. What are the conditions under which elites remain unified? Perhaps arguments that have been made about the mass public (aversion to casualties, cost/benefit analysis, etc) can be applied to elites. Can we develop a unified theory of public opinion and war? Are there important implications for theories of democracies at war? Are their any applications for policy? To illustrate the importance of this issue to policymakers, Peter Feaver states, "The Beltway wisdom is that public support collapses in the face of casualties, and by casualties I mean fatalities." Marine Lt. Col. (Ret.) Gary D. Solis, on the other hand, argues, "We've never been as casualty-averse as either the politicians said or the military thought, but that can change in an instant."
Rapporteur: David Blum
back to seminar schedule, Spring 2004