According to three academics from Duke University who are advising the Bush administration, the key to public support for war in Iraq is the perception of whether the United States will be successful in its mission. The argument is that even if the war costs mount -- real monetary costs, as well as military casualties -- Americans will continue to support the effort as long as they believe success will ultimately prevail.
The thesis is explained in a couple of research papers, and I am not one to demean such extensive and serious efforts. Duke academics Peter Feaver, Christopher Gelpi, and Jason Reifler use sophisticated statistical models to support their conclusions, and they are certainly worth examining.
I do have reservations, however, about the overall effort to establish causal connections among the various concepts. The underlying assumption is that certain perceptions about the war affect people's judgment in one area, and that in turn affects their judgment in another area, and so on until we arrive at the ultimate opinion -- whether to "support" the war in Iraq. The causal model cannot be proved, as least by the data obtained by the three authors; in this case, causality is more an act of faith than a provable dynamic.
One reason I am skeptical about the identification of a "key" attitude is that different measures of whether the United States will be "successful" in Iraq give quite different results. For example, the Hotline/Westhill poll conducted July 7-10 found 60% saying it was likely that the United States is likely "to succeed in Iraq." Yet, that same poll found that only 41% of Americans were confident that in the next year, "the U.S. will have achieved its goals in Iraq."
How can one claim the United States will "succeed" (not defined) if at the same time one doubts the United States can achieve its goals? The authors suggest that the vague question, which does not define "success," correlates more highly with war support than the question about the United States achieving its goals. If so, it is probably not because the "success" question predicts support, but rather the other way around -- that people who support the war are likely to say it will be "successful" in some vague way. When asked more precisely about achieving goals, people are much less positive. Given the large difference in results between the two questions, 41% confident of achieving goals versus 60% saying the United States is "likely" to succeed, neither question is probably the "key" to war support, but both questions -- over time -- should correlate with changing support.
Further evidence of fuzzy opinion on Iraq comes from a Gallup survey, conducted July 25-28. It did not ask about likelihood of succeeding, but rather about winning or losing. All respondents were asked whether the United States "can or cannot" win the war in Iraq, and whether it "will or will not" win the war.
The results show that 47% of Americans believe the United States can and will win the war, 6% believe it can but will not win, and 37% believe it cannot and will not win. This is a much more divided public than what the vague "success" question suggests in the Hotline/Westhill poll.
But even the Gallup results are not fixed. Half the sample was asked first if the United States "can or cannot win" the war, while the other half of the sample was asked first if the United States "will or will not win" the war. Then, the two half-samples were asked the respective follow-up questions.
The results show that when people were asked first whether the United States "can" win the war (Form A), 49% said yes. But among those who were asked this question second in Form B (after having been asked if the United States "will" win the war), 59% said the United States can win -- 10 points higher than in Form A. About half said the United States "will" win the war in each form.
These results suggest that expectations of whether the United States will "succeed" or "win" in Iraq, and even whether the United States "can" succeed or win, are highly susceptible to slight differences in question wording and question context. Such susceptibility casts doubt on the argument that the question on expectations of success is a "key" determinant of support for the war -- for the concept is too vague and its measurement too fuzzy to produce a precise figure.
Instead, opinion about success or failure is probably just one of many issues about the war, all of which are tied to each other and influence each other, but none of which can be easily identified as "the" key to all other opinions. It is tempting to latch onto simple concepts that appear to provide clear explanations of public opinion. But such opinion is too complex, too messy, to yield to such hopes.