For decades, the conventional wisdom in academic and governmental circles was that the U.S. public gets weary about military engagements as soon as the body bags begin to mount. While the conventional wisdom has been challenged as of late, few scholars have attempted to fully understand the conditions under which the U.S. public is willing to bear the human costs of war. That is exactly the task that Christopher Gelpi, Peter Feaver, and Jason Reifler set out to accomplish in their book. Their argument and findings not only contribute to the scholarly literature but also carry important implications for policymakers.

The authors acknowledge that many forces influence the U.S. public's tolerance of military casualties, but they place prominence on the intersection of two factors: expectations that the military mission will be successful and beliefs concerning whether the country was right to enter the conflict in the first place. Of the two, they argue that the former carries more weight. They also contend that these foreign policy assessments not only are relevant to the public's willingness to bear the costs of war but also influence more general assessments, such as voting in presidential elections. They argue that in the domain of electoral decision making, existing theory suggests that beliefs about the rightness of the war should play a bigger role than expectations of success because the choice is between two candidates and retrospective evaluations tend to take prominence in elections (Morris Fiorina, Retrospective Voting in American National Elections, 1981).

An impressive array of data is used to test how the intersection of expectations of success and beliefs about the rightness of entering a war influence casualty tolerance and voting in a presidential election context. In addition to looking at extant data, the authors conducted nine original surveys between September 2003 and November 2004. This enabled them to develop a much better measure of casualty tolerance than is typical in the literature, to embed some experiments into the surveys, and to have readings of casualty tolerance during an actual conflict, the Iraq war.

Early in the book, the authors take primary aim at John Mueller's argument, in War, Presidents and Public Opinion (1973), that support for war declines in proportion to the log of casualties. Using the cases of Vietnam, Korea, Somalia, Lebanon, and Iraq, they show convincingly that the simple log relationship does not hold. Rather, the effect of mounting casualties on support for each conflict seems to depend in part on whether the military is being successful. These findings serve as preliminary support for one of the key parts of their argument.

Gelpi, Feaver, and Reifler also dismiss the log of casualty hypothesis with experimental data in which they are able to directly manipulate the number of casualties and level of confidence in a nonspecific military mission. Confidence works roughly as expected, and the effects are comparable to casualties, which is consistent with their theory. I was expecting to see some interaction between confidence and casualties, as was tested in the first empirical chapter. While the authors did not test for this interaction, the combination of data from an experiment and real conflicts present a very strong case against the conventional wisdom. Since it is not really possible to directly manipulate the number of casualties in an actual conflict, the authors develop an alternative measure. They ask individuals how willing they are to tolerate casualties across different types of military conflicts. The baseline begins with no casualties and continues to 50, 5,000, 50,000, and 500,000 casualties. They combine responses to generate an ordinal scale. This measure is a vast improvement over proxies for casualty tolerance in the literature that ask whether a given war is a mistake or whether the war was worth it. However, since it is self-reported, it may still be subject to biases.

The authors employ this dependent variable in their analysis of the mission in Iraq. They find strong confirmation for their argument that casualty tolerance is a function of the interaction between the likelihood of success and an individual's perception over whether the United States should have entered the war in the first place, with the former exerting a stronger effect. For example, even individuals who did not believe that the United States was right to enter the war are about 50% likely to tolerate casualties if they believe victory is very likely. They also find support for other scholarly accounts, namely, that the public will be more tolerant of casualties if they perceive a strategic interest (e.g., Bruce Jentleson, “The Pretty Prudent Public: Post-Vietnam American Opinion on the Use of Military Force,” International Studies Quarterly 36 [no. 1, 1992]: 49–74); consensus among domestic elites (e.g., Eric V. Larson, Casualties and Consensus: The Historical Role of Casualties in Domestic Support for U.S. Military Operations, 1996); and support from the international community (e.g., Steven I. Kull and M. Destler. Misreading the Public: The Myth of a New Isolationism, 1999). However, the substantive effect of these other accounts pales in comparison to those found for the authors' theory.

Gelpi, Feaver, and Reifler then look at how the intersection of success and rightness of the Iraq war influenced voting decisions in the 2004 presidential election. In this case, they find that both factors affected voting decisions, though perceptions of the rightness of the war had a stronger effect. One limitation to the analysis of voting is that the authors do not include other issues that were prominent in the election, such as the economy. Not including these issues may inflate the effect of foreign policy assessments on voting decisions. To their credit, they did break down the analysis by those most concerned with the economy or social issues, but this is a rough approximation of the relative influence of foreign policy evaluations versus other considerations on voting decisions.

Finally, the authors explore the factors that determine perceptions of success and the rightness of the war. As expected, perceptions of success influence perceptions of the rightness of the war and vice versa; however, the overlap is not perfect. Perceptions of the rightness of the war are also more favorable if individuals believe that Saddam Hussein had weapons of mass destruction, see a link to Al Qaeda, and support the Bush Doctrine. Meanwhile, individuals have higher evaluations of success when they perceive elite consensus and a clear plan for winning the war, have confidence in the U.S. and Iraqi governments, and are hawkish. Many of these indicators were signals being sent by the administration as benchmarks of success. One thing the authors do not explore directly in the analysis is whether casualty tolerance influences perceptions of success. They do show that people do not mention casualties in an open-ended question about how they conceive of success, but this does not seem a strong enough test for dismissing concerns of endogeneity.

The picture of the public that emerges here stands in stark contrast to the public of conventional wisdom, which buckles at the first sight of body bags and cares little about foreign policy. The authors note that “members of the American public can and do form reasonable and coherent attitudes about the use of military force” (p. 241). Even though those members lack knowledge about foreign policy and some hold misinformed beliefs, they are able to form attitudes by following elite cues. These findings suggest that leaders face fewer constraints in getting involved in military conflicts. However, the public will judge decisions to enter a conflict when the next election rolls around, and so there are still some constraints for those seeking reelection.

One measure of a book's quality is whether it generates questions for future research, and this one certainly fits the bill. In the conclusion, the authors point out several avenues for future work. One avenue has to do with how the public reacts to conflicting cues from the president and the media. Another has to do with how tolerant the public is of civilian casualties. They also suggest that future scholars probe more deeply into the psychology behind support for casualty tolerance. Gelpi, Feaver, and Reifler admit that the sources of beliefs need not be cold and calculating and suggest that they may be affected by attitudes toward risk and sunk costs. This certainly piqued my interest and made me wonder to what extent perceptions of the rightness of war and success are the result of objective information, mis- or biased information, and emotional reactions.