Book Search:  


Google full text of our books:


Paying the Human Costs of War:
American Public Opinion and Casualties in Military Conflicts
Christopher Gelpi, Peter D. Feaver & Jason Reifler

Book Description | Endorsements | Table of Contents

COPYRIGHT NOTICE: Published by Princeton University Press and copyrighted, © 2009, by Princeton University Press. All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher, except for reading and browsing via the World Wide Web. Users are not permitted to mount this file on any network servers. Follow links for Class Use and other Permissions. For more information, send e-mail to

This file is also available in Adobe Acrobat PDF format

Chapter One


Perhaps the most important task that American citizens entrust to their elected officials is the decision to deploy the country’s military forces in combat. In making such decisions, leaders place the lives of American citizens—and the citizens of other nations—in the balance. For decades scholars and politicians have sought to understand the conditions under which Americans are willing to support their leaders’ decisions to use military force. In this book, we show that many conditions are important for shaping the public’s willingness to bear the human costs of war, but most important of all is the public’s expectations that the military operation will be successful.

Initially, scholars believed that the public was not capable of placing constraints on the use of force either because Americans would reflexively “rally ’round the flag” (Verba et al. 1967), or because their attitudes toward foreign policy lacked structure and content (Lippmann 1955; Converse 1964). Later work examining the reaction to the Vietnam War continued to see the public’s reaction to the use of military force as reflexive and unthinking, but drew the opposite conclusion about the direction of that reaction, seeing the public as unwilling to tolerate any use of military force that resulted in even a few American deaths (Luttwak 1996; Klarevas 2000).

Ever since the Vietnam War, policymakers have worried that the American public will support military operations only if the human costs of the war, as measured in combat casualties, are trivial.1 The general public, so the argument goes, is highly sensitive to the human toll, and this sets severe constraints on how American military power can be wielded. Political leaders who engage in costly military ventures will face their own sure demise at the ballot box. Americans stop supporting military operations that produce casualties, and voters punish political leaders who deliver such policies. Indeed, ever since the rejection of the Versailles Treaty and the rise of isolationism in the United States, but especially since the Vietnam War, the conventional wisdom has cited public reluctance to bear the costs of global leadership as the Achilles heel of American foreign policy.2 The conventional wisdom is so strong that it is enshrined in Army doctrine and regularly invoked by U.S. leaders.3

In this book, we argue that the American public is more discerning and deliberative than most pundits and policymakers expect, and thus American foreign policymakers are less constrained than the conventional wisdom implies. Casualties do not produce a reflexive collapse in public support. Under the right conditions, the public will continue to support even relatively costly military operations. In a similar way, casualties are not as toxic for public support of the president as popularly believed; combat deaths do not translate directly into political death. To be sure, the public is not indifferent to the human costs of American foreign policy, but the constraints placed by American public opinion are not as limiting as popularly believed. Instead, the public appears to take a reasonably levelheaded cost-benefit approach in forming attitudes toward military missions.

Our central argument is that—within this cost-benefit framework— when it comes to supporting an ongoing military mission in the face of a mounting human toll, expectations of success matter the most. Many factors—the stakes, the costs (both human and financial), the trustworthiness of the administration, the quality of public consensus on the foreign policy goal in question, and so on—affect the robustness of support. But the public’s expectation of whether the mission will be successful trumps other considerations. When it comes to voting on a president who has led the country into a costly war, the relative weights of factors shift; expectations of success still matter, but the most important factor appears to be whether the public views the initial decision to start the war as correct.

Of course, actual success, let alone perceptions of success, are not entirely under the control of policymakers; nor are public judgments about the rightness or wrongness of the initial resort to military force. The president has neither a free hand nor a blank check. But the image of the American public as a paper tiger—a mirage of strength that collapses in the face of casualties—is as incorrect as it is popular.

We show that our argument makes the best sense of the voluminous survey data that are now available on this subject. We make wide-ranging use of surveys administered by others, but the centerpiece of our book is data from proprietary national surveys that we designed and conducted from October 2003 through November 2004. These data, representing the results of some 8588 interviews with adult Americans, are an unmatched resource of information, the most extensive and detailed compilation of public attitudes toward casualties of which we are aware. The coincidence of our field research with the ongoing war in Iraq provided an unprecedented, albeit tragic, opportunity to gauge public attitudes toward casualties as events on the ground evolved—and also allowed us to reshape our research focus accordingly.

Indeed, it is impossible to investigate a topic like this today without having the ongoing conflict in Iraq uppermost in mind, and we do look very closely at public opinion on the Iraq war. Concern, however, for what might be called “the public’s stomach for costly military action” predates the Iraq war (as does our initial research design).

The issue, in fact, is as old as the Republic. General George Washington and the Continental Congress worried about the willingness of the American colonists to continue to pay the costs of war with Britain. “There is a danger,” the general wrote to the Congress, “that a commercial and free people, little accustomed to heavy burdens, pressed by the impositions of a new and odious kind, may not make a proper allowance for the necessity of the conjuncture, and may imagine that they have only exchanged one tyranny for another.”4 President Lincoln likewise confronted the issue in the Civil War. From Lincoln’s vantage point, the mounting human toll of the war seemed on a collision course with the 1864 election.5 His concerns seemed justified, moreover, in light of the impact that war casualties had on Republican candidates in the 1862–63 midterm elections.6 President Wilson spoke eloquently about the wastefulness of war, opining that “never before have the losses and the slaughter been so great with as little gain in military advantage.”7 The terrible human toll of World War I left what one scholar called a “dark shadow” on the American public, and enshrined the “casualty issue” as a crucial constraint on American foreign policy.8 Concern over casualties drove U.S. efforts in the 1930s to avoid involvement in another major European war, and shaped the way the war was ultimately fought.9 Of course, World War II proved to be the bloodiest American war (not counting the Civil War) and, compared with every military operation since, World War II is held up as the exceptional instance of the American public having a strong stomach for war. At the time, however, President Roosevelt and his military commanders worried greatly about public casualty tolerance and went to extraordinary lengths to manage the public’s reception of adverse news.10 On the other side, both Adolf Hitler and Japanese leaders were convinced that the ethnic composition of the American public, and the democratic government’s responsiveness to that public, meant that the economic and military potential of the United States would not be realized in combat; the United States might look tough on paper but, once bloodied, it would collapse.11

Concern over the public’s tolerance for casualties was arguably a defining feature of the major military operations of the Cold War: Korea and Vietnam. Vietnam, in particular, is remembered as the war that established beyond dispute that the American public will not support a “long and bloody conflict in a faraway land,” as one North Vietnamese leader put it.12 Each of the leaders who brought the United States into these wars saw his political headaches multiply with the mounting combat toll, and each was denied a second term as a result. America’s enemies drew the predictable inference about the United States: In the words of Chairman Mao, “[I]n appearance it is very powerful but in reality it is nothing to be afraid of, it is a paper tiger.”13 Vietnam thus raised the question, “Why do big nations lose small wars,” and the answer lay in the difficulty of sustaining popular support in the face of mounting costs.14

Since Vietnam, of course, the issue has only grown in prominence. The hasty exodus from Beirut after the tragic Marine barracks bombing in October 1983, the hasty retreat from Somalia after the infamous “Black Hawk Down” Ranger raid in October 1993, the force protection mind-set in the Bosnia and Kosovo missions—all reflect a conventional wisdom that the American public will reflexively turn on any military mission that involves a human toll. Political leaders’ fears of public casualty phobia further help explain decisions against U.S. military intervention in such places as Rwanda, Congo, or Sudan. Edward Luttwak summarized the conventional wisdom well: “The prospect of high casualties, which can rapidly undermine domestic support for any military operations, is the key political constraint when decisions must be made on which forces to deploy in a crisis, and at what levels.”15 The Weinberger/Powell doctrine further enshrined the view that public support for military operations was a scarce resource—difficult to mobilize and easy to lose.16 The view that public resolve was easily overcome was further reinforced by the fact that three influential groups bought into the idea: determined enemies of the United States, media elites, and policymakers. Thus, Saddam Hussein premised his strategy in the first Gulf war on the idea that casualties would defeat the U.S. popular will, even if it did not defeat the military.17 Slobodan Milosevic knew that he could not directly defeat NATO’s military might in the Kosovo war, but believed that inflicting even modest attrition on NATO forces would be sufficient to prevail politically.18 And in his infamous November 1996 fatwa, Osama Bin Laden quite explicitly invoked American casualty phobia in Somalia as evidence for his strategic premise that the United States could be defeated with only a relatively modest level of damage: “[W]hen tens of your soldiers were killed in minor battles and one American pilot was dragged in the streets of Mogadishu, you left the area carrying disappointment, humiliation, defeat and your dead with you. Clinton appeared in front of the whole world threatening and promising revenge, but these threats were merely a preparation for withdrawal.”19 As Steven Kull and I. M. Destler show persuasively, American media elites and policymakers agreed.20

The terrorist attacks of 9/11 did not erase the issue from public commentary. Indeed, within a few weeks, pundits were fretting about the bleak prospects for the military operation in Afghanistan and warning that victory would require an unacceptably high military commitment.21 Even with the Taliban routed and al Qaeda in full retreat, the question of casualty tolerance continued to dog the war, most famously in disputed allegations that a fear of casualties drove the administration to withhold American troops from the assault on Tora Bora in December 2001.22

Pundits were once more invoking the “lesson of Vietnam” as the war with Iraq loomed, arguing that the public simply did not have the stomach for the kind of bloody fighting required in Iraq.23 Based on reports of interviews with former Iraqi commanders, Saddam Hussein himself evidently thought as much, specifically invoking the Somalia incident as proof that the American public was casualty phobic.24 The war was barely a week old, when observers began to worry that news of combat fatalities would cause public support to collapse.25 As the security problems in Iraq persisted long after “major combat operations” had ended, the daily reminders of a mounting human toll were carried in grim headlines, often paired with analyses that warned that public support was eroding precipitously.26 The bloody toll in Iraq featured prominently in the 2004 electoral campaign, and countless observers speculated that President George W. Bush, like Presidents Johnson and Truman before him, would find his electoral hopes lost in a far and distant war. Bush did defy the expectations and win reelection, but the political toll of the Iraq war remained a preoccupation for political observers of all stripes—and for the administration as well.

The relative popularity of the war ebbed and flowed, but as this book went to press (fall of 2008), public opinion had turned sharply negative according to a number of measures, and we could observe some substantial public pressure to begin withdrawing U.S. forces from Iraq. In January 2007 the president increased the number of U.S. troops in Iraq in response to the continuing civil violence there. The president’s policy was widely unpopular, but the intensity of public pressure to withdraw troops remained somewhat difficult to gauge. Surveys demonstrated substantial shifts in opinion about the conditions for withdrawing U.S. troops depending upon the specific phrasing used in the question. For example, CBS surveys done in both April and May 2007 asked respondents a simple “yes or no” question regarding whether they support setting a specific timetable for withdrawing American forces from Iraq and found that nearly 65 percent of the public favored such a timetable. Questions that allowed respondents to express support for a specific withdrawal timetable, or for a more flexible policy that would link funding of the war to specific “benchmarks” of progress by the Iraqi government, indicated that “benchmarks” was the median position. Specifically, a Fox News poll indicated that 39 percent supported setting a timetable, 32 percent supported “benchmarks,” and 24 percent supported giving the president’s policies more time to work. Thus while the public has become increasingly dissatisfied with the Bush administration’s handling of this conflict, the public pressure to end U.S. participation in the war has been remarkably slow to materialize. At first blush, then, the Iraq war did not settle the question of whether or not the American public was casualty phobic.

The effect of casualties on American public opinion, whether measured in terms of support for the military operation or support for reelecting the leader who opted for war, is thus both an enduring and timely issue. It is also a question of great importance. Concerns about casualties drive both American foreign policy and American electoral campaigns. It also drives the behavior of America’s most determined foes. In short, the issue is worthy of the sustained attention we give it here, and so we proceed as follows. In the rest of this chapter, we contrast the conventional wisdom with the existing scholarly literature in two separate areas: first, public opinion on casualties; and second, the role of foreign policy in elections. We finish the chapter by laying out a summary of the theoretical argument that we will test in the rest of the book.

Chapter 2 assesses our argument in light of aggregate data on public opinion from surveys conducted by others during times of military conflict. We begin by focusing on five military operations that are most often invoked by the conventional wisdom—Korea, Vietnam, Lebanon, Somalia, and Kosovo. Then we look in-depth at the most recent military operation in Iraq, through the U.S. presidential election in 2004. In chapter 3 we argue that many of the debates and issues raised in chapter 2 cannot be definitively answered through an examination of aggregate data alone—as is most commonly done. We examine individual belief systems about the use of force and demonstrate that key arguments in this literature can be measured and tested at the level of individual attitudes. In chapter 4 we return to the theoretical debates described in this chapter and test them at the individual level with a series of survey experiments about hypothetical military missions. Chapter 5 builds upon the results of our survey experiments by applying our argument to an ongoing real-world conflict—the Iraq War. We examine a series of proprietary surveys completed between October 2003 and October 2004. Chapter 6 extends our argument by examining the impact of the Iraq War on the 2004 presidential election and shows how our argument can help explain the Bush victory. Chapter 7 investigates the question left hanging by our argument: If the key attitudes in support for war and vote choice are “expectations of success” and “belief in the rightness of the war,” what deeper factors explain those key attitudes? Chapter 8 is a brief conclusion, summarizing our results and identifying questions for future research.


Casualty tolerance is a difficult concept to study, but over the years a fairly substantial body of scholarship addressing the issue has accumulated. Following our own earlier work, we distinguish between a variety of terms that are often treated as synonymous in popular discourse, but that are analytically quite distinct.27

By the noun forms of casualty tolerance, casualty sensitivity, and casualty shyness, we mean the overall willingness of the public to continue to support a military operation even as the human toll is rising; in theory, the public’s tolerance/sensitivity/shyness could be absolute, high, moderate, limited, or nonexistent. By the adjectival forms of casualty tolerant, casualty sensitive, and casualty shy, we mean specific claims about how casualties affect public support. Thus, the claim that the public is casualty tolerant is a claim that casualties do not substantially undermine public support for a mission; the opposite claim that the public is casualty shy means that casualties do substantially undermine public support. The claim that the public is casualty sensitive is simply a claim that the public views casualties as a negative, preferring less if possible. Virtually all research— including ours in this book—assumes that the public is casualty sensitive in this minimal sense. This is another way of saying that casualties are a human cost of war; we would prefer the same benefit, the goals of the war, at lower cost if possible.

The conventional wisdom described earlier, however, goes a step further, and claims that the public’s casualty tolerance is so low that even historically low numbers of casualties will undo public support for a military mission. We call this extreme form of casualty sensitivity “casualty phobia.”28 Casualty phobia is different from pacifism. Pacifism is opposition to any use of force; casualty phobia involves initial support for the use of force, but the support evaporates rapidly and irrevocably at the sight of body bags.

Finally, for the sake of clarity, we will use the terms casualty averse and casualty aversion to refer to the policies and behaviors that political leaders and the military implement with regard to this issue. Thus, the military can adopt casualty averse rules of engagement, depending on their understanding of the casualty aversion policies of the political leadership; these policies are themselves at least partly in response to political leaders’ personal casualty tolerance and also their beliefs about whether the general public is casualty shy or even casualty phobic.

The conventional wisdom is that the public is casualty phobic, but the scholarly consensus is otherwise. The scholarly consensus has evolved over time, partly in response to the growing sophistication of methods of assessing public opinion and partly in response to perceived changes in the nature of public opinion as technology and America’s role in the world itself have evolved. The literature is best understood as three layers of interlocking (and not necessarily successive) debates.

The first debate concerns whether casualties affect public support for the war according to a fixed pattern of inexorable decline or whether the public views casualties and the use of force through a rational cost-benefit calculus. Early research during the Vietnam War emphasized what came to be called the “rally ’round the flag” effect—the way public support spikes during crises—and the degree to which determined political leaders can reinforce public resolve through decisive action; viewed this way, public casualty sensitivity was not a debilitating constraint on American military power.29

As the Vietnam War continued and public support eroded, the scholarly consensus shifted. Jeffrey Milstein was a pioneer in applying the sophisticated techniques of the McNamara “whiz kids” to the problem of public support for Vietnam. He found a variety of strong correlations: As the U.S. military commitment increased and as casualties increased, public support dropped, whereas public support climbed when the burden was shifted to the shoulders of the Vietnamese themselves.30

John Mueller built on this work with a landmark study of public opinion in the Vietnam War (with a comparison to the Korean War).31 Mueller is famous for arguing that public support for the Vietnam War dropped in proportion to the log of casualties: “While [the American public] did weary of the [Korean and Vietnam] wars, they generally seem to have become hardened to the wars’ costs: They are sensitive to relatively small losses in the early stages, but only to large losses in later stages.”32 Viewed in toto, Mueller’s finding cuts against the casualty phobia thesis. Public support for an ongoing military operation did not drop catastrophically with mounting casualties. On the contrary, casualties drained public support only slowly. Mueller, however, was arguing that public support dropped reflexively, and more to the point, inexorably. Mueller’s oft-quoted study thus fixed in the public mind the idea that support for Vietnam buckled as the body-bag toll mounted, and this gradually hardened into the conventional wisdom that the public is reflexively casualty phobic.

Mueller later reinforced the “inexorable decline” view with his analysis of public opinion during the first Iraq war, which emphasized that public support was far more precarious than the euphoria over the quick victory might indicate.33 Scott Gartner and Gary Segura revised this argument somewhat, noting that support for the Korean and Vietnam wars declined with logged casualties during periods when the casualty rate was low, but in periods of high casualty rates, then public support drops with marginal casualties, not logged cumulative casualties.34 Other research also showed that there was a direct link between mounting casualties, antiwar protests, and then subsequent changes in U.S. governmental policy.35

This view of a fixed pattern of declining public support with rising casualties was challenged in a series of studies that showed that public support did not inexorably decline. On the contrary, Benjamin Schwarz showed how the rally ’round the flag effect not only inured political leaders from the negative impact of casualties early on in a military operation, but also might even have driven the public to favor escalation of a military operation rather than the withdrawal suggested by the “body bag syndrome.”36 Moreover, several major scholarly investigations assessing public opinion and national security during the Cold War painted a collective picture of a “rational public,” one very capable of responding to elite debates and weighing the complexities of foreign policy. 37 The public, in this way, went through what Alvin Richman called a simple “ends-means” calculus.38 Eric Larson applied this argument to the surveys originally analyzed by Mueller, as well as public opinion from subsequent wars, and found, contra Mueller, that the complex cost-benefit calculation fit the data better than a reflexive, logarithmic response.39 Today, the scholarly consensus is nearly unanimous in favor of the “rational cost-benefit” model, and Larson’s oft-cited version serves as a point of departure for most subsequent research in the field.

The cost-benefit model, however, is not necessarily a rebuttal of the casualty-phobia thesis, and, indeed, the model raises what can be considered the second big debate in the academic literature: If the public applies something like an economistic rational calculation about war, how inelastic is the public’s “demand” for war? 40 It is at least theoretically possible for the public to have such an elastic demand—to be so “price sensitive”—that even marginal numbers of casualties cause public support for the war to collapse. Thus, in more recent work, Mueller explicitly accepts the cost-benefit model, but, writing before 9/11, he argued that the public saw so little benefit in most military missions that in effect the cost-benefit calculation was functionally equivalent to a casualty-phobic posture.41 Likewise, Louis Klarevas endorses the cost-benefit model, but then elsewhere argues that for some key categories of missions the public sensitivity to casualties is so high that trivial numbers of casualties can produce a “Somalia Syndrome.”42 Most scholars who have examined public opinion polls closely come down on the other side, however, concluding that the demand, while not completely inelastic, is nevertheless not as price sensitive as to approximate casualty phobia.43

The third debate, which our work engages head on, takes the issue of elasticity a step further: What factors shape the “elasticity of demand” for military missions? Put another way, under what conditions will the casualties cause public support for a given mission to decline more rapidly or more slowly?

For instance, there is a wide scholarly consensus that stakes do matter. The more vital the public views the military mission, the higher the price the public is willing to pay to achieve it, other things being equal. This insight, however, borders on a tautology, since the way you can be sure that the public considers the military mission to be more vital is that the public shows a willingness to tolerate more casualties in conducting it.

There is also a wide scholarly consensus that multiple factors may be at work at the same time.44 What distinguishes different authors in this third wave of debate, however, is the pride of place they give to certain factors. One can identify in the existing debate, five different claims that argue, in effect, that “other factors may also matter for driving casualty tolerance but this is the factor that trumps the others.”45

First, Bruce Jentleson argues that the “pretty prudent” public bases its casualty tolerance on “the principal policy objective (PPO)” envisioned by the military operation.46 PPOs involving “foreign policy restraint” will be accepted by the public as important and thus worth even a serious price; these included the traditional military tasks of using force to coerce an adversary engaged in aggressive action against the United States or its allies. Missions deemed “humanitarian intervention” enjoy public support only if the costs are relatively low. Still other missions, dubbed “internal political change,” are viewed as inherently dodgy adventures by a skeptical public; public support for these missions is hard to come by and easy to lose as costs mount.

Second, Eric Larson argues that public casualty tolerance follows domestic elite casualty tolerance.47 When domestic elites line up in a consensus behind the mission, public support will be robust even in the face of mounting costs, but when domestic elites are divided, then even small amounts of casualties will be highly corrosive of public support. By domestic elites, Larson primarily refers to potential political rivals in Congress.

Third, Steven Kull and his colleagues argues that public support for a military mission will be more robust if the public sees that other countries likewise support the mission and thus the United States is not obliged to bear the costs all by itself.48 Multilateral support may function as an elite cue—“this mission must be justified because lots of other states are supporting it” 49—or the public may simply prefer to have the burden distributed more evenly.

Fourth, a variety of scholars and pundits have identified what might be called the “contact” factor in shaping robustness of the public’s stomach for costly military ventures. One of the earliest systematic studies of the question was Donald Rugg and Hadley Cantril’s analysis that compared the views of person’s with draft-age family members versus those without such members to the prospect of war before the United States entered World War II; Rugg and Cantril concluded that there was no difference and, at least at that time, opposition or support was not a personal matter.50 On the other hand, Scott Gartner, Gary Segura, and Michael Wilkening showed that individuals from counties with higher casualty rates had greater opposition to the Vietnam War than individuals from counties with lower casualty rates; in other words, local losses increased casualty sensitivity.51 Similarly, Karol and Miguel (2005) looked at county-level aggregate data and found casualties within some counties reduced the proportion of the vote for Bush relative to its level in 2000, though they find that casualties had no impact in other counties. This argument is another way of understanding the popular claim advanced by Charles Moskos, Charles Rangel, and others that the general public, whose children are at risk in military combat, are more sensitive to casualties than are elites, whose children by and large do not serve in the military.52

Fifth, our own prior research identified expectations of success as the crucial factor.53 If the public believes the mission will succeed, then the public is willing to continue supporting the mission, even as costs mount. When the public thinks victory is not likely, even small costs will be highly corrosive. Note that the critical attitude specified here is expectation of eventual future success, not necessarily assessments of how the war is going right now or most recently. Of course, recent experience can shape expectations of the future, but our claim was that the future judgment was the one that matters. It is the difference between how the patient feels right now versus how optimistic the patient is that he will get well eventually; the latter is the more crucial attitude for determining one’s tolerance for enduring pain.54

Beyond these five factors are a host of demographic factors that research has shown affect casualty sensitivity. Race is a significant factor, with African Americans being more sensitive to casualties than Caucasians.55 Gender is also significant, with women more sensitive than men.56 Education and age likewise matter, though they do not have a consistent effect; depending on the case in question, sometimes education and age are positively correlated with casualty tolerance and sometimes they are negatively correlated.57

Each of these studies establishes convincingly that the favored factor matters (and most also show that other factors matter, too). It should be noted that the cost-benefit approach does not preclude the possibility that in the very long run—something like the decade-long involvement in Vietnam—time may itself shape the calculation.

Every factor that is given pride of place in analyses—whether external multilateral support, or domestic elite consensus, or prospects of victory— is likely to be negatively affected to some degree if the war drags on indefinitely. But from a policymaking perspective, the inexorable decline is slow enough to provide a window for military operations, provided the other factors are favorable. Of these other factors, the demographics are less policy-relevant; policymakers seeking to shore up public support for a military mission as costs mount—or war protestors seeking to undermine public support—are not able to do much about changing demographics. The other factors—how the mission’s purpose is framed, the degree of elite or international consensus, the perceived likelihood of victory—are indeed in play for policymakers. Accordingly, we single them out for special attention.

The existing scholarly literature makes one further observation relevant to the shaping of our current project. Most pundit commentary treats public opinion in the aggregate with sweeping statements about overall casualty phobia or overall robustness of support. In fact, however, it makes more sense to view public opinion as an aggregation of different pockets of opinion, each with different responses to casualties. Our own review of public opinion on the use of force identified a quadripartite pattern that seemed to recur across a variety of different cases. The public was made up of solid hawks (roughly 30 to35 percent) who will support virtually any military mission virtually regardless of the costs; solid doves (roughly 10 to 30 percent) who will oppose essentially any mission regardless of costs; casualty-phobics (roughly 15 to 20 percent) who support a mission provided it is extremely low cost; and defeat-phobics (roughly 15 to 40 percent) who support a mission, despite mounting costs, provided that the mission is likely to succeed, but who turn on a mission provided that it looks like it is doomed to failure.58 With such a distribution, the public reservoir of support in the aggregate is adequate even for low-stakes missions that involve the cost of human lives. The exact percentages on any given mission vary with the stakes and a host of other factors, but in broad-brush terms one inference is that a resolved president can count on at least 45 percent support for any successful mission; this degree of support is adequate to carry on even as casualties mount, given the executive branch’s privileged position on foreign affairs and the likely nature of military conflicts that the United States would face.

In sum, we glean from the existing literature four key insights that serve as the point of departure for our study:

  • Public attitudes toward casualties are very difficult to assess and may change over time.
  • The public is not casualty phobic, but casualties do affect public support for military operations by counting as the costs in a cost-benefit calculus.
  • A range of factors shape the elasticity of demand for military operations— the rapidity with which casualties might undermine public support in any given mission. At present, however, no study is able to show how much one factor matters compared to the other relevant factors.59
  • While judgments are possible about public opinion in the aggregate, in fact individuals respond to casualties differently.

In this book, we also examine the way that casualties and attitudes toward the war in Iraq affected the 2004 election results, and in so doing we encounter yet another bit of conventional wisdom—this time about the link between foreign policy and elections.60 The conventional view is summarized by the pithy aphorism attributed to then-candidate Clinton’s campaign advisors in 1992: “It’s the economy, stupid.” Public opinion in general has been considered to be ill informed and unsystematic.61 Foreign policy evaluations in particular have been suspect and not considered likely to shape vote choice. When forming attitudes about the performance of the economy, citizens have their personal experience to fall back on. But foreign policy is so removed from the everyday lives of most citizens, it was argued, that it is simply unreasonable to think that what happened beyond U.S. borders would have a large impact on Americans’ political behavior. In support of this claim, many studies showed, at best, only weak evidence that foreign affairs affects the voting decision.62

This skeptical view was gradually challenged by scholarship that identified a “rational public,” capable of making reasonable or competent decisions from limited amounts of information.63 Indeed, research showed that citizens have reasonably structured attitudes concerning foreign policy; attitudes of foreign policy affect political evaluations; and citizens respond in understandable ways to changing world events.64 The public may not be very good at quiz bowl questions about international current events, but the public as a whole has stable and reasonable opinions that change in response to changes in the real world.65 Public opinion may be “latent” on many issues, but when activated by news events and especially by prominent debates within the elite over foreign policy options, it becomes a factor that policymakers must address.66

If the public has rational views about foreign policy, it is not so unreasonable to think that the distant world of foreign affairs can have a meaningful impact on political behavior like vote-choice. Voting, as Morris Fiorina has argued, involves both retrospective and prospective judgments, as shaped by an individual voter’s political predispositions.67 How well have these candidates performed in the past and which is likely to do the best job in the future, are questions to be measured against an individual’s own ideology, especially in terms of defining phrases like “how well” and “best job.”

It has long been known that economic evaluations have an effect on presidential approval and vote choice.68 An increasing amount of evidence has emerged showing that foreign policy judgments matter as well as, and in roughly equal magnitude to, economic evaluations. In an analysis of the 1980 and 1984 presidential elections, Aldrich, Sullivan, and Borgida find that foreign policy issues were just as powerful a vote determinant as domestic issues.69 Likewise, in a time-series analysis of aggregate quarterly presidential approval data, Nickelsburg and Norpath show that the president is as much “commander-in-chief” as “chief economist.”70 Adding major foreign policy events as predictor variables to their model, these international events matter at least as much as economic evaluations. Using individual-level data from several national random sample surveys conducted from 1983 to 1987, Wilcox and Allsop (1991) find that approval of Reagan’s foreign policy is consistently a good predictor of his overall approval, though, as a predictor, the strength of foreign policy approval relative to domestic issues does depend on the salience of economic or foreign policy issues at a given time.71 Nincic and Hinkley (1991) and An-nand and Krosnik (2003) also show that foreign policy attitudes affect the evaluation of presidential candidates. One intriguing study even found a link between the casualty rates at the district level and the electoral fortunes of members of Congress running in the 1862–63 congressional elections; the higher the level of casualties in a specific congressional district, the worse the incumbent fared in the election, though the substantive effect was quite small (an incumbent lost less than 1 percent of the two-party vote for every one hundred casualties in his district).72

The precise impact of foreign policy on electoral choice does appear to wax and wane with the flow of current events. Survey responses regarding the nation’s “most important problem” suggest that the economy is nearly always salient in the minds of voters, while concern about foreign affairs varies substantially, depending on world events. Foreign affairs will play a less prominent role in elections during a relatively quiet time internationally, say 1996, than they will in an election during the middle of a controversial war, say 2004.

In sum, we glean from the existing literature four key insights that serve as the point of departure for our study:

  • Foreign affairs in general, and political salient matters like war casualties in particular, do affect vote-choice.
  • Both retrospective and prospective judgments come into play when individuals make a vote-choice.
  • Behavior seen in the aggregate—for example, elections won or lost—are a function of individual-level choices, that is, individual voters choosing one candidate over another based on a host of factors, including foreign issues.
  • These effects are likely to be pronounced when foreign issues are salient, for instance, during wartime.


The foregoing review of the scholarly literature has five implications for the research design of this book. First, because the vast majority of the scholarship evaluating public attitudes toward casualties is based on aggregate data, we will emphasize analyses of individual-level opinion. Where appropriate we will also assess aggregate opinion, but aggregate data by itself cannot settle the debates that remain. Second, because casualty sensitivity is such a difficult concept to measure, we will use a variety of measures—aggregate versus individual-level opinion, hypothetical versus real-world scenarios, and direct versus experimental question designs— to ensure that our findings are robust and not an artifact of question wording or ephemeral survey context. Third, for these same reasons, where possible, we will draw on surveys designed explicitly to tap into attitudes toward casualties; relying on surveys where the casualty question is an afterthought may be misleading. Fourth, where possible, we will use appropriate multivariate statistical techniques to isolate and compare the separate (or interactive) effects of different factors on casualty attitudes; in this way we can to make a more confident assessment about the relative weights. Fifth, our model will incorporate both backward and forward-looking questions, since the attitudes of direct interest—whether to support a military mission and which candidate to elect as president— involves both retrospective and prospective judgments.

The centerpiece of our project is proprietary opinion survey data from nine original surveys that we designed and administered to different random national samples from 22 September 2003 through 1 November 2004. The first survey was administered by telephone by the Parker Group from 22 September 2003 through 12 October 2003, and consisted of 1203 interviews with adults drawn from a Random Digit Dialed (RDD) U.S. national sample.73 The next eight surveys were all administered via the Internet by Knowledge Networks: Wave 1 (6–20 February 2004, 891 respondents), Wave 2 (25 February–4 March, 870 respondents), Wave 3 (5– 18 March, 930 respondents), Wave 4 (19 March–2 April, 889 respondents), Wave 5 (2–16 April, 881 respondents), Wave 6 (17–29 April, 899 respondents), Wave 7 (18–28 June, 900 respondents), Wave 8 (21 October–1 November, 1125 respondents). The Knowledge Networks data is equivalent to a national RDD sample.74

These data have three distinct advantages over other casualty-related opinion data.75 First, the data come from surveys designed explicitly to measure casualty tolerance and to probe the determinants of that attitude. As we explain in more detail in chapter 3, casualty tolerance is very difficult to measure, and previous measures, including our own, were not well-suited to the kinds of comparative analyses that make up the heart of this project. Second, the surveys sampled opinion over the span of a year in the midst of a war during which the combat toll more than tripled, and during which the public was exposed to countless news reports and commentaries about the human costs of war. The most sophisticated previous work in this area, including our own, was primarily based on hypothetical scenarios asked during peacetime, and this naturally raised doubts about the robustness of any findings. Third, the last survey wave was conducted just on the eve of the 2004 presidential election and is the only nationally representative casualty-related survey ever conducted in such close proximity to an election. As we will show in chapter 5, our poll is a very good proxy for a survey of the electorate, and we are thus able to probe the electoral implications of casualties in an unprecedented way.

Our goal in conducting this research is to extend our empirical grasp of how Americans structure their understanding of foreign affairs in the area of military conflict. We seek to build upon existing theoretical models and empirical evidence to improve our knowledge of how Americans respond to seeing their fellow citizens die in combat on behalf of their country and how their attitudes toward war influences their voting behavior. We fully understand, however, that ours will not be the last word on this topic. In fact, as we discuss in chapter 7, we believe that our research raises important questions for future work in political psychology regarding information processing and belief formation. We recognize that the answers to these questions of information processing and public “rationality,” cannot be fully answered without further theoretical and empirical work in political psychology.

Nonetheless, we believe that our research sheds important new empirical light on the way that Americans structure their understanding of foreign policy and how they respond to America’s experience of war. Our detailed findings are too varied to enumerate here, but we can provide a snapshot of our overall argument and the main results from the analyses presented in the chapters that follow. Our basic model of casualty tolerance is reflected in figure 1.1.

We argue that support for continuing a military operation (or, for that matter, starting such an operation) in the face of mounting combat casualties is a function of the interactive effect of two underlying attitudes: expectations about the likelihood that the military operation will be a success and belief in the initial rightness of the decision to launch the military operation.76 The more likely you think the operation will be a success and the more correct you think the original decision was, the more you will be willing to pay a higher cost in the form of mounting combat fatalities. The relationship is interactive, meaning that if one believes that the operation is doomed to failure, then also believing that the decision was the “right thing” does not have so big an effect; likewise, if you believe that the decision was wrong, then also believing that the operation will be successful does not have so big an effect. But, following our own earlier research, we argue that expectations of success will be the more important of the two factors.77 That is, varying expectations of success have a more substantial effect on one’s stomach for war than do varying beliefs about the rightness or wrongness of the war; the support of people who believe that the war can still be won but think that the war was wrong is more robust in the face of a mounting human toll than is the support of people who believe that the war is lost but think the war was the right thing to do in the first place.

Of course, other factors contribute to expectations of success and a belief in the rightness or wrongness of the decision to use force. Real-world developments intervene, such as the capture of Saddam Hussein or the failure to find stockpiles of weapons of mass destruction. The public resolve of leaders can reassure publics that success is still likely despite adverse developments on the battlefield. Presidential rhetoric can likewise serve to reinforce public conviction in the rightness of the cause. As Larson and Kull and others have argued, the endorsement of actors not directly implicated in the decision can also reinforce beliefs about the appropriateness of war. For that matter, as we explore in chapter 7, perceptions of the rightness of the decision may itself be shaped by expectations of success, and vice versa. Viewed this way, however, success still trumps other factors; the causal arrow running from perceptions of success to belief in the rightness of the war is stronger than the causal arrow running in the other direction. Moreover, demographic factors like gender, race, age, partisan affiliation, and so on, all have a separate influence on the key attitudes we are studying. Importantly, however, we find that partisan attitudes, which the conventional wisdom tends to treat as dominant, has most of its effect early on in the causal chain; that is, controlling for the influence of partisan affiliation on belief in the rightness of the initial decision to use military force, we find that partisanship has only a very modest effect on casualty tolerance.

When it comes to the electoral decision in wartime, the same basic framework operates. Attitudes about the rightness of the war and the likelihood of success interacted to shape the voter’s choice between President George W. Bush and challenger Senator John Kerry. The relative weights, however, were reversed. Whereas the prospective attitude of expectations of success was dominant in shaping casualty tolerance, it was the retrospective attitude about the rightness of the war that dominated in shaping vote choice. Individuals who believed that the decision for war was wrong but believed that we would win were willing to continue the military mission despite mounting casualties; in the choice for president, however, they were more likely to favor Kerry over Bush. Individuals who believed that the decision for war was right but believe that we would lose are not willing to continue the mission in the face of mounting casualties; they were, however, far more likely to prefer Bush over Kerry.

Throughout we are able to confirm the modest form of the claims made in the existing literature. As Jentleson has argued, individual attitudes about the rightness or wrongness of certain categories of missions in the abstract does shape individual casualty sensitivities in particular missions. As Larson has argued, individuals who perceive elite support for a mission are more likely themselves to support the mission; this, in turn, does shape casualty tolerance. As Kull has argued, an individual’s support for the mission is influenced by whether that individual believes that other states are also supporting it. And, as many people have argued, having a personal connection to the people at risk in a military operation does shape one’s willingness to pay the human costs of war. But our analyses here show that regardless of how the issue is approached, expectations of success matters the most.

Return to Book Description

File created: 11/13/2008

Questions and comments to:
Princeton University Press

New Book E-mails
New In Print
PUP Blog
Princeton APPS
Sample Chapters
Princeton Legacy Library
Exam/Desk Copy
Recent Awards
Princeton Shorts
Freshman Reading
PUP Europe
About Us
Contact Us
PUP Home

Bookmark and Share