Since the end of the Cold War, the American public has shown three different faces with respect to using U.S. military force abroad: full support for the Gulf War, transitory support in Somalia, and no support in Bosnia. An examination of public opinion and media coverage of these cases provides a paradigm of the opportunities and obstacles faced by policymakers in seeking to assess the likely public response to American involvement abroad, as in Haiti. American attitudes toward these three crises suggest that the public will be clearly disposed to act militarily in two situations: if it feels America's vital interests are at stake, and if American military force can provide humanitarian assistance without becoming engaged in a protracted conflict. The peacekeeper role evokes an ambiguous response, but the public strongly rejects the peacemaker role.

The patterns of response suggest that early in a crisis the public will seriously consider the use of force, but that even when it feels the United States has a responsibility to act in its national interest, large percentages (sometimes majorities) will favor no action unless they are swayed by presidential leadership. All of these factors may be missing in the case of Haiti.


The Persian Gulf War enjoyed popular support because it was relatively cost-free and ended well. But even in the earliest stages of the crisis, long before the outcome of the war was certain, public opinion was positive enough to convince policymakers that Americans would support the war effort. Several reasons can be cited. From the start, there was a strong belief that the United States should play a significant role in helping defend the Saudi Arabian oil fields and evicting Iraq from Kuwait. The Bush administration did a very good job of communicating to the public the purpose of the Persian Gulf deployment, including its initial defensive mission and later offensive mission. Multilateral contributions of money and manpower boosted public support for a military option. The congressional debate and vote to authorize the use of force in the Persian Gulf fortified American opinion. Finally, extensive media coverage of the crisis, and later the war, solidified public opinion and kept public attention focused on the crisis.

The amount of television coverage was unprecedented. The Tyndall Report, which tracks television news coverage, found that the amount of Persian Gulf-related news aired on the three U.S. broadcast networks in 1990, before the war began, was nearly four times more than the amount for the top story of 1989, the Tiananmen Square crackdown. And network news coverage was not the key media element in this war. The viewing of live news on CNN played a well-recognized, crucial role in the public's continuous connection to the Persian Gulf from August 1990 onward; CNN was overwhelmingly the medium of choice in this respect. Probably for this reason, public attitudes toward the Gulf deployment differed substantially from reactions to previous major U.S. military involvements. More of the American public had an opinion about the decision to commit U.S. forces to Saudi Arabia (92 percent) than about the decisions on Vietnam two decades ago (78 percent) or South Korea in 1950 (85 percent).

First reactions to American troops in the Persian Gulf, Panama, and Somalia were also more positive than first reactions to troops in Vietnam and Korea and all, as we will discuss, were considerably more positive than the reaction to the dispatch of troops in Haiti. It may be that the public, watching events unfold on television, feels it is taking greater part in go-to-war decisions than 40 years ago. Again, however, Haiti stands out as an exception. Table 1 summarizes this data.

The public seemed persuaded from the start on moral and practical grounds that force was justifiable in the Persian Gulf. In a New York Times/CBS poll in early August 1990, more Americans saw a parallel between stopping Iraqi dictator Saddam Hussein and stopping Hitler (61 percent) than between the Persian Gulf and Vietnam (42 percent).1 In the same survey, two out of three Americans agreed that the United States should have a say in who controls Mideast oil. But while first reactions to Persian Gulf troop deployment (portrayed as a move to defend Saudi Arabia and oil fields) were very positive, the conditionality of that public approval is striking. First reactions to the possibility of offensive action against Iraq were quite negative. The public opposed bombing Iraqi military targets to force Saddam Hussein out of Kuwait by almost two to one: 59 percent to 27 percent (WP).

The public remained ambivalent about going to war right up until Bush ordered air strikes. At the end of the prewar period, it was still closely divided on whether economic sanctions should be given more time to work. While large majorities throughout the prewar period agreed with the general proposition that the United States should do what was required to drive the Iraqis out of Kuwait, including using force, when pollsters asked more pointed, "hard" questions about military operations, they got more dovish answers. In mid-November 1990, 53 percent said that, based on all they had heard or read, the United States should not go to war against Iraq over the Mideast (LAT). Responses to "hard" questions changed with the U.N. Security Council vote in November that set the January 15, 1991 deadline for Iraq to withdraw from Kuwait. Before that vote, majorities opposed going to war with Iraq; afterward, majorities favored it every time the question was asked (Gallup).

The contributions of men and money from allied governments added support to this sentiment, but more important was the congressional debate and subsequent vote in early January 1991 for war. All major polls found majorities, some as high as 82 percent (Newsweek), supporting Bush's intention to seek congressional approval before using force. After the vote, support for immediate military action increased sharply (WP).

A final factor was the Bush administration's communication skills in justifying the war. In early August 1990, the public was divided over whether Bush had "explained clearly what's at stake and why he is sending troops to Saudi Arabia": 50 percent yes, 41 percent no (NYT). One month later, 77 percent said that they had a clear idea on the matter (WP). At no point during the military buildup and debate about military action did approval ratings for Bush fall below 55 percent on his handling of the crisis.


The first post-Gulf War overseas commitment of U.S. forces began with an American public more positively disposed toward and confident in its military than at any time since the Vietnam era. The initial purpose of the Somalia mission, providing humanitarian aid, was one the public easily embraced, and extensive, supportive media coverage attended the arrival of American forces there. Immediately after the deployment was ordered, 66 percent of the public surveyed in December 1992 approved sending "American military forces to assure distribution of relief supplies" (Newsweek). The next month, after the U.S. Marines landed, their mission was approved overwhelmingly by 84 percent, of whom 53 percent "strongly approved" (lat).

Nonetheless, warning signs were apparent. The public overwhelmingly understood the mission to be humanitarian, not fighting: 59 percent in December 1992 said the main objective was to "deliver relief supplies"; only 31 percent said it was to "also attempt to bring a permanent end to the fighting" (Gallup). The public expected the mission to be short: 62 percent believed the United States would not get bogged down in Somalia, but would get out quickly (WP), 51 percent believed it would be over within six months, and 69 percent said it would end within a year (lat).

Even after the first clashes with Somali warlord Mohamed Farah Aidid and initial U.S. casualties, the American public remained behind the effort. At the end of March, 84 percent of Americans approved (56 percent "approved strongly") of the use of U.S. forces for humanitarian purposes (ATI). Even in June, after the mission had broadened dramatically, with a full-scale attack on Aidid camps in retaliation for the deaths of 23 Pakistani peacekeepers, 65 percent of Americans approved of U.S. participation in U.N. efforts against a warlord (Gallup).

But the development of public support for the new mission was inhibited by American television screens going virtually blank on Somalia for half a year after the Marines landed. Public interest plummeted from 52 percent who said they followed the issue "very closely" in January to 16 percent in June (TMC, Gallup). In the fall, when the networks and public attention returned to the story, the nature and danger of the mission were very different, and support withered rapidly.

In September, as the U.S. death toll mounted and violence increased, the public divided evenly (46 percent disapproved vs. 43 percent approved) over "the presence of U.S. troops in Somalia" (Time). The same month a majority said U.S. troops should leave (53 percent) rather than stay in a limited capacity to preserve peace (46 percent). A larger majority, 57 percent, said military combat against warlords should stop (Gallup). As many as 69 percent said U.S. troops should be responsible only for food delivery (Time). The next month brought a battle between U.S. Army Rangers and Aidid gunmen that ended in tragedy and precipitated the firestorm of public opposition to the Somalian operation. Eighteen soldiers were killed; television news showed some of their bodies violated by crowds and a battered helicopter pilot who had been captured and interrogated. In October 1993 the proportion of Americans who followed Somalian events "very closely" jumped to 34 percent (TMC) as the amount of television coverage leaped out of the doldrums (although the coverage reached only half the level during the Marines' landing). Soon thereafter, 69 percent of people surveyed called for withdrawal of American troops, 43 percent right away (Gallup). Public confidence that U.S. forces would not get bogged down plummeted to 44 percent (WP) from 62 percent a year earlier. Approval of U.S. involvement in Somalia was halved to 34 percent (NBC) compared to 74 percent a year earlier.

Part of the reversal can be blamed on lack of sustained communication to the public by the Bush and Clinton administrations, especially as the mission in Somalia changed from humanitarian aid to peacekeeping between factions and then to peacemaking and "nation-building." The first significant raid on Aidid's camp, which occurred January 7, 1993, was portrayed by Bush aides as a shift of tactics rather than strategy. The Clinton administration, sworn into office days later, embraced the approach and made subsequent changes in the mission without trying seriously to justify them. During an impromptu news conference three days after a huge raid on Aidid, Clinton referred to the peacemaking effort only in passing: "We cannot have a situation where one of these warlords, while everybody else is cooperating, decides that he can go out and slaughter 20 peacekeepers."

Somalia also illustrates the capacity of the media to be almost embarrassingly enthusiastic for a mission, then turn abruptly against it when pictures of the human cost begin to appear in American homes. However, the American public's impulse to use force for humanitarian reasons survived the embarrassment in Somalia. In October 1993, after the Ranger tragedy, 56 percent of the public approved "sending U.S. military forces to Asian or African countries in order to prevent famines and mass starvation" (TMC). And in December 62 percent said the United States did the "right thing" to send troops to Somalia to "make sure shipments of food got through" (NYT). Even when asked if the intervention was worth the "loss of American life, the financial costs, and other risks involved," a plurality of 48 percent said yes.

Somalia nonetheless underscored American aversion to the use of force for peacekeeping, and particularly for nation-building and peacemaking. Asked whether the United States should keep troops in Somalia "until the situation is peaceful," two out of three said the troops should "get out as quickly as possible" (TMC). Ironically, the same survey that found majority support for the humanitarian mission in Somalia at its end also found 60 percent opposed to using U.S. forces to restore law and order if governments break down in Africa or Asia and 53 percent opposed if that occurs in Latin America or the Caribbean (TMC).


Yugoslavia began to break up into ethnic republics in 1991 with the independence of Slovenia and Croatia. But the media spotlight turned on the civil war only when multiethnic Bosnia separated in April 1992. In the first polls, taken as U.S. airdrops of food began, 67 percent of the public said the United States does not have "a responsibility to do something about the fighting," while 24 percent said it did (NYT). The proportion feeling some responsibility has risen only slightly since, judging from a mid-1994 survey that found 31 percent agreeing that U.S. "vital interests are at stake in Bosnia" (WP). The media failed to engage the public's attention in Bosnia, let alone sway sentiment toward intervention. At the outset of the war, fewer than one in ten Americans said they were paying "very close" attention to news of that conflict (TMC), and indifference remained high throughout the spring of 1992, despite reports of ethnic cleansing, rape campaigns, and other atrocities that recalled the genocidal horrors of World War II.

Moreover, the level of public interest in Bosnia over the course of the conflict has shown no correlation to the amount of media coverage. Only once did the proportion of the public that followed news of the Balkan conflict "very closely" climb above 20 percent: in May 1993, when U.S. military action appeared likely. But mostly it hovered near 15 percent throughout 1993, dipping to 12 percent in 1994 (TMC).

While most Americans were indifferent to the conflict, those who took sides favored the Bosnians. In early 1993, 30 percent said they were more sympathetic to the Muslim side, compared to only 8 percent for the Serbs (TMC); that changed little throughout that year. To test for possible religious prejudice that might explain the low absolute level of sympathy for the Bosnians, the Times Mirror Center for the People & the Press posed the question in two variants: Half of the respondents were asked if he or she had more sympathy toward Serbs or "Bosnians" and the other half were asked about sympathy toward Serbs or "Bosnian Muslims." No significant difference was found: 35 percent were supportive of "Bosnians," 33 percent were supportive of "Bosnian Muslims."

Throughout much of 1993, American air strikes against Serbian military forces were opposed by majorities who said the United States should not get militarily involved: 62 percent in April, 56 percent in May (Gallup). With Serbian attacks and intransigence toward peace efforts rising, however, support for U.S. air strikes against Serbs, in conjunction with European allies, drew 51 percent and 60 percent support in August 1993 polls (Gallup, WP).

U.S. air strikes, which some have termed the "American way of war," were clearly favored over committing U.S. ground troops to Bosnia, except if they were part of a peacekeeping force. In early 1993, the public by a 55 percent to 32 percent margin opposed the use of military force to help end the fighting there (TMC). But support for deploying U.S. armed forces as part of a U.N. operation to deliver relief supplies was 57 percent in a December 1992 poll (Gallup), particularly if they were to be part of a peacekeeping force: 68 percent in May 1993 (Gallup) and 73 percent in April 1994 (UM).

When the risk of U.S. forces being subjected to sniper re was introduced into the question, however, 59 percent in September 1993 opposed their deployment even as peacekeepers (NBC). In 1994, perhaps spurred by the slaughter of 68 civilians in an outdoor market in Sarajevo, there was a steady rise in the view that the United States did have a responsibility to do "something." But despite many hours of media coverage, as well as much editorializing commentary in the print media, the prevailing attitude remained that this was not an American fight. Table 2, using New York Times data, shows the evolution of this sense of responsibility.


In the post-Cold War world there is little consensus about what constitutes America's vital interests. In a 1993 survey, the Times Mirror Center for the People & the Press found that protecting American oil supplies and preventing nuclear proliferation were the only foreign policy goals shared by both the general public and groups of opinion leaders in nine professional fields, including business, religion, and science.

Polls have found substantial support for using force to prevent a buildup of nuclear weapons in North Korea. Surveys in late 1993 found majority support for U.S. intervention to preempt a North Korean nuclear threat but majority opposition to defending South Korea from North Korean aggression (TMC). When asked in September 1993 about circumstances that might justify using U.S. military force in five situations, public approval ranged from a high of 53 percent "if Iraq invaded Saudi Arabia" to a low of 21 percent "if Russia invaded Ukraine," with 41 percent supporting it if Mexico was "threatened by revolution or civil war" (TMC).

Other polls have approached the issue in more generic terms. A 1993 Roper survey found that virtually all Americans (94 percent) would use force to repel an attack on the United States, and large majorities would use force to strike back when Americans in a foreign country are attacked (79 percent) or close allies are attacked (74 percent). There was surprisingly strong support for interventions related to high-priority domestic issues: 82 percent to stop illegal drugs from abroad, 70 percent to police illegal immigration. At the bottom of this scale, less than 50 percent supported using force to stop genocide or protect innocent lives in a civil war.

The Times Mirror Center for the People & the Press in May 1994 surveyed public attitudes toward the use of force, depending on the justification for intervention (national interest, humanitarian cause, police role) and the kind of force to be used (air or ground). The survey also asked about use of air power "to eliminate an arsenal of nuclear weapons in an unfriendly country such as North Korea or Iran." The poll results underscored the patterns observed in the three cases considered above. It found majority support for the use of force to insure the U.S. oil supply, thwart nuclear proliferation, and provide humanitarian aid. However, there was little support for military intervention to restore law and order. In every case, support for using air power was greater than for ground forces by an average of ten percentage points.

Combining the responses regarding air power and ground troops, slim majorities of Americans supported using force to "make sure U.S. oil supplies are not cut off" (50 percent versus 40 percent), and to "prevent famines and mass starvation" in Africa, Asia, and Latin American countries (52 percent versus 41 percent). But a similar majority opposed the use of force "to restore law and order if the governments break down" in Africa, Asia, and Latin America (51 percent versus 41 percent).

Generally, conservative white males supported the so-called national security goals of protecting oil or eliminating nuclear weapons, while Democrats, women, and African-Americans showed greater support for using U.S. force for humanitarian and law-and-order causes. Most striking was the difference in the views of blacks and whites on justifiable uses of force. Specifically, the majority of Americans willing to use air and ground forces for protecting oil supplies were more often men than women, young rather than old, white rather than black (52 percent versus 44 percent), Republicans rather than Democrats or Independents, and Bush voters rather than Clinton or Perot voters in 1992. Politically vocal respondents - those who regularly listen to talk radio and/or communicated with their congressmen within the past year - were more supportive of this mission than others.

Among the majority of Americans supporting use of force to deliver famine relief, men and women were in the same proportion overall but with more younger women than younger men, more blacks than whites by a huge margin (69 percent versus 49 percent), more Democrats than Republicans or Independents, and more Clinton voters than Bush or Perot voters.

No gender gap was found in the minority that supported using force for law and order, but the less educated and poorer respondents approved significantly more than their socioeconomic opposites. Blacks were more in favor of using force for this cause than whites (59 percent versus 38 percent). Democrats were evenly split, but both Republicans and Independents were opposed by large margins. Clinton voters were more supportive than Bush or Perot voters.

Based on this data, we can construct a simple, four-part typology:

Interventionists, 31 percent of the public, who would use force to protect oil as well as for humanitarian purposes.

Noninterventionists, 29 percent, against both missions.

U.S.-centrics, 19 percent, who would use force to protect oil sup-plies but not to give humanitarian aid.

One-worlders, 21 percent, who would use force for the humanitarian but not the oil mission.


If the public's basic opinions about America's responsibility to act drive its response to the potential use of force, we believe that media coverage modulates rather than dictates public opinion. Some have given the press more credit or blame.

Many have faulted the so-called "CNN effect" - televised images of combat victims - for the refusal of the American public to tolerate casualties in foreign wars, for example. Images of body bags containing American soldiers shown in family living rooms obviously affect the public's attitude toward use of force. But this is not unique to Americans watching uncensored television. As Edward Luttwak points out, the Soviet Union never let its population see U.S.-style images of its wars, yet "the reaction of Soviet society to the casualties of the Afghan war was essentially identical to the American reaction to the Vietnam War." More often than not, what is blamed on the media is really the public acting on its deep-seated judgments about the basic wisdom of an intervention.

Neither Somalia nor the Persian Gulf War tells us whether the public would tolerate televised casualties for a sustained period in a conflict in which Americans clearly believe their interests were at stake. The media probably has the most negative effect on military operations abroad in the absence of coherent policy and firm leadership that persuasively reiterates the purpose of the use of force throughout the mission. The problem becomes crucial and more difficult in an era in which satellite communications give Americans direct access to news of momentous international events. Political scientist Maxine Isaacs argues that in such an environment, opinion leaders (perhaps even the president) play a less important role in shaping views about foreign affairs because the public can form its own judgments based on observations via the press. Examining the relationship between elite and mass opinion in response to the Tiananmen Square uprising and massacre in mid-1989 and the failed August 1991 coup against Soviet leader Mikhail Gorbachev, she concludes:

The dependence of mass upon elite opinion (expressed either by policy/opinion leaders or by the media), so long presumed by so many authorities to exist, is not total in two important late twentieth century cases . . . Congruence of elite and mass opinion is possible, but it is not foreordained. Indeed, there is ample evidence that public opinion now operates somewhat, if not entirely, separately from elite opinion.

Isaacs also found no support for the view that the president, as the most elite of all elite opinion makers, exerts extraordinary influence on the public in foreign policy:

Sometimes the public supported the President, sometimes it did not. Sometimes the public aligned with the President's critics, and sometimes the public disagreed with both the President and his critics . . . It also appeared that the President did not exert a strong influence on matters of policy with the policy and press elites who did not work for him.

She suggests that a new model is needed to explain how elite and mass opinion on foreign policy are related in the age of mass communications. In part, it would have to deal with "the phenomenon that some public events capture the public's attention while others, sometimes despite massive news coverage such as in the case of Bosnia, fail to engage the public interest." Isaacs, however, dealt with cases that did not involve the use of American forces. We believe that President Bush, his aides, and the Congress were highly influential in shaping public opinion during the Persian Gulf conflict. President Clinton and his aides were significantly less so in Somalia. In Bosnia, neither Bush nor Clinton attempted to persuade the public that the Balkan conflict was worth the risk of American lives, and the public, preoccupied in the post-communist world with domestic issues, did not clamor for war in the Balkans.

The media, which affects mass opinion through the amount of coverage devoted to an issue, if nothing else, increased support for the Persian Gulf War and, initially, for the Somalia intervention. It was also critical in causing the U.S. withdrawal when, in the absence of a coherent and articulated national policy, it presented American casualties to a public that had no underlying sympathy for the peacemaking mission. In Bosnia, the media clearly had no effect.


Following the military ouster of Jean-Bertrand Aristide in 1991 and a new surge of Haitian refugees toward U.S. shores, the American public in late 1993 approved the oil embargo of Haiti by 63 percent (WP), and approved U.S. participation in a naval blockade to enforce U.N. sanctions by 46 percent to 37 percent (NYT). But its attitude toward sending troops depended on the mission. Seventy-seven percent favored using force if the purpose was to evacuate U.S. citizens (NYT), and 69 percent approved force to prevent illegal immigration (WP). But the public strongly and consistently opposed U.S. force to restore Aristide and democracy to Haiti: 69 percent against in October 1993 (WP), 68 percent against in May 1994 (Time), and two-thirds opposed in June, July, and early September (NYT, Newsweek). Immediately before the intervention, 60 percent said U.S. troops should stay out; immediately afterwards, 52 percent said the same, a relatively small eight-percentage-point drop, even after the forces were committed (NYT).

The public was also dubious that an intervention would be quick and successful. In July 1994 less than half believed it would be as effective as the Persian Gulf War (47 percent), while 22 percent thought it would be like Somalia, and 17 percent said it would be long and costly like Vietnam (Time). Similar results were found in a September pre-intervention poll (Time).

Public attitudes immediately after the intervention were remarkably similar to initial views about the famine relief effort in Somalia, according to a CNN poll: 61 percent were confident that U.S. casualties in Haiti would be few (64 percent said the same regarding Somalia in December 1992), 50 percent were confident the United States would withdraw in a few months (52 percent were with respect to Somalia), and 52 percent were confident the effort to restore democracy in Haiti would succeed (49 percent believed that about the famine relief effort in Somalia).

In sum, the pattern of post-Cold War approval of U.S. interventions did not extend to Haiti (Table 1). The public could sour quickly on Haiti for several reasons. A majority opposed intervention before the troops landed, and even afterwards. There was significantly less rally-round-the-president and rally-round-the-decision reaction than usually occurs once forces are committed. Clinton did not get Congress' approval, although as much as 78 percent of the public wanted him to do so a week before the invasion (WP). He did not convince the public that U.S. interests are at stake. Asked a week before the operation for the main reason that the United States might take part in an invasion, 41 percent said because human rights abuses were severe, 13 percent said because U.S. interests were threatened, and 32 percent said because Clinton "will have political problems if he doesn't carry out his threat to invade" (Newsweek).

On the positive side, Clinton put together a multilateral operation, and he went the extra mile for a diplomatic solution that avoided a forcible entry of U.S. troops. But, in the end, public attitudes will probably hinge on whether the promises of short duration and low cost (in lives if not money) will be realized. The public may demand a "zero-casualty" mission, as it did for the humanitarian mission to continue in Somalia, but the Haitian peacemaking mission begins with far less support among the American electorate.

The New York Times/CBS poll, hereafter identified as NYT. For brevity, most polls in this article will be identified by the initials of the polling organizations. In partnerships, only one organization is cited. Other abbreviations are: WP for The Washington Post/ABC, TMC for the Times Mirror Center for the People & the Press, LAT for Los Angeles Times, NBC for NBC/The Wall Street Journal, Time for Tim/CNN, ATI for Americans Talk Issues, and UM for the University of Maryland.


Table 1. positive negative uncertain


August 1950 65 20 15


January 1965 50 28 22

May 1965 52 26 22

November 1965 64 21 15


November 1983 63 29 8


January 1990 72 18 10


August 1990 75 17 8

January 1991 77 15 8


December 1992 74 21 5


September 1994 41 52 7

Positive response refers to "should be involved," "not a mistake," "a good idea," and "right decision;" negative response refers to the opposite. Polls were conducted by Gallup, except for Panama and Somalia, where the surveys were done by NBC/The Wall Street Journal; the January 1991 poll on Iraq, by the Times Mirror Center for the People & the Press; and the Haiti poll, by The New York Times.



April 1994 41 49 10

February 1994 36 53 11

December 1993 26 65 9

June 1993 37 51 12

May 1993 37 52 11

January 1993 24 67 9

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • Andrew Kohut is Director of the Times Mirror Center for the People & the Press, a public service arm of the Times Mirror Co. His is also former President of the Gallup Organization and founder of Princeton Survey Research. Robert C. Toth, a former foreign and national security correspondent of the Los Angeles Times, is a Senior Associate at the center. Carol J. Bowman, Director of Research at the center, assembled the survey data for this article.
  • More By Andrew Kohut
  • More By Robert C. Toth