Informed Public Opinion About Foreign Policy: The uses of deliberative polling

Henry E. Brady,
Henry E. Brady Dean and Class of 1941 Monroe Deutsch Professor of Political Science and Public Policy - University of California, Berkeley
James S. Fishkin, and Robert C. Luskin

June 1, 2003

Only one-sixth of the respondents to a January 12, 2003, poll for the Knight-Ridder newspapers knew that none of the September 11 terrorist hijackers was an Iraqi citizen. One-third said they did not know whether the hijackers were Iraqis, and almost half (44 percent) said that some or most of them were from Iraq. Almost two-thirds thought that “Iraq and Al Qaeda—Osama bin Laden’s organization—are allied and working together to plan new acts of terrorism.” And two-thirds said that the United States should “take military action to disarm Iraq and ensure that it cannot threaten other countries with nuclear, chemical, or biological weapons.”

It is hard to know exactly how far these beliefs were causally linked—to what extent the plurality belief that some or most of the hijackers were from Iraq led to the majority belief that Iraq was allied with Al Qaeda and from that in turn to the majority belief that the United States should take military action to disarm Iraq. But the combination of support for military action with such striking ignorance and misinformation is troubling—the more so, considering that two-thirds of the same Knight-Ridder respondents thought they had “a good understanding of the arguments for and against going to war with Iraq.”

Strong feelings undergirded by scanty knowledge are hardly new to American public opinion about foreign policy. In a book and article of the mid-1950s, Harold Isaacs found among Americans a new awareness and sense of “worried urgency” about Asia, accompanied by images of the continent that were little more than “scratches on our minds.” Much the same now could probably be said about American public opinion toward the Arab world. When Isaacs wrote, the United States was already waging war in Korea and would later do so in Vietnam. Now we have just finished waging war in Iraq. Whatever one’s position on these wars, we think it worrisome that public attitudes toward them, whether supportive or critical, typically rest on such limited information. Is there some way of doing better—of glimpsing better informed and more considered public opinion concerning America’s role in the world?

One approach is Deliberative Polling, which surveys a random sample of Americans both before and after they have had a chance to deliberate about certain issues. The sample is first interviewed, then invited to gather for a weekend. They are sent carefully balanced (and publicly available) briefing materials. On-site, they discuss the issues in randomly assigned small groups led by trained moderators and then pose questions developed in the small groups to panels of experts and political leaders. Parts of the discussions are broadcast on television, either live or in taped and edited form. At the end, the participants are asked again the same questions posed initially. The resulting distribution of opinion represents the conclusions the public would reach if people knew and thought more about the issues.

This proposal is not just theoretical. On January 10-12, 2003, we collaborated with MacNeil-Lehrer Productions, the National Issues Forums, and others to conduct a Deliberative Poll on foreign policy issues as part of a project called By the People: America in the World. The participants in the 2003 poll in Philadelphia were a scientifically chosen random sample of 343 American citizens, aged 18 and older. They matched the country’s adult citizenry in age, education, sex, region, ethnicity, race, occupation, employment status, military service, religious belief, religious attendance, and union membership. In microcosm, the sample was America. What did they conclude?

The Deliberative Poll on Iraq and International Security

Before coming to Philadelphia, the participants completed a 40-minute telephone interview about their views on fighting terrorism, controlling weapons of mass destruction, protecting human rights, promoting democracy, providing foreign aid, solving global environmental problems, promoting international trade, and other topics. At the end of their two days of deliberations, they completed the same questionnaire with some additional questions.

By the end of the weekend, the participants were more likely to see Iraq as a threat but also far more insistent that the threat be dealt with only within the framework of the United Nations Security Council and on the basis of evidence from UN weapons inspectors. Eighty-seven percent ranked Iraq as a threat, as against 74 percent of the control group (a separate, contemporaneous survey of citizens who did not deliberate), and only 46 percent, compared with 57 percent of the control group, wanted to shift America’s focus from Iraq to terrorism.

More participants (77 percent) than control group members (73 percent) agreed that the United States should invade Iraq if the UN inspectors found evidence that convinced both the United States and the Security Council that Iraq had weapons of mass destruction, but only 37 percent of the participants, as against 46 percent of the control group, thought that the United States should do so if the UN was not convinced—if evidence was found “that convinces the US but not the UN Security Council that Iraq has weapons of mass destruction.” And only 22 percent of the deliberators, compared with 31 percent of the control group, thought that the United States should invade Iraq “if there is no new evidence found by the inspectors but the US still has reasons to believe that Iraq has weapons of mass destruction.”

More generally, the participants grew warier of go-it-alone approaches, and not just with respect to Iraq. Support for acting alone to stop the spread of weapons of mass destruction fell from 58 percent to 44 percent, for acting alone to stop terrorism from 67 percent to 52 percent. Support for working with the UN to stop the spread of weapons of mass destruction rose from 85 percent to 92 percent, for working with the UN to stop mass killings from 82 percent to 93 percent.

The participants also changed their views regarding what the United States should do following a successful invasion of Iraq. They emerged more interested in rebuilding Iraq’s economy (96 percent versus 79 percent), trying to ensure Iraq’s friendliness to U.S. interests (86 percent versus 76 percent), and working with international organizations such as the United Nations in rebuilding the country (82 percent versus 61 percent).

Other Results on Global Affairs

Beyond Iraq and international security, the participants emerged more willing to take some responsibility for solving the world’s problems—ranging from military security to economic privation, the promotion of democracy, and the environment. The share disagreeing that “this country would be better off if we just stayed home and did not concern ourselves with problems in other areas of the world” increased from 63 percent to 86 percent, while the share agreeing that “it is the job of the US as a global leader to protect the world from countries that are likely to use weapons of mass destruction” rose from 55 percent to 69 percent.

Similarly, the shares of participants preferring to deal with problems at home before dealing with world hunger or the global AIDS epidemic fell dramatically, from 53 percent to 29 percent in the first case and from 50 percent to 20 percent in the second. The share not supporting the goal of promoting democracy abroad declined from 38 percent to 23 percent. On environmental issues, the share wanting to solve environmental problems through international agreements increased from 70 percent to 87 percent, and the share willing to require higher gas mileage standards for vehicles increased from 65 percent to 81 percent.

Increases in Knowledge

These changes of opinion came about amid a great deal of observable learning. Before deliberation, many people thought the United States spends much more on foreign aid than it does. The share knowing roughly what proportion of the U.S. budget goes to foreign aid (less than 1 percent) more than tripled, from 19 percent to 64 percent. Participants showed the largest knowledge gain in this policy area, but in every area they registered some increase in knowledge, and in several the increase was substantial. The shares knowing that President Bush opposes recent international agreements to control greenhouse gases, that the United States does not have a veto in the World Trade Organization, and that more than 30 of every 100 adults in the African countries with the highest rates of infection have AIDS or the AIDS virus increased from 37 percent to 53 percent, from 40 percent to 52 percent, and from 54 percent to 68 percent, respectively. Across the nine knowledge items in the survey, the share answering correctly increased an average of 12 percentage points.

Some sense of the effect of this learning may be gleaned from the change in opinion about foreign aid. Once the participants realized just how little is being spent on foreign aid, support for increasing foreign aid leapt from 20 percent to 53 percent, a majority. More generally, the participants’ knowledge and their opinions clearly changed dramatically because of their involvement in the Deliberative Poll. Some of this change probably preceded the deliberative weekend because the prospect of participating had already heightened participants’ interest and attention to foreign affairs. Our data show that they watched the news on TV, read newspapers, and discussed public affairs more often, paid more attention to TV news and newspaper articles about foreign affairs, and grew more interested in both foreign affairs and politics in the United States. Other changes undoubtedly came during the weekend itself, as they discussed foreign policy with their peers, asked questions of experts, and reflected on America’s place in the world.


The National Issues Convention ended just as the Knight-Ridder poll was published, on January 12, but the Deliberative Polling results drew far less media attention. Why so little coverage? Shouldn’t informed public opinion count for more than uninformed public opinion?

One response might be that a single Deliberative Poll cannot make the entire American population more knowledgeable about foreign policy. A Deliberative Poll involves only a few hundred Americans who are given a chance to learn, think, and deliberate. Why should the media take seriously the views of such a relatively small group? The answer is that they are a scientifically chosen random sample and their views therefore represent what the American people would think if they became similarly more knowledgeable about foreign policy. As noted, those who came to Philadelphia learned a great deal about foreign policy. They were clearly better informed after deliberating.

Another response might be that the outcome of the Deliberative Poll depended partly on the briefing materials and the experts to whom the participants were exposed. This is no doubt a central issue. Indeed, it is a fundamental issue in democracy—how can expertise be combined with democratic choice? We will not claim to have the final answer, but a well-designed Deliberative Poll does provide a sensible approach. The briefing materials for Philadelphia were prepared in consultation with experts taking many different perspectives, including former members of Democratic and Republican administrations and prominent academics. The experts who appeared in Philadelphia offered similarly varied perspectives. They included Lawrence Korb of the Council on Foreign Relations (a former assistant secretary of defense in the Reagan administration), Peter Brookes of the Heritage Foundation, William Niskanen of the Cato Institute, Anne-Marie Slaughter of Princeton University’s Woodrow Wilson School, Princeton Lyman of the Aspen Institute, Kevin Martin of Peace Action, Dean Baker of the Center for Economic and Policy Research. For the final session, the Bush administration chose Richard Haass, director of policy planning at the State Department, as its own spokesperson. As his counterpart, By the People chose Zbigniew Brzezinski, Jimmy Carter’s national security adviser. A full range of opinions was represented. All the briefing materials and all the deliberative sessions were open and available to the media.

These concerns call to mind those voiced by critics of scientific opinion polling during its infancy in the 1930s and 1940s. Some concerns led to such methodological improvements as the move from quota sampling to random sampling. Other concerns were unfounded. Deliberative Polling is similarly in its infancy and will probably require some modifications as we learn more about it. But the value of going beyond policy attitudes based only on “scratches on our minds” is great enough to justify investing in this method and taking it seriously. By giving substantial coverage to Deliberative Polls, the media could stimulate a broader debate about what information and knowledge people need to make informed pronouncements about foreign policy.

Certainly a better informed public is worth listening to. On the PBS broadcast from the National Issues Convention, Brzezinski was asked, toward the end of an extended dialogue with the members of the Deliberative Poll, “whether foreign policy has become so specialized and so complex that it is impossible for average citizens to have informed opinions about foreign policy.” “I probably would have said yes,” he replied, “but my experience this morning makes me think otherwise. I was told—I hope that’s true—that you are really a cross-section of the American public. If you are, you’re damn good.”