John Zogby's Creative Polls

In a recent New York Times Magazine cover story about animal rights, journalist Michael Pollan reported that 51 percent of Americans believe that "primates are entitled to the same rights as human children." It was a surprising finding, but one that Pollan simply attributed to a "recent Zogby Poll." When Pollan's article came out, you can only imagine the celebration at the Doris Day Animal League, a group dedicated to establishing legal rights for chimpanzees. The league's role in commissioning the survey went entirely unmentioned in the Times story. By hiring the renowned pollster John Zogby, the group had essentially purchased an objective fact, one that entered into the conventional wisdom via the nation's leading Sunday magazine.

Whomever you blame for this small propaganda coup, it's hardly unique. Media coverage of polling results often neglects to mention the self-interestedness of the sponsor, and John Zogby is a leading enabler. Today, Zogby International's polling reputation may be second only to that of the hallowed Gallup Organization, which makes having a Zogby Poll extremely desirable for advocacy groups across the political spectrum. Animal rights is a lefty cause, but one recent Zogby Poll conducted for the libertarian Cato Institute found that "two-thirds of likely voters support personal Social Security accounts" -- i.e., partial privatization. Another, conducted in 1997 for the anti-tort group New Yorkers for Civil Justice Reform, found that Empire State citizens "overwhelmingly believe that the cost of lawsuit awards is too high." And a International Poll, conducted for the right-wing Newsmax Web site, found in late 1999 that two-thirds of Americans wanted Congress to consider a second impeachment proceeding against then-President Clinton. It helped that the poll primed respondents with speculative allegations that the president traded nuclear technology to the Chinese in exchange for campaign cash.

What these polls have in common is that they reveal "findings" that their sponsors wish the public to believe as facts. And Zogby's standing as a reputable pollster buys instant credibility.

There's nothing new about dubious surveys: An infamous Roper Poll released in 1992 came to the wild conclusion that 3.7 million Americans had likely been abducted by aliens. And Zogby International isn't the only firm available for advocacy groups, candidates and corporations in need of creatively framed findings and message testing. But among high-profile pollsters, Zogby is unusual in the extent to which he has blended partisan and interest-group polling with credibility-enhancing contracts for media outlets such as Reuters, NBC News, MSNBC, and numerous newspapers and television stations.

As Zogby himself acknowledges, the repute he derives from media polling helps him sell his services to more self-interested clients. The lucky groups end up with the Zogby brand name attached to findings that advance their agendas. "Media organizations should have people who absolutely aren't polling for interest groups," observes Robert Blendon, who directs Harvard University's Program on Public Opinion and Health and Social Policy. Blendon notes that most major media polling conglomerates, such as the ABC News/Washington Post Poll, maintain firewalls between their work and outside interests.

Frequent Zogby collaborator John K. White of The Catholic University of America believes the pollster does his best to divine what the public really thinks, but, as Zogby himself concedes, the ultimate decision on whether to make public a particular poll rests with his clients. By contrast, Andrew Kohut of the Pew Research Center for People and the Press notes that when he was president of the Gallup Organization, clients who sought surveys for public-relations purposes had to release the results no matter what they showed.

Gallup likewise doesn't work for political candidates. In the last election, however, John Zogby brazenly polled for a Democratic opponent to Rep. Tom DeLay (R-Texas), paying out of his own pocket because he wanted to provide a "fresh challenge" to the Republican House whip. Zogby also polled for New York's millionaire independent gubernatorial candidate Tom Golisano, declaring in late October, "I'm ready to mortgage my house and predict that Golisano comes in at least second, barring anything unforeseen." Golisano came in third with 14 percent of the vote.

In the past half-decade, meanwhile, numerous Zogby Polls for various special interests have relied on creative phrasing to give the impression of wide public support for the view that the given client is promoting. In response to a question about the wording of the impeachment survey, Zogby responded, "If we had anything to do with the wording of that question, then I guess I have a problem with it." He telephoned back to add that it was "probably not the best wording, but, I mean, I think it's defensible." Zogby acknowledges that he retains control over question phrasing. Indeed, in the world of interest-group polling, clients often submit proposed questions or concepts, but much of what they are buying is the polling firm's expertise in devising wording that produces results.

Zogby protests that he can't control the misuse of sound survey data by interest groups and incautious journalists. And, in fairness, Zogby is just one link in a chain of misinformation. Any criticism of him is also, inevitably, a criticism of major media organizations whose skeptical faculties, when it comes to polling, are suspended.

The Gullible Media

Indeed, key to Zogby's success is a credulous media, particularly cable news. In the unregulated polling industry, journalists are, by default, the chief arbiters of quality. For years, Zogby has been regularly exalted as "the nation's most accurate pollster," in the words of FOX's Bill O'Reilly -- a distinction Zogby owes to his pinpoint prediction of the 1996 presidential outcome. It doesn't hurt that Zogby is a bright and charming television personality in a polling profession that has its share of geeks.

Because Zogby works for both left and right, it's often assumed that he serves the causes of truth and objectivity. Unlike partisan pollsters, who are known for giving their own parties some padding in surveys, Zogby is generally invited on the air without anyone from the "other side" for balance. "I can't think of any pollster other than Zogby who regularly works for people on both sides and is touted by people on both sides," notes University of Virginia political analyst Larry Sabato. "That's quite an accomplishment. Whether it's good or bad is another question."

In the summer of 2001, journalist Cynthia Cooper alleged on Women's eNews that Zogby had conducted a poll for an "unidentified conservative client" that reached the questionable conclusion that a majority of Americans would support legislation requiring welfare recipients to use birth control in order to be eligible for benefits. Cooper also noted that Zogby's refusal to disclose the poll's sponsor violated the American Association for Public Opinion Research's (AAPOR) code of professional ethics and practices.

Zogby confirms that he did the poll. But he adds, "There is nothing that forces me to reveal [a sponsor's identity]. If I'm issuing that as a Zogby Poll, you know, then I'm fine and willing to take the heat." Zogby also opines, "The credibility is in the numbers, not the sponsorship." In fact, an advice sheet to journalists from the National Council on Public Polls (NCPP) warns otherwise. "You must know who paid for the survey," it reads, "because that tells you -- and your audience -- who thought these topics are important enough to spend money finding out what people think. This is central to the whole issue of why the poll was done."

Yet Zogby is right about his freedom from regulation; he is not compelled to reveal his sponsors. Because industry self-regulation is weak, self-interested polling is often mislabeled, and the media seem not to care.

Look more closely at the Doris Day Animal League survey. The New York Times Magazine report that 51 percent of Americans think "primates are entitled to the same rights as human children" goes far beyond anything in the actual poll. First, the poll didn't ask about primates -- a category including anything from pygmy mouse lemurs to gorillas -- but about chimpanzees. Second, the actual question gave respondents four options to choose from: In brief, they could say that chimps ought to be treated "like property," "similar to children," "the same as adults" or "not sure." Given this particular set of choices, option two was the obvious pick -- almost as if respondents were steered toward it. And after 51 percent had chosen "similar to children," the Zogby survey inexplicably translated "similar" into "the same" in its conclusions -- a very big difference. The Doris Day Animal League then reported this in its press materials.

Organizations such as the NCPP and the AAPOR have guidelines and standards stressing openness, balanced questions, transparency and so forth. But as self-regulators, they've rarely censured individual pollsters. One of the exceptions is Republican pollster Frank Luntz, who was reprimanded by the AAPOR in 1997 because he "repeatedly refused to make public essential facts about his research on public attitudes about the Republicans' 'Contract with America.'" The AAPOR, however, has not taken on Zogby.

It wasn't always clear that John Zogby would end up a pollster: For a while he was a consumer activist in his hometown of Utica, N.Y., and at one point even ran for mayor. In the early 1980s, he was heavily involved, along with his brother James Zogby, in Arab-American political activism. But since the founding of his company in 1984, Zogby, a second generation Lebanese American, has become a dominant figure in the polling industry. Today no one doubts Zogby's political insightfulness, and the fact that he still works from Utica allows him to inject a helpful outside perspective into the cliquish world of Beltway politics.

Yet Zogby is also very much the businessman, one who has seen his firm grow steadily over the past several years into an outfit with some 500 full- and part-time employees. Roughly two-thirds of the 300 to 500 polling projects conducted each year by Zogby International are corporate or private-sector work; business clients run the gamut, from Coca-Cola to Philip Morris to Microsoft. Such corporate contracts, of course, tend to be the most lucrative in the polling business.

To a significant extent, the entire edifice rests upon Zogby's well-remembered success in the 1996 presidential race between Bill Clinton and Bob Dole. "All hail Zogby, the pollster who conquered the 1996 election," wrote Washington Post pollster Richard Morin after the results came in. Zogby had forecast an 8.1 percent Clinton margin, and the actual margin was 8.4 percent. Most of Zogby's media-polling contracts date from 1996 and afterward. It didn't hurt that he was also quite accurate in the 2000 presidential contest. (Not that presidential races are necessarily the best way of judging pollster accuracy: By the end, most polls are within a few points of one another.)

Zogby does have has his detractors among the polling fraternity. "The pollsters have a view of Zogby that doesn't seem to be shared by the news organizations," observes Warren Mitofsky, who sits on the polling review board of the NCPP. Zogby's performance deteriorated somewhat in the 2002 elections (which he says prompted an internal audit). According to an NCPP postmortem, Zogby got five races wrong out of 17 polled on a nonpartisan basis. His final Colorado Senate poll, for instance, put Democratic challenger Tom Strickland ahead of Republican incumbent Wayne Allard by a margin of 49 percent to 44 percent. (Allard actually won with 51 percent to Strickland's 45 percent.) Following the election, Zogby put out a mea culpa comparing his firm to the New York Yankees, which despite failing to win the 2002 World Series was still "the best team in baseball."

Moving Right

Zogby describes his personal political history as "very left Democrat." His brother James, head of the Arab American Institute, was an adviser to Al Gore's presidential campaign. Yet 1996 helped establish John Zogby as a favorite pollster with the political right. This crossover potential has allowed him to work for groups from the Club for Growth to the National Environmental Trust while still being labeled objective by the media.

To see how Zogby earned his cachet with conservatives, consider the context of the 1996 elections. The year 1996 was a watershed one for polling because many mainstream media organizations, including the CBS News/New York Times Poll, significantly overstated Clinton's lead, predicting a double-digit victory. Following the election, Everett Carll Ladd Jr., director of the Roper Center for Public Opinion Research, even wrote an influential article for The Chronicle of Higher Education (reprinted in The Wall Street Journal) titled, "The Election Polls: An American Waterloo." Ladd called the polling "so flawed that the entire enterprise should be reviewed by a blue-ribbon panel of experts." Though much overstated, Ladd's scathing critique -- especially its Wall Street Journal version -- fed into a widespread sense of disillusionment with polling among political conservatives. Republicans, argued Ladd, view polls as tools of the liberal media and are less likely to respond to them, creating a pro-Democrat bias of precisely the sort that plagued the 1996 election.

Into this breach stepped Zogby. Throughout the 1996 election season, his polls had shown a far closer race than others' had. Critics had expressed disbelief but he had been vindicated by the final election result. Zogby claimed to remedy the perceived problem of Democratic bias by weighting his data according to a previously determined distribution of party affiliation: 34.5 percent Democrats, 34 percent Republicans. This somewhat subjective approach gives some academic public-opinion specialists serious heartburn. As the University of Michigan political scientist Michael Traugott explains, "There's no known distribution of party identification in the sense that we think of a known distribution of sex or race. All we know comes from other survey data, so it has to be an estimate by definition."

Even if all of Zogby's techniques couldn't be taught in a university course, that critique seemed irrelevant after he called the election correctly. By February 1998, Zogby had been asked by Rush Limbaugh to do a poll on the Monica Lewinsky scandal, one that, by emphasizing moral questions, would differ from surveys showing widespread support for the president in the face of Kenneth Starr's inquiries. Zogby obliged. The five-question poll opened with the following: "Suppose you are ready to hire a candidate who is well-qualified for the job, but then you find out that they like to have consensual sex with subordinates. Still hire them?"

Another question asked whether it was "immoral" for a U.S. president to have "consensual sex with a 21-yr-old intern." Sure enough, two-thirds of respondents did indeed consider it immoral. Of course, their only other option was that such behavior was "acceptable," something even Clinton's defenders probably didn't agree with.

The Prospect's reviewer of the Limbaugh poll, Cornell University communications professor Dietram Scheufele, notes, "It is possible that the answers to some of these questions were influenced by questions that were asked before, i.e., by question-order effects." In other words, the premise of the first question could influence responses to subsequent questions. Unless the questions were randomly rotated, this would skew later answers.

Rush Limbaugh bestowed on Zogby the "my favorite pollster" mantle, a kind of calling card for use among political conservatives. By October of 1998, Zogby had reiterated in National Review his findings about the public's opinion of the Lewinsky affair. That year Zogby also did some 60 polls for the Republican Congressional Committee. It's no wonder that many today still think he's a Republican pollster.

More precisely, Zogby is a pollster who works with a lot of Republicans, and in ways that are not always disclosed. Most journalists were probably unaware that some of Zogby's so-called American Values Polls were a joint venture with an organization called Associated Television News, which has a very strong Republican pedigree. Associated Television News is run by Bradley O'Leary, a longtime Republican consultant known for his legendary fundraising abilities and for doing direct mail for the National Rifle Association (NRA). Zogby told the Prospect that O'Leary's role in the surveys wasn't always made apparent but, "Anyone who asked, to the best of my knowledge, was told." However, when columnist Arianna Huffington asked Zogby about the funder of an American Values Poll in April of 2000, according to her column, he responded, "I can't say who it is, but he publishes a newsletter in which he prints the poll's results." Presumably that newsletter would be the O'Leary Report.

The strongly Republican slant of the O'Leary-Zogby surveys is unmistakable. One released in October 2000 found that voters favored George W. Bush over Gore on "20 out of 25" campaign issues. Or, as the Zogby International/Associated Television News press release put it, "Bush Overwhelms Gore On Presidential Campaign's Major Public Policy Issues." That's a pretty convenient finding for a longtime Republican consultant just before a presidential election, which may be why Associated Television News was only identified in the release as an organization that "has covered domestic and international news for 20 years" while Zogby International was described as "a respected, non-partisan polling firm."

The poll contrasted purported candidate positions on different issues, and asked respondents to choose which they favored. Bush-Cheney always came first, Gore-Lieberman second. The poll used loaded language such as "partial-birth abortions" (a term coined by anti-abortionists) and tended to define the Gore-Lieberman position in a politically unappealing way. Cornell's Scheufele also notes that the poll created "false dichotomies" by forcing respondents to answer complicated public-policy questions in a simplistic either-or format. For example:

Bush-Cheney say we need to test teachers and better train those who do not meet minimum standards. Students must meet minimum academic requirements before passing, and more funds should be allocated to help state programs. Gore-Lieberman say more teachers should be hired at higher wages, more classrooms should be built, and the federal government should take more control over our educational system to achieve a better balance between rich and poor school districts.

Unsurprisingly, with this framing, 53 percent approved the Bush-Cheney position to just 34 percent for the Gore-Lieberman position. Here's another: "Bush-Cheney say that tax refunds should be returned to those who were overtaxed. Gore-Lieberman say that tax refunds should be used to fund the federal government." Hanging the albatross of the "federal government" around Gore-Lieberman's neck -- while painting Bush-Cheney as the champion of the "overtaxed" -- sounds more like a Republican National Committee press release than a poll.

When the conservative Washington Times covered one of the polls, neither O'Leary's Republican efforts nor his NRA work was mentioned. On Zogby's Web site, meanwhile, a December 2000 American Values Poll with flattering results for the NRA also made no mention of O'Leary. Other American Values Poll results invariably favored conservatives. When asked to explain the striking Republican slant to these surveys, Zogby said: "Call Karl Rove at the White House and ask what he thinks of me. He'll tell you that he hates my guts."

Cato Calling

Among Zogby's more dubious findings have been his polls on Social Security for the libertarian Cato Institute. Academic research has shown that public opinion on Social Security reform varies greatly depending on the questions asked. If respondents are merely asked whether they think people should have the option of investing part of their Social Security income in private accounts, they approve by a margin of roughly 2-to-1. But the response changes dramatically if people are clearly warned that such privatization could have negative consequences, such as cuts in guaranteed benefits. "As soon as the public is given a sense of what the risks are that we entail to ourselves as individuals by partially privatizing Social Security, people then are against it," explains Fay Lomax Cook, director of the Institute for Policy Research at Northwestern University.

Zogby insists he asks "balanced questions" on Social Security, but consider his Cato polls. The latest, in the summer of 2002, began with this question: "There are some in government who advocate changing the Social Security system to give younger workers the choice to invest a portion of their Social Security taxes through individual accounts similar to IRAs or 401(k) plans. Would you [support or oppose]?"

Sure enough, without any mention of risk, 68.1 percent of the 1,109 likely voters sampled vouched their support. The same Cato poll even managed to use the Enron scandal to demonstrate support for privatization:

With which statement do you most agree? A: The Enron scandal shows the dangers of the stock market and why we must maintain Social Security as it is and not allow individuals to invest their payroll taxes in personal retirement accounts. B: The Enron scandal proves that people need more choice and more control over their retirement savings, including allowing workers the option to invest part of their Social Security taxes in a personal retirement account.

Here 63.3 percent chose option B, perhaps because A did such a poor job of framing the argument that business scandals should cause wariness about privatization. Who could object to "more choice and more control"?

Other pollsters have approached Social Security questions more carefully. A December 2002 Los Angeles Times Poll presented the complexities of Social Security privatization and found that 55 percent disapproved of "allowing younger workers to divert their payroll tax money from Social Security into private investment accounts." In the Los Angeles Times Poll, those who approved of partial privatization -- just 38 percent -- were subsequently asked a follow-up question: "Would you still support this proposal if it meant a reduction in the guaranteed benefits retirees receive through the Social Security system?" Thirty-nine percent of the sub-group then said they would be opposed.

Later questions in Zogby's poll made some slight allusion to privatization's risks but no mention of the possibility of a reduction in guaranteed benefits for retirees. Cato relentlessly publicized the finding from the first Zogby question. "Two-Thirds of Likely Voters Support Personal Social Security Accounts," announced Cato's Web site, citing the "respected independent polling firm Zogby International." Zogby protests that he has no control over how his clients and the media cite his results, but he certainly appeared to lend his endorsement by attending Cato press conferences and other events to discuss his findings. United Press International headlined its story about the survey, "Poll says majority wants Soc. Sec. reform," which quoted Zogby on the alleged popularity of the Cato program. "Republicans should wake up and realize they have a winner," Zogby said.

Afterward, supporters of partial privatization had Zogby's poll to cite. In a December 2002 Weekly Standard article arguing that Social Security reform was still very much on President Bush's agenda, Fred Barnes referred to Zogby's Social Security polling as if it came from an objective source rather than a pollster employed by Cato. In a National Review Online article published in September 2002, meanwhile, Stephen Moore and Thomas L. Rhodes of the conservative Club for Growth -- which, as previously mentioned, has also used Zogby's polling -- cited the 68 percent figure. Moore and Rhodes did disclose Cato's role in the poll but didn't mention that Zogby's question omitted possible risks of privatization.

Finally, consider an August 2001 Zogby poll for the conservative Seattle-based Discovery Institute, which advances the theory of "Intelligent Design" (ID), a more subtle successor to Biblical creationism, as a rival to evolution in high-school science classes. The Zogby Poll asked:

Do you strongly agree, somewhat agree, somewhat disagree, or strongly disagree with the following statement: 'When Darwin's theory of evolution is taught in schools, students should also be able to learn about scientific evidence that points to an intelligent design of life.'

An impressive 78 percent of respondents agreed with the statement; 53 percent of them strongly agreed. At first glance this might seem innocuous enough -- who could oppose the teaching of scientific evidence? But how many respondents grasped that "intelligent design of life" is used as a synonym for divine creation? Also, as Eugenie C. Scott of the National Center for Science Education points out, the premise that scientific evidence supporting ID actually exists is a highly dubious one. The American Association for the Advancement of Science has specifically stated that ID is not science.

As with his polling for Cato, Zogby's Discovery Institute work has been widely cited by ID proponents. In part, Zogby is just testing messages for interest groups, which he reasonably calls a "legitimately defined methodology." But Zogby is also trading on his reputation as a legitimate, media-certified pollster to help groups disseminate inflated claims about public opinion based on inventive wording. In his defense, Zogby says he has refused to work for some clients, including ones who were pro-Confederate Flag and militantly anti-gay, and emphasizes that he ultimately controls question wording. "Apparently this doesn't pass your smell test," he says. "I'm telling you, it passes mine."