There has been a lot of poll- and pollster-bashing this primary season, some deserved but much undeserved. All in all, the polls have done a pretty good job of forecasting primary outcomes, and have easily outperformed the pundits.
But, the landscape is strewn with polling casualties this year. Bob Dole fired his pollsters for failing to predict his loss in Delaware. Exit polls have been blamed (wrongly) for miscalling the order of second and third place finishers in Arizona. And just about everyone has despaired over the more lamented than apparent volatility of the polls in New Hampshire.
Little Delaware was the big problem for pollsters this year. Neither media nor campaign surveys saw the Steve Forbes win coming. According to the Washington Times, Public Opinion Strategies was sacked by the Dole campaign for reporting an 11% point lead in the state. A Louis Harris poll taken two days before the primary for the Wilmington News Journal also showed Dole with an 11% point lead. The message may be that primaries that receive little attention from the candidates and the media are more difficult to assess than higher profile races.
The polls were right for the most part in other states, especially the crucial ones. In New Hampshire, four out of the five polls that surveyed through Monday night correctly predicted the winner and second place finisher. The final NY Post/Fox News poll in Arizona had Forbes ahead of Dole and Pat Buchanan. Similarly, the South Carolina outcome was forecast accurately by the media sponsored polls. The Junior Tuesday and Super Tuesday primary polls all predicted Dole’s sweep (see table at the end of the document).
Exit polls have come under fire for the erroneous early call by networks that Buchanan would finish second and Dole third in Arizona. In fairness to the exit poll, the prediction was based not only on its data, but also based on the estimates of a statistical model that used the actual vote as it came in on election night. Further, the exit poll did not find Buchanan with a statistically significant lead over Dole. But Voter News Service (VNS) poll director Murray Edelman does not blame chance error alone. He found and advised his clients that there had been a consistent over-representation of Buchanan voters in a number of earlier primary exit polls. Knowing that, VNS itself did not call second and third place in Arizona until the early hours of Wednesday morning.
The pundits were far less accurate than the poll takers in pre-primary forecasts. Their predictions of winners were correct only about half the time: they picked the winner in only 15 of 29 forecasts made about Iowa, New Hampshire, Arizona and South Carolina, as reported in the Hotline. (Predictions were taken from the following issues of Hotline: February 12th, February 19th, February 26th, and March 1st.)
Not a very good record considering that they did have access to polling data that was largely correct …
Wither the Perot Vote
While candidates from Bob Dole to Bill Clinton are trying to figure out how to court the Perot vote, it is apparently disappearing before their very eyes. Recent national surveys have found fewer and fewer “Perot voters”. As an independent candidate in 1992, he captured 19% of the popular vote. Just after the election, 17% of voters in nationwide polls said they had cast a ballot for the billionaire industrialist. Since then, it has been all downhill, either because of denial or failing memories. In 1995, Perot voters slipped to 15%, and they are but 11% in the latest Center surveys.
Polling on the Buchanan Message
After New Hampshire, the national polls immediately sought public reaction to Pat Buchanan and his message. Their findings belie the short-lived media hype about the resonance of the former commentator/speech writer’s positions. Newsweek bannered a March cover “Preaching Fear — Why America Is Listening.” The New York Times front page headlined its analysis “Social Issues Give Buchanan Boost.”
The polls themselves tell a different story. While Buchanan’s economic message strikes a chord with economically anxious Americans, the public is divided, at best, over his stands on social questions. For example, the CBS/NYT poll found half the public (52%) saying that homosexuality is morally wrong. But the other half says either it’s okay (12%), or they do not care about the moral question of homosexuality (34% ). This mirrors our finding that 45% believe that homosexuality should be accepted as a way of life, while 50% say it should be discouraged by society.
Similarly, CBS/NYT found little consensus about whether the government should do more to promote traditional values (40%), or whether it should stay out of values promotion (54%). And contrary to Buchanan’s view, the public opposes government restrictions on the sale of pornography to adults by nearly two- to-one, according to the same poll.
The issue of abortion galvanized hard core Buchanan brigades in state after state, but nationally, opinion runs strongly against his position. By a 70% to 24% margin, the public opposes a constitutional amendment to ban abortion, according to a Time/CNN survey. An earlier poll also found the public opposed to changing the laws to make it more difficult for women to get an abortion although by a narrower margin (60% to 32%).
Exit polls after the primaries indicated that Buchanan’s economic message is a much bigger draw than are his social positions. Nationally, he tapped into broad discontent with corporate downsizing policies. A Center for The People & The Press survey last fall found unfavorable opinions of business on the increase, as the public blamed big companies for looking after the concerns of bosses and stockholders instead of its workers and customers. But these attitudes did not translate into calls for action. As recently as last October, Americans did not blame the country’s problems or their personal problems on Wall Street or on business corporations. Buchanan’s success may change that view, however, as newly minted populists such as Dole now deplore the treatment of American workers at the hands of greedy corporate bosses.
Opinion about protectionism, a subset of the Buchanan economic message, is less conclusive. Both the Time/CNN and the CBS/NYT polls found solid majorities saying that the U.S. should impose tariffs in order to protect American jobs and industries. However, both surveys also show that there is another side to opinion about free trade. Americans split evenly as to whether trade with other countries creates more jobs for the U.S. (39%) or loses more jobs for the U.S. (40%). The public also acknowledges a distinction that affects its pocketbook: GATT and NAFTA have been mostly bad for workers, it says, but these treaties have been mostly good for consumers. Perhaps as a consequence, the public opposes U.S. withdrawal from NAFTA by a 48% to 34% margin at this time (Time/CNN).
Buchanan gets more unequivocal support on the immigration issue. Nearly two-in-three think that immigrants burden society, rather than strengthen it. The public ranks stopping illegal immigration as a top foreign policy priority, and in the Time/CNN poll, as many as 56% favor suspending legal immigration for five years.
But Buchanan’s support is greater than the sum of attitudes toward his positions. In every primary state, his supporters were more turned on by the perception that he is “standing up for what he believes” than by any particular issue position. Buchanan’s ability to project sincerity, strength of conviction, and an understanding of the needs of working people are the hot coals that fueled his campaign.
AND YOU THINK THE REPUBLICAN PARTY IS DIVIDED …
The results of a recently released U.S. Information Agency opinion survey conducted in Bosnia-Hercegovina underscore the long term difficulties of nation-building there. (The survey was commissioned by the USIA’s Office of Research and Media Reaction.) The horrors of the recent war have left virtually no one untouched, but sadly represent just the latest chapter in the history of ethnic hatred rending the former Yugoslavia. According to the personal interview survey, majorities or near majorities of every ethnic group reported that family members were either injured or assaulted during the war. One-third of the Croats and Muslims said they personally had been combatants, as did a 54% majority of Serbs. Fully 42% of Croats were injured in the fighting, as were 19% of Sarejvo residents and 13% of Serbs.
Bottom line attitudes about the prospects for peace are as grim as the recent experiences of most residents. While acknowledging that the Dayton peace accords are better than war, most Serbs and Croats say that war has destroyed the possibility of living peacefully with each other. Each of these two groups also believe it can feel completely safe only when it is the majority nationality in its country. On the other hand, the predominately Muslim inhabitants of Sarajevo, Tuzla and Zenica are more sanguine about the chances of living peacefully with other ethnic groups. Not surprisingly, perhaps, since they are the plurality population, a solid majority of these mostly Muslim inhabitants believe that minority groups can feel safe in their country.
For the section, “Polling on the Buchanan Message” the following surveys are used in the analysis:
CBS/New York Times poll, conducted among a nationwide random sample of 1,223 adults, interviewed February 22-24, 1996. All CBS/NYT references within the text are based on this poll.
“Voter Anxiety Dividing GOP; Energized Dems Backing Clinton.” Times Mirror Center for The People & The Press, November 14, 1995. Washington, DC.)
Time/CNN poll, conducted among a nationwide sample of 1,058 adults, February 21-22, 1996. All Time/CNN references within the text are based on this poll.
“The Vocal Minority in American Politics.” Times Mirror Center for The People & The Press, July 16, 1993. Washington, DC.
“The New Political Landscape.” Times Mirror Center For The People & The Press, October, 1994. Washington, DC.
“Public Opinion of the UN: Strong Support, Strong Criticism.” Times Mirror Center for The People & The Press, June 25, 1995. Washington, DC.