The public's view of President Jimmy Carter was deeply negative in November 1979, thanks to enormous inflation, an oil crisis, and Americans held hostage in Iran. Interviewers for a national AP/NBC poll paused on the second night of interviews, while Carter went on television to declare that the hostages had to be released. Within minutes after the news conference, the pollsters found dramatically higher ratings for Carter and an extraordinary anger among the American people.
In 1986, President Ronald Reagan's second term was humming along, with Gallup Poll job approval ratings in the mid-60s. Then revelations about U.S. arms sales to Iran and profits funneled to Nicaraguan rebels began to appear in the press, and Reagan's standing plummeted.
Why do events make such a difference? The answer is quite simple: Public opinion doesn't just register randomly on the pollsters' screens. It changes in reaction to events, to new information, to fresh analysis. Thus, it follows that polls taken before a major event or news story may well show quite different opinions than those taken after.
While dramatic revelations can have an immediate impact, news, events, and changing circumstances often have a gradual, cumulative effect on the public's views. The more we already know about a situation or public figure-or the more deeply held our views-the more likely our opinions will change more gradually, if at all.
That's why it is always critical to know exactly when a poll was conducted-when the actual interviews took place. The context of news and events that surround a poll can provide vital information to help understand why opinion shifts as it does. This can be particularly true with relation to the President and in political campaigns. The outcome of a debate or the revelation of a major scandal about a contender can be key determinants of public opinion.
Take the phenomena of the "convention bounce." Every four years, the major political parties hold conventions to anoint their presidential candidates. Every four years, these staged events attract an amazing amount of attention from the news media (although less from the TV networks than in the past). In the days immediately after a national convention, the standing of the party's nominee jumps in the polls.
The convention bounce is simple to understand. For about a week, the party and its nominee receive a great deal of air-time and ink, most of it favorable. (The opposing party's candidate often suspends campaigning during the convention.) The convention is a feel-good event and it works, at least briefly. But usually within a week or two, the bounce is gone and the horse race numbers are back more or less where they started.
While events do move public opinion, it does not follow that events always must sway opinion. Many a poll taken in 1998 about the Clinton-Lewinsky scandal demonstrated that more news does not necessarily change opinions. Much to the dismay of his critics, Clinton's job rating has seemed to go up with every new revelation.
Last year, Princeton Survey Research Associates conducted a major poll for Americans Discuss Social Security on options for changing the government retirement and disability system. One such option is to invest some of the billions in Social Security taxes in the stock market for a higher return. The very day the interviewing was closed, the stock market plunged. Down and down it went throughout August. Surely this stock market swoon would change opinions. It did not. A quick repeat of the key questions in a short survey in September found little change in the levels of Americans' support for or opposition to investing Social Security funds in the stock market.
This example demonstrates that public opinion is often based on core values, strongly held beliefs or long experience. Attitudes on religion, family, and patriotism have deep roots. Such opinions are not changed-at least not immediately-by a sudden, surprising event.