|November 4, 1999
Volume 30 Number 7
|A publication for the faculty, staff, administrators, and friends of California State University, Chico|
CBS Pollster: Good Polls Mirror the Public and Keep Politicians Honest
"I'm responsible in 1996 for saying the worst eight words in the English language, 'I'm really sorry, Dan, we made a mistake," said Kathleen Frankovic, who recently visited CSU, Chico as a Presidential Visiting Scholar. Frankovic, director of Surveys and producer for CBS News, had to tell Dan Rather that the exit poll prediction of a turnover in the 1996 New Hampshire Senate race was wrong, and to tell him after he'd called the race on national television. Although all the news organizations got it wrong, since they were all working with the same data, Frankovic found some redemption. "We were proud to say we took it back an hour before any other network took it back." Fortunately, she said, they usually get it right.
Frankovic, brought here by the Department of Political Science, discussed the history of polling, 1998's polls, and future challenges as part of the Visiting President's Scholars Lecture Series. A political scientist, Frankovic moved from teaching at the University of Vermont, to CBS News in the 1970s, where she is now in charge of the CBS News/New York Times Polls.
In 1998, the year of Clinton's impeachment hearings, more Americans answered more survey questions than ever before. In fact, "CBS News, alone or with the New York Times, conducted forty-eight national opinion surveys, and we weren't alone in increasing the number of polls," Frankovic explained.
Polling has a long history. In 1824, when many people did not directly vote for their electoral college representatives, straw polls showed who people would vote for if they could vote. "The straw polls had an impact. Four years later, there were only two states that didn't have a public say in electing the president, and turnout doubled between 1824 and 1828," Frankovic said. Polling was extolled as a democratizing activity, giving voice to public opinion.
Over the years, polling became a way for journalists to cover politics without only interviewing political leaders, but scientific polling didn't emerge until the 1930s. George Gallup, Elmo Roper, and Andrew Crossley used new methodologies, such as selecting a representative sample from the population. Instead of sending mass mailings, they could interview 2,000 or 3,000 people and still be more accurate than earlier studies.
Polling really took off in the 1970s. Frankovic explained the expansion "was due to the cynicism that was developing first among the press and then the public, sort of a residual effect of the Vietnam war and of Watergate." Journalists didn't trust the politicians' opinion polls and began to conduct their own. Technological improvements made this possible. By the mid-seventies, over 90 percent of American households had telephones. Surveys could be reliably and accurately conducted over the phone from a centralized calling room where interviewers could be directly supervised.
In 1998, polls were conducted with every new revelation and action, from the initial Monica Lewinsky story, through the State of the Union address, to President Clinton's grand jury testimony and the impeachment hearings.
Throughout its history, polling has had its detractors. In 1896, the Democratic Party leader described straw polls as "a scheme of fraudulent debauchery and the first step to do away with popular elections under the law." In 1998, Henry Hyde "blamed opinion polls...basically saying that because there were opinion polls, Bill Clinton wasn't impeached," Frankovic said. "The attacks on polls obviously are made strategically by trailing candidates."
While politicians may attack polls for questionable campaign reasons, Frankovic noted, "There are legitimate concerns that we have about polls, methodological questions that are absolutely critical to ask about them." These concerns include sample size, sample selection, interviewing dates, response rates, and question wording and order.
Frankovic told how the importance of question wording was recently demonstrated when the Roper Center reported that a quarter of Americans believed that the Holocaust had never happened. This result was based the question: "Do you think it is possible or impossible that the Holocaust never happened?" As it turned out, the Roper Center's numbers reflected the confusing nature of the question and not the opinion of Americans.
"Even small changes in question wording can sometimes change the answers," Frankovic explained, giving the example of two questions asked by The Gallup Poll in August 1998. In response to the first question, "Please say if you have a favorable or unfavorable opinion of Bill Clinton," 55 percent responded "favorable." When asked, "Thinking about Bill Clinton as a person, do you have a favorable or unfavorable opinion of him?" 44 percent said "favorable." This was reported as a drop until the question wording was examined and a retraction issued.
The challenge to pollsters is to be accurate, and "despite all the attacks on public opinion polls, [they are] really extraordinarily successful," Frankovic said. Polls are generally accepted. The greater challenge, according to Frankovic, is to make sure pollsters disclose their methodology and make realistic interpretations of the data they collect.
Frankovic concluded, "News polls manage to do two things: they keep the government honest by not allowing misrepresentations of public opinion to be presented to the public....News polls present a mirror to the public, and actually permit individuals to understand where they fit into the political system....democracy matters, and people's opinions do really count." -- BA
Chico | Admissions | Bookstore
| Catalog | Schedule | Library
California State University, Chico
400 West First Street
Chico, CA 95929-0040