Can the Results of Pollsters Be Trusted? Knowing Which Ones to Rely on Is Vital
Gribbin, August, The Washington Times (Washington, DC)
Never before have public opinion polls and pollsters garnered so much attention - or been so influential.
Poll results have kept President Clinton's foundering presidency afloat. In coming weeks, they are expected either to buoy or help scuttle his scandal-scarred reign.
It's certain that polls will dictate candidates' strategies in the midterm congressional races and the 2000 presidential campaign. In fact, polls and pollsters have become indispensable to office seekers, to major news organizations, and - as many scholars and most pollsters see it - to the very functioning of democracy.
Yet polls differ widely. Some can be misleading. Results often conflict. Clearly, some are useless and some pollsters simply shills.
So which can be trusted? And who are these people who question a mere 800 or 1,000 of the country's 270,258,000 residents and purport to reveal the entire nation's opinion?
A new study by The Washington Times concludes that a relatively small number of public opinion and political polls are done well and are truly trustworthy. Scientific sampling had no part in producing the names of the most eminent, influential pollsters operating today. They were journalistically teased from numerous interviews with academics, politicians, pollsters and other insiders - in other words, from people in the know.
The names emerged from talks with that special group of researchers who serve politicians, or conduct policy-issue polls, or both. Among the immense, unknown number of survey researchers in universities and the business community, such practitioners are relatively few. The annual directory of political consultants published in Campaigns & Elections, a trade magazine, lists just 221 "survey research and analysis" firms serving "the campaign industry."
In the past, many of those pollsters came from academia and held advanced degrees in political science, history or statistics. More recently, less-tutored practitioners have been coming from the back rooms of various campaign headquarters. Knowing comparatively little about statistics or survey strategies, they cobble together teams of subcontractors to do interviewing, tabulating and the myriad chores that demand technical knowledge and experience.
They can do that because, as veteran political pollster Bruce Blakeman of Wirthlin Worldwide reports, "There are no standards. Anybody can become a pollster just by hanging out a shingle."
A pioneer some 70 years ago in political and public opinion polling was a former journalist, George "Ted" Gallup. He polled for a newspaper consortium and, in 1936, became one of three pollsters who introduced what then passed for "scientific" assessment of the voters' minds. (The others were Elmo Roper and Archibald Crossley, who became well-known for a time and are famous in polling history.)
After 1936, the late Mr. Gallup and his polls earned widespread respect. The Gallup Organization, now run by George Gallup Jr., is regarded by professionals as the nation's premier public-opinion research firm.
The Princeton, N.J.-based researcher produces the independent Gallup Poll on a huge range of topics, including public affairs. It tops nearly every expert's list of pollsters most highly respected for trustworthiness.
The runners-up are:
* Louis Harris & Associates Inc., New York
* The Pew Research Center for People and the Press, Washington
* Princeton Survey Research Associates, Princeton, N.J.
* Yankelovich Partners, Norwalk, Conn.
* Zogby International, Utica, N.Y.
Zogby International's founder, John Zogby, broke into polling in 1989 and currently runs the hottest firm producing election and public-affairs surveys. He won recognition when he became the only pollster to correctly call New York's gubernatorial race in 1994 and, among other coups, predicted the 1996 presidential race to the exact percentage point. …