Editorial: Detailing the pitfall in polling

Tuesday, Nov. 23, 1999 | 9:09 a.m.

The results of a Los Angeles Times poll released over the weekend indicate that Vice President Al Gore and Texas Gov. George W. Bush hold commanding leads as they try to win their party's nominations. In the nationwide poll Gore was selected by 52 percent of Democrats while his challenger, Bill Bradley, received only 34 percent. Bush meanwhile was favored by 59 percent of Republicans and his closest challenger, Sen. John McCain of Arizona, had just 13 percent. But Harvard University's Shorenstein Center on the Press, Politics and Public Policy suggests that we should view skeptically polls, such as the Los Angeles Times survey, because they don't accurately gauge voter support for the candidates.

The highly respected Shorenstein Center notes that most polls -- as did the Los Angeles Times -- ask people to register their choices based on a list of candidates provided by the pollster. But this polling technique forces respondents to essentially choose one candidate, which has the effect of hiding the true percentage of undecided voters. In contrast, a recent Shorenstein Center poll -- conducted about the same time as the Los Angeles Times poll -- instead asked this open-ended question: "Which candidate do you support at this time, or haven't you picked a candidate yet?" When put this way, 64 percent said they don't have a preference so far. The Shorenstein Center question is a better barometer, an acknowledgement that it is highly unlikely voters have made up their minds this soon in the race to capture the White House.

The Shorenstein Center survey also found that the public still isn't focused on political campaigning, noting that the overwhelming majority doesn't know the candidates' positions on important issues. For instance, respondents to the Shorenstein Center poll were asked whether Sen. John McCain supports or opposes campaign finance reform. The Arizona Republican's pledge to overhaul campaign finance laws is a campaign centerpiece and generated considerable media attention, but 77.2 percent said they didn't know his stand on the issue and another 12.8 percent mistakenly thought he opposed this reform.

Polling can provide an accurate snapshot of voter attitudes, but hypothetical presidential matchups this early in the campaign season have little relevance -- except possibly for campaign insiders and political news junkies. The Shorenstein Center study convincingly shows that these polls actually are little more than beauty contests that fail to accurately convey the actual support for presidential candidates. Until voters start paying attention, and genuinely know the candidates' views, polls are more a measurement of name recognition than a demonstration of how well a candidate's ideas are resonating with the public.