October 2- 8, 2003

city beat

Gale Warning

Lies, Damned Lies and Polls

by Daryl Gale

A quote generally credited to humorist Mark Twain goes something like this -- there are three levels of untruth: lies, damned lies and statistics. Never is that more true than during campaign season. Statistics, like those that come flying at you fast and furious in the form of public opinion polls, can leave a voter’s head spinning. Do you really care if the incumbent is two points down, the challenger is three points ahead and the margin of error makes the race a statistical dead heat? Sure, polls are handy reference guides for campaign flacks, pundits, insiders and columnists, but do they have any real influence over the average voter?

I put that question to some flacks, pundits and insiders this week and got some interesting answers.

"When you talk about polls, you have to differentiate between polls commissioned independently, like media polls, and those commissioned by the candidate or his party or some other special interest," said Neil Oxman, one of Philly's best-known political strategists and media consultants. "Media polls are there to gauge public opinion, and for the most part, act in the public interest. But the short answer to your question is yes: Early polls can affect a candidates' ability to raise money and that certainly has an affect on the outcome. I polled for Sam Katz in '99, and I can tell you that some of those early polls absolutely solidified Katz's status as a serious contender."

Oxman also said that the neck-and-neck nature of this year's mayoral race means that polls are going to be less important than they were four years ago.

"We could send 10 pollsters today out into the streets of Philadelphia and virtually all would come back with numbers that were even, when you consider the usual 3 or 4 percent margin for error," he said. "In the end, people generally vote for who they think is the best candidate, regardless of his polling numbers."

Clay Richards, pollster for Quinnipiac University, agreed with Oxman up to a point. The average voter cares about polls, sure, just not as much as pollsters think.

"The average voter cares which way the wind is blowing, so to that extent they watch the polls," Richards told me. "Certainly not as closely as you or I watch them, but they do watch them. I think polls tend to affect the general turnout as much as which candidate voters pick. It's a double-edged sword."

If your chosen candidate is up by 20 points in the most recent poll, he explained, you may be tempted to stay home on Election Day, because he doesn't need your help to win. On the other hand, if your guy is down by 20, you may abandon him because nobody wants to back a loser. He also reiterated Oxman's point that all polls are not created equal.

"Polls are easily manipulated, which is why independent polls are so important," said Richards. "Because responses vary greatly depending upon how the questions are asked, it's that much more important that voters be able to trust that the information contained in a poll isn't tainted by partisanship."

That taint of partisanship is why people don't trust polls, and that mistrust is healthy, Richards said. People should be skeptical, because polls don't just gauge public opinion, they can also influence it.

The guy who told me that polls don't really mean much to the average voter was a bit surprising. G. Terry Madonna of Millersville University is easily the best-known pollster in the state and probably the most quoted.

"The short answer is no," Madonna said curtly. "This thing has been studied ad nauseum and the evidence suggests that polls are more important to campaign enthusiasm than to public opinion. There's a sense that the competitiveness of a campaign drives the turnout, but the fact is, if you're the kind of person who pays attention to politics and political campaigns, you're more likely to be a voter. There are chronic nonvoters, and no polling data would get them into the voting booth."

Madonna does agree with my other pundits in that he told me that the wording of the polling questions and the sequence of those questions can drastically affect the outcome of a poll, and therefore people should find out who commissioned the poll and what their stake is. I asked representatives of both mayoral campaigns about the numbers, and for once, this was something they could agree on. Probably the only thing.

"Polls give people something to hang their rhetorical hat on," said Mayor John Street's campaign spokesperson Dan Fee. "But in the end, it's the candidate's job to get his supporters to the voting booth, regardless of the polls."

I could have saved the paper the price of a phone call, because Katz spokesperson Nathan Raab told me essentially the same thing.

"We're encouraged by the polling numbers," Raab said, "but the only poll Sam is focused on is on Election Day, and he'll leave no stone unturned between now and Nov. 4."

In the meantime, I'll be closely examining the polling data, because that's my job and they make me do it. You, on the other hand, only have to go to the voting booth and cast your ballot, without regard for the margin of error.

Daryl Gale’s weekly radio show, Dialogues, with co-hosts Rotan Lee and Bill Miller, is burning up the airwaves Fridays 7-10 a.m. on WURD (900 AM) in Philadelphia.