Are Presidential Polls Accurate?

Every day, it seems, the American public is bombarded with new polling data trying to discern the state of the presidential race between Democrat Barack Obama and Republican John McCain. The polls fluctuate wildly, which is one of the reasons I take them with a grain of salt, but they do consistently convey one message: Barack Obama is ahead and is likely to win the White House on November 4.

But are the polls accurate? That's the question asked by Michael Barone, writing for the Wall Street Journal on October 22. I have, in the recent past, questioned the importance of national polls in a contest that is decided on a state-by-state basis. I also have never put much stock in exit polls.

According to Barone, however, polls are not quite meaningless. In "Are the Polls Accurate," he says, "Can we trust the polls this year? That's a question many people have been asking as we approach the end of this long, long presidential campaign. As a recovering pollster and continuing poll consumer, my answer is yes - with qualifications." Here's why.

The first problem with polls, Barone says, is that polling is an imperfect science. "Academic pollsters say," he writes, "that to get a really random sample, you should go back to a designated respondent in a specific household time and again until you get a response. But political pollsters who must report results overnight have to take the respondents they can reach." That results in a weighting of respondents to try to approximate the population as a whole.

A second problem is the growing number of households who do not have traditional telephone lines, instead choosing to rely on cell phones for their telephone service. Some have argued that since young people are more likely to rely only on cell phones, and since young people overwhelmingly favor Barack Obama, the Democratic candidate's poll numbers are probably lower than his actual level of support. Barone says, "Gallup and Pew have polled such households, and found their candidate preferences aren't much different from those with landlines; and some pollsters have included cell-phone numbers in their samples."

Third, Barone says, is the growing number of people who won't participate in polling, by choice. "We can't know for sure," Barone writes, "if they're different in some pertinent respects from those who are willing to answer questions." But these are not the only issues concerning polling. According to Barone, "...this year especially, many who ask if we can trust the polls are usually concerned about something else: Can we trust the poll when one of the presidential candidates is black?"

Barone dismisses the so-called "Bradley Effect," citing a series of polls that correctly called the California and Virginia cases often cited in discussions of the alleged phenomenon and saying Daniel Hopkins, a researcher at Harvard University, who, "...after examining dozens of races involving black candidates, reported this year...that he'd found no examples of the 'Bradley Effect' since 1996."

Transitioning to the presidential race, Barone writes, "And what about Barack Obama? In most of the presidential primaries, Sen. Obama received about the same percentage of the votes as he had in the most recent polls." But there was one oddity: "In most primaries Mr. Obama tended to receive higher percentages in exit polls than he did from the voters. What accounts for this discrepancy?"

Part of the difference between exit polls and actual votes can be attributed to voters themselves, Barone says. Since many voters refuse to take part in exit polls, "...it seems likely that Obama voters - more enthusiastic about their candidate than Clinton voters by most measures (like strength of support in poll questions) - were more willing to fill out the exit poll forms and drop them in the box."

What this means, according to Barone, is that Senator Obama will probably get about the same percentage of votes as he does in the final polls before voting. If current numbers hold, Obama would win a majority of the popular vote, "...something only one Democratic nominee, Jimmy Carter, has done in the last 40 years."

On election day, Barone warns to be careful of exit polls for a couple of reasons: 1) exit polls in recent presidential elections have shown Democrats doing better than they actually did; and 2) exit polls in this year's primaries showed Obama faring better than the actual results.

"The exit poll gives us," he says, "and future political scientists, a treasure trove of information about the voting behavior of subgroups of the electorate, and also some useful insight into the reasons why people voted as they did." But, he cautions, "...political polls are imperfect instruments. Reading them right is less a science than an art. We can trust the polls, with qualifications."

In other words, don't take what you are seeing too literally. It might just turn out to be different from what it seems to be.