It’s getting to be crunch time in Iowa, and suddenly there is dearth of major media pollsters reporting the preferences of potential Iowa Caucus voters. Maybe that will change between now and January 3rd, when Iowa holds the first-in-the-nation delegate selection contest for 2012. But don’t bet on it.
Most major media pollsters avoid polling in Iowa altogether, in part because – as Janet Elder of the New York Times wrote four years ago – “All of the eccentricities of the Iowa caucuses make polling Iowa something most pollsters would rather not do.”
There are many reasons why pollsters are leery about polling in Iowa, but none more problematic than obtaining a representative sample of the small number of voters who will actually cast ballots. As Elder wrote, “To try and define the electorate and come up with a framework for polling is to rely as much on luck as savvy.”
CBS’s Kathy Frankovic estimated that using random digit dialing would take 10,000 phone calls to reach about 500 self-identified likely voters. That’s a humungous commitment of resources to produce unreliable results. Some pollsters, therefore, use registered voter lists, but those are not completely accurate and in any case don’t include people who choose to register the day of the election.
There’s another problem – respondents in the polls don’t necessarily accurately predict their own behavior (i.e., some probably lie, saying they will vote when they won’t; and others mean it when they say they will vote, but lose the fervor when election day comes around). And, of course, even as the election approaches, there are many potential voters who haven’t firmly decided whom to support. Most pollsters don’t report (or even measure) that number, but the latest poll by CNN/Time, about two weeks ago, found just 37% firmly committed to a candidate. That’s a whole lot of uncertainty still hanging around out there.
Despite the problems, several media organizations have dared to put their reputations on the line by polling in Iowa: CBS News, New York Times, CNN, Time, ABC News, Washington Post, and Bloomberg.
The list of the major media pollsters, who have conducted polls in Iowa, is shown here:
Note that CNN/Time and NBC/Marist have conducted two polls, while the others have all conducted just one.
But have these news media organizations actually put their polling reputations on the line? Not so far. They still might, if they are willing to poll close enough to the election so that their numbers can be seen as predicting the final results. Right now, they can (legitimately) claim that the poll results they have reported are only the proverbial snapshot in time, too far away to describe how voters will actually behave on election day.
If 2008 is any guide, however, the chances are overwhelming these organizations will not risk conducting polls that can be tested against the actual election. That way, they can give the illusion of telling us what the voters are thinking without having to be tested on their accuracy.
It’s understandable when pollsters refrain from polling in certain contests, because they realize how difficult it is to produce valid results. Gallup, Pew, and USA Today have all taken that position toward the Iowa Caucuses this year. George Gallup himself never polled for primary elections, much less for caucuses, because of the unreliability of such poll results.
But when some major media pollsters drop down into a state, implicitly claiming they can produce accurate results, but then deliberately avoid being tested on their polls come election time, it shows their hypocrisy. We are supposed to trust their results when it’s too early to test them, while they don’t trust their polls to provide accurate results when they can be tested.
Which of the five polling organizations listed above are willing to be tested?
My bet: None.
I’ll let you know what they do right after the vote.
David W. Moore is a Senior Fellow with the Carsey Institute at the University of New Hampshire. He is a former Vice President of the Gallup Organization and was a senior editor with the Gallup Poll for thirteen years. He is author of The Opinion Makers: An Insider Exposes the Truth Behind the Polls (Beacon, 2008; trade paperback edition, 2009). Publishers’ Weekly refers to it as a “succinct and damning critique…Keen and witty throughout