Presidential Polls And The Bridge Of Death

God knows how many election polling outfits there are these days.  Once upon a time, there was just Gallup; then it became Gallup and Harris; now there are dozens and perhaps hundreds.  Who knows how skilled they are at their sampling, their weighting of likely voters, and the other factors that separate meaningful polls from floss and ear wax?

Of course, the main problem with polls is that you don’t know whether the respondents are telling the truth.  If only the penalty for giving a false answer to a polling question was like that imposed at the Bridge of Death . . . .


Cell Phones, Land Lines, And Survey Results

Public opinion surveys have been a staple of American politics for years.  They have a proven track record — at least, they do when the pollsters figure out how to identify an appropriate sample that mirrors the people who actually will cast ballots in the election and then reach those people to learn their views.  If you can’t accurately do both, you risk results that are as misguided as the infamous 1936 Literary Digest poll that embarrassingly predicted that Alf Landon would beat FDR in a landslide when in fact the exact opposite occurred.

In the modern cell phone and smart phone world, can pollsters know with any assurance that they have reached an appropriate sample of voters?  For years, pollsters relied on land line telephones to conduct their surveys.  Recently, however, many Americans have dropped their land line phones as a nuisance and unnecessary expense.  In 2007, nearly 13 percent of American households had no land line phone.  By 2008, that number had jumped to 20 percent and it has only increased since then as millions more — including Kish and me — have gone totally wireless.

The question for pollsters is whether the demographic and political characteristics of wireless households are different from those of households that still cling (bitterly?) to their land lines.  Some pollsters think that may be the case, reasoning that cell phone-only people probably are younger, unmarried, don’t own a home, and so forth.  That may have been true initially, but my guess is that as wireless-only status has become more common, and even old farts like Kish and me have joined that segment of the population, the differences have been minimized.  The important point, in any case, is that no one really knows.

So, in these days leading up to Election Day, let’s not pay too much attention to the polls and their competing results.  The only poll that really matters is the one that will occur on November 2, and all registered voters — be they wireless Gen Xers or land line fogies — will have an equal opportunity to be counted.