Polls Apart

The difficulty of conducting accurate political polls grows daily. Our more mobile society, which depends more on unlisted cell phones than traditional landline phones, is more difficult to measure accurately pollsters find, as several important demographic groups, the young, for example, have disappeared from the usual sources of telephone numbers used by pollsters.

This year, those missing demographic groups just might participate in the November presidential election in numbers large enough to confound the pollsters. Anopther confounding factor is the willingness of poll subjects to truthfully indicate whether they will vote for a black man.

This article in CQ Politics takes a close look at the difficulties in obtaining a valid and reliable poll.

Experts inside and outside the industry question whether pollsters are capturing big enough samples of the population at a time when Americans are increasingly on the move and more likely to be at work or in their cars in the early evening, when many surveys are conducted.

This is all creating considerable angst in a business that remains haunted by a series of Election Night gaffes that helped set the stage for the contested 2000 presidential outcome, and that was shaken again by a series of erroneous forecasts of an Obama victory in January’s New Hampshire primary.

Take a few minutes to read this worthwhile examination of one of our favorite election year activities.