In the 2010 election, SurveyUSA contrasted the vote of home-phone and cell-phone respondents in 7 different candidate elections. On average, cell-phone respondents voted 9 points more Democratic than did home-phone respondents. Had SurveyUSA not made special effort to reach the cell-phone respondents, its poll results would have been more Republican.

Polls that exclude cell-phone respondents, such as those conducted by many SurveyUSA competitors, risk mis-characterizing public opinion. SurveyUSA estimates that by Election 2012, half of voters age 25 to 34, and one-third of voters age 35 to 44, will not be reachable on a home-phone. Make certain the polls you commission, and the competing polls you analyze, properly account for cell-phone respondents.