Half of voters under the age of 34 and one-third of middle-aged voters are missing from election polls unless pollsters make a special effort to include them. These “cell-phone” respondents, who are not reachable on a home telephone, are excluded from opinion polls unless a pollster goes out of its way, at additional expense, to include them. Here is how serious the situation is today, extended through the 2012 election.
Most state-wide opinion polls, including all the polls conducted by some SurveyUSA competitors, do not include cell-phone respondents. This is indefensible. It is worse than just “cutting a corner.” It is equivalent to running half a race and claiming to have run all of it. Of the few state-level pollsters who do attempt to include respondents not reachable on a home-phone, none is as systematic as SurveyUSA in showing you how the answers from home-phone respondents differ from cell-phone respondents. No local research company has done more side-by-side testing on how best to account for cell-phone respondents than has SurveyUSA. SurveyUSA is proud of how aggressively it is working this problem; our attention to detail is reflected in SurveyUSA’s 20-year track record.