The Cell Phone Gap
Here's something to keep in mind when looking at polls: Did the pollsters call cell phones as well as traditional landlines? Pollsters and analysts have long worried that polls that relied only upon those reached on landlines --which still is the case in most polls -- might be skewed because an increasingly large number of people, especially younger people, only have cell phones. There's pretty strong evidence now that they are right, meaning many surveys may be under reporting support for Barack Obama since he has considerably strong backing from younger voters.
The Pew Research Center for People & the Press, one of the most meticulous and thoughtful polling organizations around, tried to assess the problem by conducting three major polls since the end of the primaries that included both cell phone and landline samples. "In each case, including cell phone interviews resulted in slightly more support for Obama and slightly less for McCain, a consistent difference of two to three points" in the gap between the two candidates. While the difference is small, it could be significant in a close election.
Many polls compensate for the cell phone problem by increasing the sample of younger voters. But Pew says its research shows that may not work, either. When they studied the cell phone-only samples of young voters and compared them to landline voters in the same age category they found that "cell-only young people are considerably less likely than young people reached by landline to identify with or lean to the Republican Party, and even less likely to say they support John McCain." Fifty-two percent of registered voters under 30 who were reached by landline said they favored Obama while 39% went with McCain, a gap of 14 points. But when it came to the comparable cell-phone only group, Obama's support jumped to 62% and McCain's dropped to 27%, a whopping 35-point margin.
The report has set off a lively discussion at pollster.com, where one commenter suggested that the cell phone effect could be offset by the so-called Bradley effect, under which polls often indicate stronger support for an African American candidate than is born out by the actual voting results.
Finding out whether a poll included cell phones isn't always easy, but those that do usually mention it. You can probably assume you're looking at a landline-only poll unless it says otherwise.