Nate Silver has an interesting graph looking at "likely voter" surveys from an array of recent polls. The surveys in yellow included cell phones. The surveys in gray did not.
As Silver says, "The cellphone polls have Obama ahead by an average of 9.4 points; the landline-only polls, 5.1 points." Obviously this isn't bulletproof evidence of a cell phone effect, and it's unclear if cell phone users vote at rates similar to the rest of the population. But it's suggestive. And, eventually, an effect of this sort is inevitable. Households without landlines tend to be concentrated among demographics that vote Democratic -- namely, the young -- everyone has known that, as some point, excluding cell phone users would begin to harm polling data. But in 2002, 2004, and 2006, the effect was not particularly pronounced. 2008 may be the year that it tips, and by 2010, you may just not see polls that don't include cell users.
As Silver says, "The cellphone polls have Obama ahead by an average of 9.4 points; the landline-only polls, 5.1 points." Obviously this isn't bulletproof evidence of a cell phone effect, and it's unclear if cell phone users vote at rates similar to the rest of the population. But it's suggestive. And, eventually, an effect of this sort is inevitable. Households without landlines tend to be concentrated among demographics that vote Democratic -- namely, the young -- everyone has known that, as some point, excluding cell phone users would begin to harm polling data. But in 2002, 2004, and 2006, the effect was not particularly pronounced. 2008 may be the year that it tips, and by 2010, you may just not see polls that don't include cell users. 
