Via the always-excellent Future Majority, Pew Research has released a new report about the impact of missing cell-phone-only voters in telephone polling.

An analysis of six Pew surveys conducted from September through the weekend before the election shows that estimates based only on landline interviews were likely to have a pro-McCain tilt compared with estimates that included cell phone interviews. But the difference, while statistically significant, was small in absolute terms

There's more below the fold...

According to the exit polls, roughly 20% of all voters didn't own a landline.  This finding is one more indication that as time goes on, cell-phone polling will be an ever-more-necessary component of public opinion research.

This is not happy news for pollsters, for any number of reasons.  First of all, I believe they are prohibited by law from using auto-dial to reach cellphones--every number must be hand-dialed, which is significantly slower.  Secondly, refusal rates are higher--people still see a cell as a "personal phone", and view calls from organizations they don't know as an intrusion.  Years of being mostly free from telemarketers on cell phones means that people are much less willing to accept unknown calls than they were in the days of the land line, especially after the passage of the Do Not Call registry.

So what to do? There is some combination of eating the higher costs, absorbing the lower accuracy and non-phone methods of polling that will minimize these problems, but the industry is still very much looking for a formula.

Ed. note: Please correct me if any of the assertions above are mistaken; it's entirely possible.  And don't think that this will be our last word on the subject.