Do cellphone-less polls have a conservative bias?

A new Brown poll has the quarterback of pension politics football team leading Providence’s most oft-plagiarized mayor in a likely Democratic primary for governor in 2014 with 24 percent still undecided. But it’s at least worth noting that the Taubman Center has often been off mark with election forecasting.

“A new Brown University survey of Rhode Island voters finds Democratic State Treasurer Frank Caprio has widened his lead over Independent candidate Lincoln Chafee in the race for governor,” an October 2010 press release says.

The Taubman Center predicted Caprio would win by 7 and he lost by 13 – that’s a 20 point swing. Sure, Caprio then told off the president on hate radio, but Brown botched other races that year too: It said David Cicilline would win by a whopping 39 to 20 margin and he won 50 to 44. And Ralph Mollis only beat his Republican challenger by one point, not the 11 point swing Brown predicted. In 2006, a Brown poll thought Governor Don Carcieri would hold on to his office by 12 points. He won by less than two.

In 2012, Abel Collins own internal poll was almost as accurate as the Taubman Center in predicting his electoral results, only in the opposite direction.

It’s no secret that polls can get it wrong, and a huge reason for this is they don’t often account for those without landlines. What’s interesting is that, at least with 2012 presidential polls, the ones that got it wrong tended to anticipate a more conservative electorate.

Here’s hoping that’s the case with this Brown poll.