One thing that most analysts Right and Left can agree upon… the 2012 Presidential Election is expected to be CLOSE. It’s no surprise that Barack Obama and Mitt Romney continue to circulate around the same precious group of Swing States. And even if the candidates themselves aren’t watching the daily rollercoaster of electoral polling, their campaign staff are glued to their computer screens, ready to respond to even the slightest variance.
But what if the polls themselves aren’t as accurate as they seem? No, not because of some viscious plot to skew them towards one person or the other. It’s possible that polling just isn’t keeping up with the realities of the 21st century.
A 2010 Pew Research Study revealed that there is a big gap amongst landline respondents and cell phone-only respondents. It turns out that landline respondents are more likely to be Republican, while cell phone respondents are more likely to be Democrats. This makes sense in some ways… the Democratic party is on average much younger and more diverse than the Republican party, so it’s no surprise that fewer Democrats would own a landline or still use it as their primary form of telephone communication.
The study found that during the 2010 race, Republican House candidates led by a twelve-point margin nationally when the data only polled land-line users. But when combined with data from cell-phone users, that lead shrank to just six percentage points. This was from the very last set of polls conducted before the election results, and when the actual votes came in, the real lead for Republican candidates turned out to be just seven percentage points… much closer to the combined data. The GOP advantage was inflated due to a mis-match in polling methodologies.
Even after the results of this study were widely known, some polling agencies still continue to poll only landline users. They skip cell phones entirely when submitting their data. Rasmussen’s polling methodology still openly admits to focusing on landline calls, and polls ZERO cell phone users. The Gallup Poll is in transition to incorporate more cell phone respondents, and has recently come under hot water from the Obama administration for not doing so sooner.
So if this disparity persists in our national polling, what implications could it have for the Presidential race? Are cell phone users being under-polled? We’ll all know the answer in the days following the election.