Correcting the Polls Gives Romney a 7.8% Lead

You know something is wrong with the polling data, but you can’t put your finger on it. You see that every poll is over-sampling Democrats or using 2008 turnout models, but you don’t know how to mentally adjust for that and determine what the polls really mean. You’re not alone, and now there is a solution.

You’re probably familiar with the Real Clear Politics average of polls where all major polls are averaged together. Well, now get familiar with the UnskewedPolls.com average of polls. While the site isn’t unskewed, as the name suggests, it sure is better than the leftward skew we’re seeing in the rest of the media. Instead, this website applies a single, realistic, skew and applies it to the polls from major news organizations.

President Obama won the 2008 election when Democratic turnout exceeded Republican turnout by seven percentage points. Yet a recent Gallup poll used a turnout model for the 2012 election assuming Democratic turnout would exceed that – 11 more points than Republicans. While a Republican voter like me knows that Gallup poll is inflating Obama’s numbers, there’s no way for me to know by how much.

Enter UnskewedPolls.com and the Rasmussen turnout model.
Rasmussen has been using a turnout model based on 2004, and performs monthly party identification tracking polls. They don’t take the exceptionally high 2008 Democratic turnout and they don’t take the exceptionally high 2010 Republican turnout (even though one could argue that the 2010 model is probably more appropriate given Republican excitement this year). Instead, Rasmussen tracks current party affiliation.

UnskewedPolls.com applies the Rasmussen party affiliation results to other polls, effectively applying a single turnout model across all polling results. So an NBC/Wall Street Journal poll that gives Obama a 50% to 45% lead over Romney becomes a lead for Romney, 51% to 44%. A NY Times/CBS News Poll that has Obama leading 49% to 46% becomes a Romney lead of 51% to 44%.

I’ll readily acknowledge that the entire venture may be an exercise in wishful thinking. But... 

• If, like me, you don’t believe Democrat voters are more excited to vote in 2012 as they were in 2008
• If, like me, you believe the media has a vested interest in promoting Obama over Romney
• If, like me, you don’t understand the statistical underpinnings of polls but recognize that something just isn’t right

…then UnskewedPolls.com will give you some hope for the November election.

Of the twelve different polls provided on the website, every single one shows Romney leading Obama. I have to wonder why a couple of high-profile polls aren’t included in the list – Gallup and Rasmussen are particularly notable – but the results are undeniable.

The average of all twelve polls gives Romney a 51.8% to 44% lead going into October, and that’s good news for Republicans. I think that’s closer to reality than an Obama lead given the results of his policies and the stagnant economy. Even if you’re a Democrat, you should admit that you’re not as excited this year as you were in 2008, and the polling doesn’t match up. But we’ll all find out in November.

H/T PJ Media