There are two ways for Republicans to approach President Barack Obama's growing lead in the polls. A) Introspection: a reasoned self-examination of why voters seem to be rejecting their message and their candidates; or B) Denial: a fit of paranoid delusion that blames the widening gap on a conspiracy among pollsters to skew the results in favor of the Democrats.
Okay. So let's say Republicans are right, and the presidential polls are skewed in President Obama's favor, intentionally or not. So how unskewed would these polls have to be to give Mitt Romney a reasonable chance of winning the White House? That's the question Darryl over at my old haunts HA asked himself, rerunning his regular polling analysis* while introducing a bias in Romney's favor. Darryl's conclusion?
Recent polls would have to be skewed 6 points in Romney's favor to give him a narrow electoral lead.
The graph is clear…to eek out a win, Romney has to move the electorate across the board by a remarkable 6%. That is, he is 6% behind in the polling now.
[...] To win with at least a 95% probability, Romney needs to shift things by 8%. To have the kind of lead that Obama now enjoys—with a solid 100% probability of winning—Romney needs a 10% shift.
That's an awfully big skew Republicans are imagining. If Romney goes on to win in November, this could prove to be the biggest polling fuck-up since "Dewey Wins!" Or... this could just prove how incredibly pathological the Republican Party truly has become.
* FYI, Darryl is a professor at the UW who does this sort of statistical analysis for a living. In 2008 his meta-polling analysis predicted the outcome within two one electoral votes, closer than the much ballyhooed Nate Silver.