Today, CBS News and the New York Times released their latest numbers on Florida’s governor’s race. The poll (hold onto your seats for this globe-shattering news) had the race at nearly dead even with Rick Scott holding a one-point lead.
On one hand, this should be good news for Crist. In early September, the same outfit had Scott up five; in a late September/early October poll they had Scott up four; and now that lead has tightened to Scott being up just 1 point. This is undoubtedly a good trend for Crist.
But there is something seriously amiss with the party breakdown of this poll that if this were the poll and the definitive benchmark, would make every Crist backer very – VERY – nervous.
Forgetting the fact that the writer originally had Charlie beating Daryl Jones in 2006, or that the Ind/NPA percentage is simply way too high, there is an important aspect of this poll that simply cannot be overlooked. This poll breaks down party as follows:
Democrats – 36%
Ind/NPAs – 30%
Republicans – 34%
By this reckoning, Democrats should outnumber Republicans by two full percentage points.
That didn’t happen either time Barack Obama was on the ballot and it won’t happen this year. In off-year elections, turnout skews in favor of Republicans. The question is not a matter of “if,” but “by how much?”
That is not to say that Team Crist isn’t doing a great job reshaping the electorate. It is. In fact, it is fair to say that as of this writing, the Democrats are closing in on the Republicans in terms of actual votes cast quicker and better than most observers had expected. But will they actually outpace the GOP as an overall percentage of the electorate? No. Not even the most optimistic Crist-backers will claim that as a goal – ok, some of the most optimistic will claim that, but you get the point.
So if we recalibrate the numbers and reweight these findings to make it a more reasonable Dem -3, then Charlie would likely be 4 or 5 points down. THAT would be bad news.
Here is what is also perplexing about this poll. The original unweighted sample was very different from the final weighted one. In the original, GOP voters outnumber Democrats by almost three points — a number pretty close to what we believe is highly probable. We have to wonder why they skewed it five points more Democrat.
Bring on the New York Times conspiracy theorists.