Before we release the rest of the results of our polling, I felt it necessary to address something.
Maine Today’s Rebekah Metzler wrote a piece on our polling data, and in it devoted a good deal of time to the notion that MECPO – at my request – built a “Republican enthusiasm gap” into our poll in an attempt to reflect the state of the race. After devoting some decent space to my party affiliation, employment, and my completely unconnected consulting work, the implication was clear that in some fashion I was trying to steer the results in one way or another.
I can understand why this is confusing, but I wanted it understood by our readers and anyone inquiring about the poll that we built in no such enthusiasm gap.
What I asked MECPO to do was come up with the most realistic likely voter model based on who we have seen voting for the past several elections. That did in fact include the most recent primary, as well as the 2009 referendum questions – but it also included previous elections as well. At the end of the day, our sample included the following:
- Democratic - 32.8%
- Unenrolled - 32.3%
- Republican - 29.2%
- Green - 5.7%
It should be noted that the highly respected Democratic polling firm, Public Policy Polling, which more than any other polling firm in the last two years, has been getting virtually everything right, had a very different sample when they conducted their polling in September:
- Republican - 39%
- Democratic - 37%
- Independent/Other – 25%
Nobody called the integrity of that poll into question, and it included a very low (by comparison) sample of unenrolled voters, and had a net advantage of two points for Republican voters. The only criticism anyone mustered up about it was a slight (and meaningless) age demographic skew. No one criticized it for having a “built in Republican enthusiasm gap”.
Indeed, the poll was described as having an “unusual methodology” (which it did not), and my exact quote given said nothing about trying to build in an enthusiasm gap:
“I wanted him to capture what a likely voter is really going to be this year,” said Gagnon, who has done contract work with both Scontras and Jason Levesque, the Republican running in Maine’s 2nd Congressional District. “(Smith) looked at the primary, he looked at the 2009 referendum vote and then some previous election cycles as well, and came up with a formula that I think came out with a very good indication of who is actually going to be coming out to vote.”
In other words, I asked for – and got – a model that gave us a good sample of who was going to come out and vote. Just like every other pollster in the universe. In a year of higher Republican enthusiasm and likelihood of voting, I think it would have been entirely appropriate to have a much higher GOP sample in our poll than we eventually did. The reality is MECPO’s model was actually tilted further to the left than basically most other public polls conducted thus far – with the exception of the most recent Critical Insights poll.
So, as you review the numbers from the gubernatorial results, and look at our upcoming results for the first congressional district, generic legislative ballot, referendum questions, and so on, please keep in mind that the idea that we pushed a poll to be weighted to the right is not only categorically false, but if it were true we did an extremely poor job of cooking the books, given the net disadvantage and low overall sample percentage for Republicans in this poll.
That is all. Stay tuned for more numbers.