Upon further review, it looks like the only thing that accounts for Kasich going from a four-point lead in the Ohio Poll in September to an eight-point lead today is that the Ohio Poll changed it’s likely voter model to predict even more of a Republican turnout than it did in September.
Here’s the partisan breakdown for the gubernatorial candidates in September:
In order that’s: Strickland, Kasich, and “Don’t Know.”
Now, here’s the partisan breakdown in today’s Ohio Poll:
In order that’s: Strickland, Kasich, Other, and “Don’t Know.”
Even with the addition of “Other” in the mix the partisan breakdown is essentially the same. Yet one poll showed a four-point margin, the other eight?
The only significant shift I could find—anywhere in the poll is that in Southwest Ohio, Kasich went from 11 points up to 29 points up and that people who say their neither support nor oppose the Tea Party favor Kasich 51%-41%—a complete flip from September.
At best, that might explain why the Ohio Poll in its statistically weighing went more in the GOP’s favor, but that’s pretty weak tea to make such an assumption, especially against the contrary and overwhelming, and more relevant, evidence presented in early voting patterns.
I just received a memo from Aaron Pickrell of the Strickland-Brown campaign, and they reach the same conclusions:
“We strongly dispute a 7% Republican advantage in the turnout of this election. If you did an apples to apples comparison of the sample used in September and the percentage of the vote each candidate receives – it would be Kasich 48% to Strickland 46%.”
In other words, applying the raw data in today’s poll but using the same Likely Voter Model as they did in September would generate a result of 48% to 46%. A tightening of the race and a result that is more in line with other independent polls. (Today’s Ohio Poll is in Quinnipiac territory.)
“Yet, even those polling numbers belie what we are seeing in early voting. Despite the so-called "enthusiasm gap," Ohio Democrats are turning out in greater numbers than Republicans. The investments that we have made in recent years in the Ohio Democratic Party have resulted in the creation of one of the strongest state parties in the nation. While the Republicans have virtually no turnout operation, the Ohio Democratic Party has a truly statewide get-out-the-vote operation that is identifying and turning out our supporters weeks before the November 2nd election.”
“No one pretends this is going to be an easy election – but today’s Ohio Poll does not offer an accurate snapshot of the race. Governor Strickland is ahead by a couple of points in this race and with the superior early vote and GOTV operation – there is no question he will be successful on Election Day.”
In case you were wondering, the Strickland campaign did the math and figured out the change in the partisan breakdown:
This is based on a sample that is 49% Republican, 42% Democratic and 9% Independent. The poll conducted in September had a sample of 45% Republican, 42% Democratic and 12% Independent.
The Ohio Poll doubled the GOP’s gap between Democratic voters, although their poll shows no data to really support such an increase. Furthermore the poll assumes that the biggest partisan group in the electorate, enough to make up nearly half of it, is the partisan group that makes up the smallest amount of the registration voting population. Again, that’s not likely.