Update on Gallup numbers
While writing my explanation, Gallup released the final poll before the election. They get 55-40 as their final result, assuming a 45% turnout model. So, even when they boost turnout by 13% over the max for the last 40 years, they still get a massive Republican lead that is twice what it was in 1994. By the way, the most Gallup has ever been off is 4.5% in 1998. If you assume Gallup is off by 4.5% in this election, the result would still be the all-time high for Republicans.
The results are from Gallup's Oct. 28-31 survey of 1,539 likely voters. It finds 52% to 55% of likely voters preferring the Republican candidate and 40% to 42% for the Democratic candidate on the national generic ballot -- depending on turnout assumptions. Gallup's analysis of several indicators of voter turnout from the weekend poll suggests turnout will be slightly higher than in recent years, at 45%.
One final note: I put so much weight on Gallup's results for three reasons: 1) historical record, 2) methodology, and 3) because it matches what I see in the congressional and senatorial polls.
Jay Cost is coming down on your side as well (http://www.weeklystandard.c...)
Posted by yara @ 11/01/10 06:07 AM
No flames or impolite behavior. HTML will be stripped. URLs will be transformed into hyperlinks.
Comments must be approved before being published.