A few months ago, I wrote that, based on the so-called generic ballot (surveys that as, “If the elections for Congress were being held today, which party’s candidate would you vote for in your Congressional district?”) and some research by Bafumi, Erikson, and Wlezien, the Republican Party looked to be in good shape in 2010.
Recently, Harry Joe (who took a statistics course from Joe Bafumi at Dartmouth!) crunched the early poll numbers, does the adjustment based on the Bafumi/Erikson/Wlezien regression, and comes up with a forecast that the Republicans will win by 8 percentage points (that’s 54%-46%); that is, 1994 all over again, but this time without the benefit (yet) of any Newt Gingrich-like figure (although I’m sure there will be many proto-Gingriches around to claim the credit if this all happens). Here’s Harry Joe’s plot:
As John, Jamie, and I have discussed (see also here), the translation from votes to seats is not trivial. Still, if the Republicans really get 54% of the two-party vote or anything close to that, they’re in excellent shape.
Which would not mean that Obama is a lame duck by any means. Recall Reagan in 1982 and 1984.