There’s been recent speculation that 2012 will be a double or even “triple-flip” election. In this scenario control of the Presidency, the House, and the Senate would all (or two out of three) change party hands in opposite directions, giving us President Romney and Speaker Pelosi (and maybe Senate Majority Leader McConnell). The scenario presupposes an anti-incumbent election in which disgust with government and the economy is so acute that voters throw the bums out indiscriminately.
Alan Abramowitz, demolishing the idea, uses almost exactly two thousand words. But in confirmation of the cliché, I’m happy to ignore the words and simply reproduce instead his two perfect scatterplots (data sources left out of the graphics below, but in original).
For the House, here are Democratic and Republican incumbent losses, by year:
And for the Senate:
We are done here. Time to move on.
These graphs say less than they seem to, even when you remember that history is bunk. If you trust historical trends, they merely say that only one party can lose a large number of incumbents in any given election. Fine. But the qualifier is “incumbents.” The Senate will have a number of open seats (CT, NE, TX, AZ at least), and there are a lot more Democrats up than Republicans-a legacy of the thumping of ’06. And I don’t know the incumbent situation in the House, although IIRC first-term Representatives lose far more often than those who have survived their first re-election.
It is easy to imagine the Senate going Republican, with the House going Democratic. What I have a harder time imagining is that the House could go Democratic with Obama losing. So I don’t buy a triple flip. But a double flip seems possible, if unlikely. My own guesses, in order of likeliness (ymmv):
P-S-H
D-D-R (status quo) (most likely)
D-R-R (single flip) (fairly likely)
D-R-D (double flip) (not unlikely)
R-R-R (double flip) (not unlikely)
R-D-R (single flip) (unlikely)
R-R-D (triple flip) (very unlikely)
To complete the possibilities, you should consider the clean sweeps: R-R-R and D-D-D. Neither are at all likely, but they are possible. Note that D-D-D still leads to stalemate unless the Bemocrats have the guts to kill the filibuster.
Quite so. I had R-R-R, but omitted D-D-D. (I also omitted R-D-D.) To amend my comment appropriately:
P-S-H
D-D-R (status quo) (most likely)
D-R-R (single flip) (fairly likely)
D-R-D (double flip) (not unlikely)
D-D-D (single flip) (not unlikely)
R-R-R (double flip) (not unlikely)
R-D-R (single flip) (unlikely)
R-D-D (double flip) (very unlikely)
R-R-D (triple flip) (very unlikely)
I dunno, those graphs seem to say that in each house only one party looses in each election. Seems plausible that, in that case, both the House and the Senate could flip. In is entirely possible for voters to blame the Republicans for failures of the House and Democrats for failures of the Senate and for both to flip and have the data points fit onto those purdy pictures without looking out of place. What they do show is that it is historically untrue that voters wipe out the incumbents of both parties in either house individually because of general dissatisfaction.
What´s the correlation between the changes in the House and Senate?
Eyeballing, it´s quite strong but not unitary.