Let us say that the modern era of American politics began in 1965.
Since then, there have been seven midterm elections coming after the election of a new President. Let's consider them.
1966 -- Johnson (D) takes a bath as conservatives rebound from the disaster of 1964. Loses 3 seats in the Senate and 48 in the House.
1970 -- Nixon (R) does okay, gaining a seat in the Senate and losing 12 in the House.
1978 -- Carter (D) loses 3 Senate and 15 House
1982 -- Reagan (R) loses 26 House but none in the Senate
1990 -- Bush (R) very modest losses, one Senate and 8 House
1994 -- Clinton (D) gets brutally hammered, 8 Senate and 54 (!) House.
2002 -- Bush (R) becomes the first President in almost 70 years to make gains in a midterm, picking up 2 Senate and 8 House seats.
Let us say that a Senate seat is worth four points and a House seat, one. (Yes, yes. And a real methodology would include governors and state legislatures as well. Toy model, bear with me.) How do they rank?
Bush II (R) +16
Nixon (R) -8
Bush I (R) -12
Reagan (R) -26
Carter (D) -27
Johnson (D) -60
Clinton (D) -86
Average Republican losses are -7.5, or a Senator and a handful of Representatives. Average Dem losses are -64.3. The best Dem showing is about the same as the worst Republican. Two of the three Dems suffered losses that can fairly be called disastrous, that would impose significant political constraints on the rest of the terms, while three of the four Republicans did just fine.
(By now Carlos' fingers are twitching. Toy model and yes-I-know.) Obviously these are very small sample sizes. And if I had chosen to include the midterms of 1974 (disaster for the Republicans) or 1962 (very modest losses for the Dems), the picture would blur. Still... if Obama were to have a decent midterm, he'd be the first Democratic President since JFK to do so.
Apropos of absolutely nothing, really. Back to writing now.
Doug ... Randy read this and took it seriously.
Randy, when Doug writes "apropos of absolutely nothing," he means it!
Posted by: Noel Maurer | April 30, 2009 at 11:30 PM
Johnson's losses in 1966 were, as you note, attributable in large part to the Republican rebound from the 1964 debacle. They weren't a herald of any great realignment in political power - note that Nixon recaptured the Presidency for the Republicans two years later with only 1% more of the popular vote than Humphrey (though Wallace's strong third-party candidacy complicated things). Clinton's even bigger losses in 1994 also were thought to be the start of a huge realignment, but ultimately didn't mean much.
Posted by: Peter | May 01, 2009 at 08:08 AM
Now, let me break your toy. Though we are fond of you, Uncle Douglas.
First; '94, and prolly most of the Republican gains are about the post-Civil Rights Act realignment; after Clinton, there really were no more Dixiecrats to toss--and most had hung on because post '64, you could vote split ticket between your Congressional Delegation and the Presidency.
Also, in the context of the model, Obama's gains from the 2010 midterms probably won't be all that impressive; more impressive than most Democrats, certainly, but there's really nowhere for us to go in the House (Dems pick up Joseph Cao before the midterms, btw, but we could lose, like, 5-10 seats). The problem isn't a failure of will so much as a lack of viable seats to grab. The Party of Raygun Ron is down to its most durably reactionary or fundamentalist seats (Michelle Bachmann!) which means that some of the moderate districts we grabbed in 2006 and 2008 will go back the other way. But it's the House.
OTOH, we stand a really good chance of taking the Republican Party to the cleaners in the Senate. (picking up Missouri, Ohio, New Hampshire, North Carolina, with wildcards at Florida, Arizona, and Kentucky). There's some tough defense in IL (hoping now that Burris understands he cannot run for reelection) and CT, but even so, there's a good chance that the GOP in the Senate will face the squalid, ugly messes dumped at its doorstep by its House counterpart.
Posted by: Luke | May 01, 2009 at 11:25 AM
...playing Devil's Advocate here: why were most of the Dem losses from the Civil Rights realignment concentrated in that one cycle? Why '94 and not '90, '92 or '96?
Midterms are strange. And there do seem to be patterns to them, depending on who's in power and (especially) whether it's first term or second.
Doug M.
Posted by: Doug M. | May 01, 2009 at 12:26 PM
Doug, you don't have to ask why 1994 was a key election. I know that you know the answer.
Which doesn't change the fact that Luke is right.
Right now, I wouldn't bet on Democratic losses in the House, not after the recent by-election in New York. But a lot can happen in a year. As for the Senate, well, Doug, what do you think?
Posted by: Noel Maurer | May 01, 2009 at 06:12 PM
Oh, Douglas, stop begging the question; retirements, blah blah old people. Oh, and then Richard Shelby and Ben Nighthorse Campbell flipped--did you count that?
Noel; the NY-20 may have a GOP registration advantage, but a few hundred votes on the absentee balloting operation is hardly a decisive verdict, given the way North East Republicans are trending (fleeing the party).
Tough races for Dems: AL-2, ID-1, LA-2 MD-1, MS-1, and the NY-20.
We could lose one or two. But the DCCC has lots more cash on hand than the RCCC, so yeah. But there's nowhere else in the House to make inroads.
Posted by: Luke | May 01, 2009 at 07:24 PM