The remaining thirteen seats are Republican opportunities ‘that will not fall easily.’ Add in the non-Freshman Democrats that the GOP may target (which now appears to include just a handful of seats), and the potential is there again to flip control of the House. That’s provided that Republicans hold most or all of the vulnerable open seats they have.

It would be foolish to predict a great Republican year based on the political climate today, but Rothenberg provides a helpful reality check for those inclined to the opposite extreme. If the cards fall the right way, it’s entirely possible a Republican will be sworn in as Speaker in January, 2009. ~Brian Faughnan

So if a dozen improbable things happen, something even more improbable might very well happen.   

There are some districts, particularly TX-22 (DeLay’s district) and FL-16 (Foley’s district), that will be difficult for Democrats to hold, assuming that Republicans turn out for their candidates.  That’s one area where the GOP is going to run into a lot of problems.  Democratic turnout in a presidential year is typically higher than it is at midterm and off-year elections anyway, and we are already seeing gaps opening up in party ID, fundraising and candidate recruitment.  If Republican voters are as demoralised as they seem to be, turnout may also be unusually low for Republicans, which could combine with an energised Democratic base to create more gains for the Dems on top of holding what they already have.  (For instance, NM-01 is a realistic pick-up for the Dems.)  Depending on the GOP ticket, the base’s morale may get worse rather than better.  A major third party challenge from the right could actually help the GOP in Congressional races by bringing conservatives to the polls who might otherwise have stayed home, but such a challenge is unlikely to materialise. 

Some Democrats have the fear, and I think it is probably an over-hyped fear, that a Clinton nomination would imperil closely-split districts and jeopardise the majority in the House.  There is a bizarre idea out there that a winning presidential candidate can have a kind of reverse coattail effect in every “purple” and “red” state.  This assumes that there will be a lot of split-ballot voters in “purple” states who elect Clinton but vote out the Democrats in the House, while there are few or no split-ballot voters in the “red” states who vote for the GOP candidate and select Democrats for Congress.  This is probably not how it will happen. 

The logic of this seems to be: Democratic presidential victory is very likely, in part because of the deep dissatisfaction with the GOP in many formerly red, now purple, states, but a particular Democratic nominee will actually help the GOP in these same states where they are becoming less popular (e.g., Ohio, Pennsylvania, etc.).  I suppose if enough people believed Clinton to be the left-wing gorgon conservatives see when they look at her, that might inspire them to vote for divided government and switch control of House again.  However, I know of no instance when a party won the presidential election and lost control of the House on Election Day.  In fact, I don’t know of any instance when a party won the White House and even suffered a net loss of seats in the House.  (Actually, as shown in the comments, there were six seven elections in the twentieth century where the presidential winner’s party lost seats, so I was wrong in assuming that it hadn’t happened.)   The “coattails” phenomenon may have become much weaker in recent cycles, but it seems implausible that the GOP can gain much ground in the House next year unless it wins or at least runs extremely competitively in the presidential election.  All signs indicate that this will not happen, which makes predictions of a GOP comeback in the House even more far-fetched. 

Update: So my claims about there being no cases of the winner’s party losing seats were quite wrong.  What about the exceptional cases?  Does 2008 seem likely to be another exception?  1908 and 1988 appear to be examples of voter fatigue with the ruling party that had been in the White House for eight years or more, while 1960 and 2000 stand out for being fairly close presidential elections and in one case the declared winner received less of the popular vote.  1992 was complicated by Perot’s run, but the combination of Bush and many Perot voters would help explain the GOP gains in that year.  What the Republicans have to hope happens is that next year will be like 1960, in which they may narrowly lose the White House but come storming back in the House after a midterm debacle.  However, this scenario seems unlikely because of the nature of next year’s election.  Wartime or  post-war elections (1920, 1952 and 1968 are the examples I have in mind) coming at the end of multiple terms of the same party in power tend to result in big gains for the other party in the House, even if the other party has already made gains in the previous midterm elections (as happened in 1918, 1950 and 1966).  So I was badly wrong about that initial claim, but I think the argument I am advancing here still makes sense.

Advertisement