November 9, 2009
This is an op-ed JK and I wrote after the 2009 elections. Unfortunately, we were slow off the blocks and by the time we got turned down by the NYT (they demand three days of exclusive consideration and took all three!) even our local op-ed pages were done chewing on the election results. I thought it was a good piece, which is why we tried to get it into the old gray lady, but in retrospect I wish I had withdrawn it from the Times after 24 hours and submitted it elsewhere. Lesson learned. In any case, I still think it's a good piece so decided to post it on my blog. Of course, I also put a reminder in my calendar for the first week of November, 2013, when I will dust it off, plug in current numbers and try again!
Reading it this week, after the headlines have screamed about how Dorgan and Dodd's retirements herald doom for the Democrats, I have some additional thoughts which I may or may not get around to posting. Stay tuned, readers. All one of you. (Hi, Mom!)
Republican victories in gubernatorial elections held in New Jersey and Virginia this week have been widely interpreted as rebukes to President Obama and warning signals for Democrats to expect big trouble in 2010 if things continue along the current path. This interpretation likely overstates the case, even without considering the somewhat contradictory information offered from the Democratic victory in New York’s twenty-third congressional district.
First, the history of off-year elections in Virginia and New Jersey suggests they are not bellwethers of the midterm congressional races. Second, the information already available about 2010 -- a midterm election where the president’s party holds a majority in Congress under conditions of economic stress -- is already far more significant than anything gleaned from Tuesday’s outcome. The idea that the president or congressional leaders should now stop and rethink everything because the statehouses in Richmond and Trenton will have new occupants is fanciful.
Until 1989, results of gubernatorial elections in Virginia and New Jersey were mutually independent. Democratic or Republican success in one state was uncorrelated with party performance in the other. Since then, the states’ respective gubernatorial votes have been closely aligned, suggesting that national conditions do exert a significant influence on election results in Virginia and New Jersey. In fact, the party of the sitting president has lost both gubernatorial contests in all six election cycles. The average penalty for bearing the incumbent president’s party label has been about ten percent of the two-party vote in both states.
Of course, the president’s party typically pays a penalty in midterm congressional elections as well. This is the “decline” half of the ‘surge-and-decline’ theory posited by political scientists. In presidential election years there is typically a surge of support for the party of winning presidential candidate. Some congressional candidates who would not otherwise have won ride the president’s coattails into office. Two years later, however, these once-lucky candidates lose their re-election bids. The stronger the surge, the theory goes, the steeper the decline.
That trend, once iron-clad, has faded in recent years (Democrats actually gained seats in the middle of Bill Clinton’s second term, as did Republicans in the middle of George W. Bush’s first) but one would still expect that in 2010 the Democrats will lose some of the 21 seats they picked up with Obama’s victory, particularly in light of the state of the economy. So do this week’s results in Virginia and New Jersey tell us anything we don’t already know about the Democrats’ prospects in 2010? Actually, no. Since 1989, when Virginia and New Jersey gubernatorial results started to track each other, the average decline in House seats held by the president’s party has been about 15 House seats, with considerable variation over the five cycles. That variation in seat loss (or gain, as in 1998 and 2002) has been completely uncorrelated with the gubernatorial results from the previous year.
Consider two examples. During George H.W. Bush’s term, Democrat Jim Florio won big in New Jersey, but the following year Democrats picked up only eight House seats. In 1997, the landslide was in Virginia, as Republican Jim Gilmour trounced his opponent by thirteen points. His party subsequently lost seats in Congress, breaking a 52-year streak in which the out party had gained House seats in every midterm election. Of course, one can point to idiosyncrasies to explain both of these examples, but that is exactly the point. While voters in Virginia and New Jersey have consistently chosen governors from the party locked out of the White House, results in those states have also reflected current local conditions. The same is true for the congressional elections held one year later.
All the crowing, teeth-gnashing and dissection that followed Tuesday’s election is understandable; the results provide terrific grist for the media mill. But consider what we already knew on Monday: The party of the president always fares relatively poorly in off-year gubernatorial elections. Virginia and New Jersey, in particular, always elect governors from the out party. The president’s party usually loses seats in a midterm election. So it is not much of a news flash that the road ahead for Democratic congressional candidates is rough. This is a dog-bites-man headline that provides no new information or reason for revising strategy in the White House or on Capitol Hill. Rather than serving as a bellwether, the results from Virginia and New Jersey actually suggest that the current political dynamics look a lot like those of previous years. The traditional decline in support for an incumbent president is carrying over to other candidates bearing his party banner. Recently elected Democrats will likely struggle to hold office -- particularly those in marginal districts. And the pundits will breathlessly declare the dawn of new political era.
