Special elections’ magic math

Humans are a meaning-seeking, storytelling species. When we saw stars and planets move across the sky, unwilling to believe it was just rock and burning gas transiting through space in some mathematically ordained pattern, we looked for the meaning in their motion and told tales about gods and heroes.

So it is, more prosaically, with special elections. We scour every tea leaf, ferreting out the meaning we insist is hidden inside, believing that somehow those results will presage the next general election. By the time you read this, and know the outcomes of the specials in New York and Nevada, barrels of ink will have been spilled speculating on the import of those results for November 2012. Just as predictably, those commentators will be taken to task by another set of analysts who, decrying the naiveté of the meaning-seekers, will see only random happenings, the failures and success of individual campaigns, disconnected from any larger trends.

Where does the truth lie? Do special elections tell us anything meaningful about what will happen in November? The answer? A definitive “sometimes, sort of.”

In most specials, as in most House elections generally, the incumbent party holds onto the seat. We should not expect those elections to foretell the future, and they don’t. According to data collected by professors from the University of Dallas (which I’ve updated), since 1900, in half the cycles when Republicans won more specials than Democrats, Democrats actually gained seats in the next general, losing seats in the other half. In those cycles when Democrats won more specials than the GOP, Democrats lost seats in 44 percent of the generals and won seats in the rest. 2009-10 specials illustrate the point: Democrats won a majority of the specials but were wiped out in November. 

In short, trying to figure out what will happen in November 2012 by looking at how many specials each party won is no better than guessing, a conclusion confirmed by detailed statistical analysis.

Examine these data from a different perspective, however, and it gets more interesting. Instead of considering all special elections, factor in only the more unusual contests that result in a change of party control. More than two-thirds of the time, when Democrats suffered a net loss of seats in specials — that is, in cycles where the GOP took more seats away from Democrats in specials than Democrats took away from Republicans — Democrats went on to lose seats in the general election. In 82 percent of cycles when Democrats picked up net seats in specials, the party went on to gain seats in the general.

Other things being equal, the statistical analysis suggests that for every net seat change during specials, six seats will change hands in the general. 

There is less certainty here than meets the eye, however. The statistical noise associated with the six-seat shift is fairly hefty. 2009-10 again illustrates the point. Democrats lost a net of just one seat in specials (Democrats took one away from the GOP, while Republicans gained two) leading up to 2010, but of course the general brought massive losses.

Where does that leave us for 2012? As of this writing, Democrats have gained one seat in a special election: Kathy Hochul in New York. If Republican Mark AmodeiMark Eugene AmodeiGOP revolts multiply against retiring Ryan GOP lawmakers back discharge petition to force immigration votes McCarthy faces obstacles in Speaker bid MORE were to have won in Nevada it would foreshadow nothing, as that is already a GOP seat. If David Weprin has won, based on the specials now on the calendar, it will be impossible for the GOP to come out with net gains. If Weprin has lost, it’s a net change of zero, with Oregon’s special in January holding the key to predicting the future — plus or minus a lot.

Mellman is president of The Mellman Group and has worked for Democratic candidates and causes since 1982. Current clients include the Majority Leader of the Senate and the Democratic Whip in the House.