Texas Rep. Steve Stockman has announced that he will challenge Texas Sen. John Cornyn in the primary. He plans to attack Cornyn from the right, saying:
We are extremely disappointed in the way he treated his fellow congressmen and broke the 11th commandment and undermined Ted Cruz’s fight to stop Obamacare. And now, it looks like Cruz was right and Cornyn was wrong. He sided with the president, essentially, in making sure Obamacare became law while Cruz did everything possible to stop it.
It doesn’t seem like the most auspicious moment for Stockman. He has only $32,000 in campaign cash and some lingering questions about his personal finances. And it’s not as if Cornyn is some RINO. According to Simon Jackman’s estimates, Cornyn is more conservative than almost 75 percent of his fellow GOP senators.
The conventional wisdom is that running more ideologically extreme candidates in a primary is risky. It’s not that they all lose — see, for example, the aforementioned Sen. Cruz, who bested a relative moderate, David Dewhurst, in the 2012 Senate primary — but that they’re more likely to lose (see, for example, not-quite-senators Todd Akin, Richard Mourdock, Sharron Angle, Christine O’Donnell, etc.). New research by Harvard’s Andrew Hall confirms this.
Hall identified more than 500 U.S. House races between 1980-2010 in which a more moderate candidate ran against a more ideologically extreme candidate. He then compared candidates in “coin flip” primary elections — those where one candidate beat the other by only a tiny fraction. The advantage of studying such close elections is that the winning and losing primary candidates are essentially similar in other key respects. The only difference is that some won by a nose and some lost by a nose. The key question is what happens in the general election when the more extreme candidate wins by a nose, as opposed to losing by a nose.
Hall found that when the extreme candidate won the primary, that candidate won about 11 points less in the general election. And as the ideological gap between the moderate and extremist candidates grew, so did the penalty for nominating the more extreme candidate. Even more remarkably, the electoral penalty appeared to persist for the rest of the redistricting cycle — as you might expect if a party nominated an extremist candidate, lost the general, and then had to live with an incumbent of the opposite party for years. As a consequence, there was a shift in roll-call voting too: these districts were more likely be represented by someone who tended to vote in exactly the opposite direction as the party who nominated the extremist wanted. This is how nominating an extremist can backfire.
But maybe none of this matters. After all, Texas is Texas. Pretty much any Republican has a very good chance of winning the general election in 2014. Indeed, Hall finds that this backfire effect on roll-call voting is more pronounced in competitive districts than safe districts.
Still, Hall’s research raises some red flags worth considering. The issue for the GOP isn’t so much the 2014 Texas Senate race. The issue is that, in general, the party would be better off — that is, it would control more seats and be better-positioned to steer policy — if it could discourage primary challengers in races where negative consequences are more likely. And Stockman’s example — particularly if successful — may only reinforce the desire of other conservatives in the party to mount similar challenges. When those challenges happen in states or districts that aren’t quite as red as Texas, the party may suffer, just as it has in Nevada, Delaware, Indiana, and Missouri.