Are Republicans Throwing Away Their House Majority?


Call it the Todd Akin effect: nominating an extreme candidate can halve a party’s chance of holding on to a House seat, according to a brand-spanking new political science paper. It’s a finding that has important implications for the future of the increasingly-polarized Republican Party and the often-contentious debate over gerrymandering.

Previous research had demonstrated that incumbents who went outside the mainstream were punished at the ballot box, but measuring the effect of nominating an extreme candidate in a primary who might go down to defeat is a bit trickier. Andrew Hall, a graduate student in Government at Harvard, developed a measure based on who donates to a given candidate. Basically, if donors to a particular primary candidate tend to donate to other extreme candidates (as verified by an independent measure of ideology), then it’s very likely that the candidate in question is also to the far right or left wing of their party.

This measure in Hand, Hall then set about figuring out what happened when they won. After performing some tests to isolate out the effects of candidate ideology, Hall found an enormous effect: when an extreme candidate wins a House primary in either party, there’s “a 11–13 percentage-point decrease in the party’s share of the general-election vote, and a 38–49% decrease in its probability of victory.” The more extreme the candidate, the more towards the higher end of these estimates you’d expect the extremism penalty to be.

Once a party has taken a seat because of their opponents’ extremism, it’s gone for a long, long time. Incumbency confers huge advantages on House members, which means that the extremism penalty isn’t just a one-election thing according to Hall’s data: “the nomination of an extremist today continues to cause an equally large decrease in the party’s expected probability of victory and vote share even four terms, or eight years, later — the farthest downstream that redistricting allows us to examine.”


Halving a party’s chance to win a House seat for years is an insanely large effect and, unsurprisingly, Hall finds that it has implications for what happens in Congress down the line. Normally, you’d expect having more ideological candidates win primaries would help a party pass its agenda in Congress, as, if they win, they’d be more likely to vote consistently for the Republican or Democratic line. But that’s not what happened: “Nominating the extremist causes a marked shift in the opposite direction on roll-call voting,” Hall writes. “Randomly getting an extremist in the Republican primary causes the district’s subsequent DW-NOMINATE score [a common measure of ideology] to move left by 0.30–0.55 points.” That, of course, is because Republicans are hugely less likely to hold a seat they nominate an extremist for. Turns out it’s hard to get someone from the other party to vote for your issues.

Hall’s findings are preliminary — they’re in a working paper, and hence haven’t been peer-reviewed. But if he’s right, there are two interesting implications for American politics.

First, they point to yet another way that the Republican Party is playing Russian Roulette with its political future. It’s well-established at this point that, while both parties have gotten more ideological in recent years, the Republicans have gone way further off the deep end than Democrats (here’s the newest piece of evidence on that point). This “asymmetric polarization,” an aftershock of the post-civil rights political realignment of the South, means that Republicans are significantly more likely to nominate “extreme” candidates than Democrats are. Hall’s research suggests that this could ultimately erode the Republican advantage in the House.

Now, Republicans have other, off-setting advantages in the House. Most notably, Democratic groups have a tendency to cluster tightly in cities while Republicans are spread out, which means that Republicans are efficiently distributed to control a majority of House seats. The extremism penalty is unlikely to offset the GOP’s geographic advantage on its own, but combined with unfavorable long-term demographic trends, it could play a role in weakening the GOP’s stranglehold on Congress’ lower body.

But that’s assuming at least reasonably fair districts. Hall’s research finds that there isn’t much of an extremism penalty in “safe” Congressional districts, which makes sense: in a heavily Republican district, it’s unlikely that there are enough voters willing to defect to Team Democrat in the event of an extreme Republican nominee to tip the election in the Democrat’s favor.


Which brings us to the second point, gerrymandering. If, as Ian Millhiser has argued in these pages, gerrymandering creates a glut of “safe” Republican districts, then Hall’s extremism penalty won’t end up being much of a penalty at all.

There’s huge disagreement on this topic. Nate Cohn believes that there might be more “safe” Republican seats in a world without gerrymandering (at the expense of mildly Republican ones), and a body of political science research suggests that gerrymandering has at best a moderate enabling effect on GOP radicalism. The gerrymandering and extremism argument may be tough to parse, but Hall’s research suggest it’s critically important to understanding the future of the Republican Party and, by extension, American government.