Thursday, March 3, 2011

The Gaffe Factor

Matt Yglesias reminds everyone that gaffe-centric press coverage of elections is foolish, because gaffes don't matter very much.

He's right!

I'll break it down a bit more:

In general elections, most people vote based on party, with a bit of economic circumstances and (the overlapping factor of) approval for the incumbent president thrown in. Everything else is relatively minor. At best, a gaffe might have a very short-term effect, so one that happens in the days before an election might make a very small difference. Anything earlier -- including during presidential debates -- won't.

In nomination politics, during the (current) invisible primary phase, what's mainly happening is that party leaders of various types are assessing whether the candidates will support their issue demands and whether the candidates are apt to win. Gaffes can matter indirectly to the second of those; a candidate who develops a reputation for being gaffe-prone might lose support from those party elites who are focused on winning.

During the primaries and caucuses, gaffes may well be important. Voters are seeking to differentiate between very similar candidates, and someone who says something stupid during the week before a primary could easily lose support that week (although it's unlikely to be a long-term effect; a gaffe now wouldn't hurt next February). However, remember that the primaries and caucuses phase of the nomination contest is heavily structured by what's happening before the voters get involved...George W. Bush and Al Gore had basically wrapped things up before Iowa in 2000, for example, so either could have easily withstood even a significant gaffe -- and it wouldn't have mattered what the other candidates said. And (via Nyhan) even when the circumstances seem ideal for a gaffe to matter, it's still rare for it to make much difference, as two John Sides posts about 2008 show.

So that's it. Basically, it's just not that big a deal.


  1. Unfotunately though, its never possible to tell what matters until after the fact (Macaca, Katrina, Dukaukis in the tank, Carter and the rabbit, "I have sinned in my heart", etc.) Some things seem to unexplicably get traction while things that do matter don't. How can mere mortals tell them apart?

  2. On the other hand, in a close election, you wouldn't want to say something in a pre-election debate like, oh, "There is no Soviet domination of Eastern Europe and there never will be under a Ford administration." And you certainly wouldn't want to answer a follow-up by reiterating the point more emphatically. Especially if you've been gaining until then and are on the verge of a come-from-behind victory.

    Of course, one could argue over whether this was a "gaffe" or a revelation of the difficulties of selling detente-ist policies in competitive election campaigns where it's easy to get painted as an appeaser. Ford and Carter both seemed to think it was a gaffe, though, and possibly one on which the '76 election turned:

  3. But the point is that the Ford thing, Macaca, and all the rest of those things (1) probably didn't matter much if at all, and (2) seem to us to have mattered because of the way press coverage works, and (3) may seem to the participants to matter a lot, because they have a bias in favor of believing that such things matter.

  4. But are they not a legitimate thing to judge presidential candidates on?

    Regardless of skills, a foreign service officer will never be appointed ambassador if they have a case of foot-in-mouth disease. (The political appointees are a far different story) We also don't want a president who walks into a room with foreign leaders and insults them, even if by mistake. This is a complete non-issue for the House, and an almost non-issue for the Senate. But Presidents? Foreign relations is a big part of their job.

    So, while gaffes might not affect elections in any systematic way, maybe some types of them should?

  5. I question whether Ford's comment on Eastern Europe "probably didn't matter much at all." That was the first election I followed closely (also the first in which I voted), and I remember vividly how Carter's 30-point lead kept shrinking until Ford was clearly in striking distance. Then came that debate. Yes, the way the press covered it mattered, but that's just another way of saying that a gaffe actually can make the difference (by changing the "media narrative" or what-have-you). If Ford had climbed just a bit higher, flipping 6,000 votes in Ohio and 8,000 in Mississippi, he wins the EC, so hitting a speed bump like that late in the race could have been the difference.

  6. Jeff,

    I'm pretty confident that it turned out that the debate didn't matter, although I don't remember any specific study to cite, so I could be wrong. A lot of the "momentum" stuff was probably just people returning to their party; IOW it wasn't predictive of future movement. It was a month before the election (Oct 6) with another, intervening debate still to come...just on the surface of things, it seems unlikely to me.

  7. OK, duly noted, thanks. In general, I agree that gaffes are way over-covered and over-interpreted.


Note: Only a member of this blog may post a comment.