It’s the season to start choosing presidential nominees, and people are talking about electability again. But Your Humble Blogger has come to believe that electability is bunk.
Back in 2003, I wrote up each of the candidates for my Party’s Presidential nomination. I specifically avoided writing about whether the candidates were electable or not. I wrote:
I do not want to talk at all about electability yet. I will, ultimately, look at that before I cast my vote; I don’t dismiss it as a valid topic for discussion. It’s not the only topic for discussion, though, and if you want to talk about it now, I strongly encourage you to find a site where that is being encouraged. I just don’t want to talk about it yet. I figure that if I don’t think about that part of things at all between now and Labor Day, it’ll still leave five months to think about it before anybody has to cast a vote in a primary or caucus, and that should be plenty of time.
Four years later, I wrote about electability again:
The Party has to decide (a) who would make a good President, (2) who would match up well with the other candidate or candidates, and (iii) who would move the Party in the direction we want to go (taking into account the effect of top of the ticket on the other candidates. Generally, I take electability into account. I do not want the next President to be from the Other Party. The fact that all of my Party’s candidates are far superior (in my estimate) to all of the Other Party’s candidates is not a coincidence. On the other hand, if I am feeling confident, I will likely prefer a candidate I judge to have a, oh, 56% chance of winning and extraordinary policies, instincts and abilities to a candidate that I judge to have a 59% chance of winning and quite good policies, instincts and abilities. But those percentages are impossible to judge with that kind of accuracy, aren’t they?
In fact, I now think those percentages are impossible to judge with any kind of accuracy at all. None. So now I do dismiss electability, and I will not take it in to account. And around the internet, people keep talking about whether particular candidates are electable or not, which are more electable than others. And I am tempted to comment every time in every place that electability is bunk, and keep doing it until everybody agrees with me. So to keep from doing that, I’ve decided to post it here, on my own Tohu Bohu where it belongs.
Political scientists stuff study this stuff and argue about it a lot, and there really is almost no evidence that it's possible for a party to increase their chances of winning the President by choosing the most ‘electable’ candidate. For one thing, the evidence is very strong that the actual candidates and campaigns make very little difference in the outcome, compared to (a) the state of the country, (2) the popularity of the incumbent (which is related to that) and (iii) how long the incumbent Party has held the office. Those things matter a lot to the outcome; everything else matters at the margins, and not predictably. In the end, almost everyone who would vote for any R will vote for whoever the R is, and almost everyone who would vote for the D will vote for whoever the D is, and of the people who want, most of them will vote for or against the incumbent based on what they think of him, and whoever the remaining few percent vote for (or whether they don't vote) isn't based on policy positions.
Specifically, the discussion in polisci is whether a candidate viewed as on the extremes does or does not lose perhaps two points in the outcome compared to one who is not viewed that way. Current thinking seems to be that it does not make even that much difference. But more importantly, being perceived as an extremist does not correlate with the candidate's actual policy views—I don't think it's possible to predict at this time whether Sen. Harris or for that matter Sen. Biden would be viewed as a ‘far left’ nominee by the time November 2020 comes around.
Finally (I hope) the primary system effectively makes it so candidates who aren't ‘electable’ don't become the nominee—by the time the nomination occurs the nominee will have got the votes of moderates and progressives, rural and urban and suburban voters, men and women, red state residents and blue state residents. Any candidate that is so bad at appealing widely to the populace as to prevent them winning in November will be unable to win over the long year leading up to it.
All of which means that you are much better off supporting whichever candidates you think would be good at doing the job, rather than the ones you think are more likely to win in November. And unless your job requires it, there's no particular benefit to you, the candidates, the Party, or the nation in choosing a single candidate at this point rather than supporting several candidates that you think would be good.
Tolerabimus quod tolerare debemus,