Tuesday, September 22, 2015

Winnowing the GOP Field with Jane Austen

Scott Walker has left the Republican presidential primaries: the first dropout who was once considered a major contender for the nomination. That, and the departure of Rick Perry, leaves us with only fourteen or fifteen candidates left. In fact, the real number is much smaller than that, because of an economic concept called the Pareto principle; there have never been sixteen choices, because the Pareto principle cuts the number down to a smaller number of practical options. But since I am a book nerd rather than a math nerd, I am going to illustrate this statistical idea with my old friend Jane Austen and her Pride and Prejudice.

In Pride and Prejudice, Mr and Mrs Bennet have imprudently had a family of five daughters and no sons. Since there's no son, Mr Bennet's estate is going to distant relatives when he dies, leaving his widow and five daughters in poverty. The five girls' only hope is to marry well. But since there are five of them, there is the real danger that they will crowd each other out so that none of them gets married. Other characters ask why the Bennets have allowed all five daughters out "into society," meaning the marriage market, at once instead of letting one daughter out at a time, so that Daughter Number Two wouldn't start going to balls until Daughter Number One was married. The Bennets take a laissez-faire approach which hopes that all five girls can find husbands; the neighbors fear a Tragedy of the Commons in which the glut of Bennet sisters on the marriage market keeps any of them from being married.

Neither scenario is correct. Three sisters get married, two do not, and the two who do not are never in even the remotest danger of a marriage proposal. The Bennett sisters offer three, rather than five, real choices for potential suitors. Two of the five sisters are eliminated from consideration by the Pareto principle, which says that any option which comes behind another in all criteria being considered is thereby eliminated from consideration: "Pareto dominated," as they like to say at 538. If I'm choosing a motel for the night, I might balance my choice between low cost on one hand and amenities like good cable or a free breakfast on the other. Maybe I'll go with Chain A, which costs more but provides HBO and a waffle buffet, or the much cheaper Chain B, with its basic cable and continental breakfast. But I am not even going to consider Chain C, which costs $20 more than Chain B but doesn't offer ESPN or breakfast. Why would I pay more not to get a Danish? Chain C is Pareto dominated, eliminated from the field of choices.

Now, the eldest of the five Bennet sisters, Jane, is universally agreed to be the best-looking. She is also easily the nicest of the five. Daughter Number Two, Elizabeth, is the second-prettiest (as we know from the behavior of a failed suitor who only considers looks), but not especially nice; she can have a sharp tongue. On the other hand, Elizabeth is by far the smartest. Meanwhile youngest sister, Lydia, is far and away the easiest of the five. (I tried to find another term like "affable" or "agreeable," but none of these are quite accurate. Lydia is distinguished from her four sisters because she's the one most likely to have sex before marriage.) And of course the sisters are all equal on some criteria, such as wealth and family background. A man who needs a rich dowry, or who can't bear the thought of Mrs Bennet as a mother-in-law, is going to rule all five sisters out of consideration.

An especially shallow suitor (and the novel does have one), would just rank all five sisters by physical attractiveness: Jane first, Elizabeth second, and Mary last. By that standard, Elizabeth would be Jane's closest rival. But most people looking for a spouse are at least a little smarter than that, and choose along more than one axis, so that Jane and Elizabeth are never actually in competition with each other for serious suitors. If you're looking for pretty and nice, Jane is the clear winner, and someone looking for someone like Jane is not going to consider Elizabeth at all. Elizabeth is less physically attractive AND more likely to take your inventory with other people there listening.

On the other hand, there are suitors who value intelligence, and who might prefer Elizabeth to her prettier sister, especially if they put less of a premium on beauty alone. Elizabeth is more than pretty enough to marry, and much, much smarter than most other young marriageable ladies in the novel, so she attracts her own suitors. But there's virtually no overlap between men who pursue Elizabeth and men who pursue Jane. If you're interested in beauty and sweetness, Jane Pareto dominates Elizabeth. But Elizabeth isn't universally dominated because she has something that her sister doesn't. If you're interested in beauty and brains, Jane isn't really in the running because, while she is not stupid, neither is she unusually bright. Jane and Elizabeth each have their own distinct group of suitors and occupy different parts of the decision space.

Likewise, a caddish suitor looking for someone to sleep with rather than to marry is going to be much more interested in Lydia. Jane and Elizabeth may be prettier and, in the abstract, sexier. But they don't crowd Lydia entirely out of the decision space because she offers something they don't. If you're looking for premarital sex, the fact that Jane is hotter doesn't make her a better choice. The sexier sister who won't sleep with you is not a better choice than the still-perfectly-sexy sister who will. So Lydia, Jane, and Elizabeth present three distinct alternatives, appealing to three different kinds of men.

But the other two sisters, Mary and Kitty, are eliminated from consideration because neither offers a genuine alternative to the other three sisters. Instead, they both come off as inferior imitations of one of those three. Mary, who is clearly the homeliest, has staked all her chips on showing off her brains. Her problem is that Elizabeth is still much, much smarter. Mary's attempts to seem smart are painfully laborious, all too clearly the product of ponderous study, while Elizabeth is quick as lightning. And, worse yet, Elizabeth is also much prettier than Mary. Poor Mary can't win, and doesn't. No suitor is going to pursue Mary while Elizabeth is available.

Likewise, Kitty is a paler imitation of Lydia, almost surely the second-easiest. (She clearly knows about some of Lydia's illicit romance and keeps that secret, whereas the other three sisters would almost certainly narc on Lydia immediately.) But second-easiest, in this case, means not as easy. Anybody interested in Kitty is going to be more interested in Lydia. So, like Mary, Kitty is Pareto dominated. She is one of the eliminated choices and has to live vicariously through Lydia's imprudent adventures.

Now, our crowd of Republican candidates likewise represents a number of significant alternatives, each with its own sector of the decision space, and a number of also-rans who are basically ruled out. The candidates are competing on different strengths, most obviously on their conservatism and their electability, but there are other characteristics that resonate with Republican primary voters; the exact list is up for debate. Performance of authority seems to be salient, so that Fiorina doing her best alpha-dog act at the second debate helped her enormously. And, alas, there is clearly a subset of GOP voters that is looking for the best racist dog-whistler.

Trump doesn't compete on electability at all. But he performs authority well, he traffics in various fringe beliefs that are current among some of the party base, and  he doesn't so much do the racist dog whistle as he calls his racist dogs at the top of his voice. If you were planning to run as a maverick outsider and pick up support with some subtle racial signaling, Trump has you beat on every level. He is more of an outsider than any first-term Senator or far-from-DC governor can claim to be. He is also more maverick-y than anyone else, being not merely a maverick but a bull in a china shop, untethered by any restraint or sense of prudence. You can't be more outrageous than Trump. And if you were hoping to pick up a few white-pride voters, Trump had you beat out of the gate when called Mexicans rapists in his announcement speech. Other "outsider" candidates looking for that particular slice of white support are Pareto dominated by Trump.

There are other examples. Huckabee and Santorum are both running as not-very-electable champions of Christian conservatives. But Huckabee is both more appealing to Christian conservatives and more electable than Santorum is (meaning not so very electable, but not as hopeless as Santorum). This leaves Santorum no air to breathe at all. If Bobby Jindal was hoping to be the non-white hardcore conservative candidate, Ben Carson (even more hardcore and less white) has him beat. If Jindal was hoping to be the Chance to Reach New Voters, Rubio has him beat (because the Latino vote is much bigger than the South Asian vote).

Where we are really not seeing much competition is in the Electability sweepstakes, with the candidates whose basic appeal is that they can win in the general election. Right now the primary voters don't seem interested in electability at all; the most recent polls show Trump, Carson, and Fiorina, three candidates who haven't won a single election between them, with more than half the combined support of GOP voters. As I've argued before, the most surprising thing is not how well Trump is doing but how poorly Jeb Bush, the presumptive electable alternative, is doing. And no one has yet emerged as the main electable candidate, the way Mitt Romney emerged last time around. The 2012 Republican primaries featured one main Electable Option, Romney, and a bunch of competitors for the role of Uncompromising Conservative. This time we have a clear Uncompromising Outsider, pretty much safe from challenge on his native turf, and no solid Electable Mainstream Option. In 2012, no Republican could hold onto the Lydia Bennet role for more than a week or two. This time, no Republican has seized the Elizabeth Bennet role for even a week.

The Democratic primary, on the other hand, already has a pretty clear and recognizable shape. There's a party-establishment favorite, Clinton, whose main appeal is her electability, and a dark-horse challenger, Sanders, whose main appeal is his ideological closeness to the base. Then you have a couple of also-rans like Martin O'Malley or Jim Webb,  Pareto dominated by Clinton because they are at once less liberal and less electable than she is. And you have the non-candidate, Biden, sitting out because he would be dominated by Hillary if he got in now (he's slightly less electable and equally mainstream), but that would flip around if Hillary were suddenly undone by a major scandal. (If Hillary turned out to be, say, selling weapons to Iran and using the money to fund the Nicaraguan contras, to choose a purely hypothetical example, she would suddenly be less electable than Biden.)

Where I would expect to see movement in the Republican campaign is on the mainstream, electable side. Trump cannot be beaten at Trump's game. Candidates like Cruz or Rand Paul are going nowhere this year. Neither is Carson, really, even if he's outpolling some more likely contenders right now. But someone could conceivably take over the mainstream/establishment/viable-in-a-general-election role that Jeb Bush hasn't managed to keep or win. Yes, some of the other mainstream/moderate candidates are hopeless. No primary voter would vote for George Pataki, who is both to the left of Jeb Bush and less viable in the general election than Jeb Bush, when they could simply vote for Jeb Bush. The same basically goes for Lindsay Graham. But while Jeb himself can't manage to break ten percent in the polls, one of the other candidates in his general category could overtake him. People like Rubio and Kasich, or even Christie despite his serious problems, would be smart to stick around.  They aren't really competing against all 13 of the other candidates. They aren't even competing against Trump. Not yet. They're competing against Jeb Bush for the position of Reasonable Party Dad that he mysteriously can't nail down. Then whoever manages to establish himself in that role can go toe-to-toe with Trump under the "Yes, but I can beat Hillary" banner.

And if none of the "electable" establishment candidates emerges as a major contender, then we will be on new and unexpected ground.

cross-posted from, and all comments welcome at, Dagblog


Yastreblyansky said...

Beautiful! Though I guess a little dated. Cruz is turning out to be Caroline Bingley, right? Who inevitably wins if, and only if, Lizzie doesn't exist.

Doctor Cleveland said...

Hi, all coming from the annual Jon Swift Memorial list. Welcome!

The main comments thread for this post is here: http://dagblog.com/politics/winnowing-gop-field-jane-austen-19911#comments

and my comments filter on blogger blocks nearly everything once the post is past a certain age. But please feel free to join the existing thread.

And yes, Yastreblyansky, Ted Cruz does have some Caroline Bingley vibes going on, in that he should be a serious contender but he's just too damned hard to like.