Tuesday, February 23, 2016

Trump to Political Science: Drop Dead, Episode IV: A New Hope

It is a period of civil war.  Rebel candidates, striking from a not-so-hidden base (seriously, the guy writes his name in 'yuge' letters, even on the buildings he didn't build), have won their first two victories against the evil party establishment...

I promise, I'm going somewhere with this.

This is part four in a series in which I describe the many ways that Donald Trump is embarrassing the discipline of political science.  I have defined Type B Trump denialism as a refusal to admit that Trump's unexpected strength as a candidate shows underlying weaknesses in political science models of nomination politics.  Today, the "spatial model."  See?  Space!  It connects!

Fine.  Have my lunch money.

Back to business now.  The idea of the spatial model is pretty simple and intuitive.  In political science, we have been using it heavily since Anthony Downs published An Economic Theory of Democracy in 1957.  Put every voter on a line, representing the liberal-conservative spectrum.  The median voter along that line is the one who has 50% of the electorate to the left, and the other 50% to the right.  Represent candidates' platforms with points along that line.  Whoever is closer to the median wins, so candidates should converge towards the median.  So, if the median is 0, the Democrat is at -1, and the Republican is at +2, the Democrat should win.

Of course, we aren't there yet.  This is a primary.  That means two things.  First, the important median isn't the country's median, and second, we've got more than two candidates.  Yes, and both are important.

Since we are in the Republican primary phase, the median should be more conservative than the overall electorate, and that creates pressure on candidates to move, not to some centrist platform, but to some conservative platform based on the electoral Willie Sutton principle.  The most conservative candidate should win, so move to the right, right?

Problem:  where does Trump fit on that spectrum?  Nowhere.  Trump isn't a conservative.  He isn't a moderate.  He isn't a liberal.  He is ideologically incoherent.  Since mounting his rebel attack on the Republican Empire, he has performed the rituals of conversion, such that his current platform includes tax cuts, opposition to abortion, etc., but in the past, he has advocated single-payer healthcare, tax increases specifically on the wealthy, abortion rights, etc.  What would he do if elected?  Aside, obviously, from making America great again?  He has positioned himself most vocally on immigration.  That is an issue that cuts strangely across ideological lines.  Self-identified conservatives are split on the issue, with Chamber of Commerce types advocating more open borders for economic reasons, with social conservatives treating it as a law-and-order issue.  Among the groups who help to define liberalism (I am conceptually influenced by Hans Noel here) are unions, with many union members preferring restrictive policies to keep supply from pressing down wages on low-skill jobs.  The fact that immigration cuts strangely along ideological lines was what deluded some people into thinking that immigration reform might happen, even with divided government.  Hah!

Rather than going for the straight-up conservative vote, then, Trump seems to be looking for what E.E. Schattschneider would have called a cross-cutting issue, changing the dimensions along which the rebellion is fought.  Basically, then, the policy "space" is malleable.  The proper reference is Star Trek, then, not Star Wars because in the former, they use warp drive for propulsion so...  I promise I'll stop.  Please don't give me a wedgie.

Basically, though, if this all came down to a left-right thing, Trump would have a problem because he is so clearly not a real conservative.

Then again, there's the number of candidates issue.  Think about the weird dynamics of a three-way race.  Just to bug Rick Santorum.  Let's say the median is 2.  If two candidates are at exactly 2, and a third is just to the left of 2, the candidates at 2 will get around 25% of the vote each, and the candidate just off-center wins with 50%.  Adding more candidates to this kind of model makes things blow up like the Death Star...

I promised I'd stop.  Anyway, weird stuff happens with more than two candidates, and Trump got into this thing with enough competitors to give Rick Santorum a seizure.  Once it is down to two, then, the more conservative candidate should win, then, right?  Well, PredictWise has the betting odds pretty much down to Trump and Rubio, but as long as Cruz stays in the race, the best salvation for the spatial model is that Trump is winning because Rubio and Cruz are splitting the conservative vote, both being actual conservatives.  This may turn out to be the real test, then.  If it comes down to Trump and a real conservative, does the real conservative consolidate the vote and defeat The Donald?  One could easily imagine Cruz dropping out, and his voters going to Trump rather than Rubio as an antiestablishment vote.  If that's what happens, then the basic spatial model will really be failing us, unless the Republican primary vote isn't as conservative as some would have us believe (stay tuned for the next installment).

So, as the contest moves today into a hive of scum and villainy more wretched than Mos Eisley, remember that if the spatial model were right, then someone who is so obviously not a conservative should have real problems in a Republican contest.

Sorry.  Hey, if I had gone the Star Trek route, I could have worked in William Riker!

Never mind.  I'll go back to jazz.

No comments:

Post a Comment