Monday, January 2, 2017

Assessing democracy in the aftermath of Trump's victory, Part XIII: Policy positions

What about policy?  I've spent the last however many posts in this interminable series talking about the concept of valence characteristics, and how voters assessed each candidate, but the point of valence models is that the valence "dimension" is supposed to exist separate from but in addition to the candidates' policy positions.

So, let's talk about that.  When we use valence characteristics in a model in political science, we do so in a "spatial" model.  Usually, that spatial model has one dimension-- the liberal-conservative dimension.  (Why?  Because the math blows the fuck up with more than one dimension.  The subject of another post!)  In order for that dimension to make sense, we need to be able to put everyone-- voters and candidates alike-- along the liberal-conservative dimension.  Can we?  Um, no.  It's a model, and as George E.P. Box said, "all models are wrong.  Some are useful."  The question we always confront with the "unidimensional spatial model" is whether or not it is wrong enough that it is just no longer useful.

Can we put Hillary Clinton on the liberal-conservative scale?  Yup.  Keith Poole & Howard Rosenthal developed a method called "NOMINATE" which calculates ideology scores for legislators based on how they vote.  Clinton was a Senator, and as a Senator, you could put her on a scale from -1 to +1 at -0.4030 in her last term.  You can get these scores yourself at Voteview.  That score can explain nearly every vote she ever cast.  How liberal is -.4?  Pretty darned liberal, but not Bernie Sanders-liberal.  That intolerably self-righteous, dumbass fuckin' hippy is at -.7170.  On the flip-side, Ted Cruz (whose name is, itself, an insult) is at .8780.  Who is at 0?  Susan Collins.  What does it mean to be at 0?  It means having no principles.  I'm just a misanthrope.

But what about Trump?  There's the problem.  We can't put him anywhere on that scale.  I, as a political scientist, can assert that Trump will sign whatever tax cuts that Paul Ryan and Mitch McConnell pass, but consider him from a voter's perspective.  What if all you had from Trump was his history of policy statements and advertisements?  Could you formulate any reasonable assessments of what he would do?

No.  And that's a problem for any spatial placement of Trump.  It is, in fact, a reflection of Trump's lack of policy beliefs because he has never really thought about policy before, but this is an important point from the electorate's perspective.  Trump asserted repeatedly that he would repeal Obamacare and replace it with "something terrific."  Yes, an intelligent voter should realize that this means Trump is a transparent con artist, but from a spatial placement perspective, there is no way to put "something terrific" anywhere on the liberal-conservative spectrum, and that's my point here.  Clinton's position was to make some incremental changes to Obamacare.  Trump's position, on the basis of his own statements, didn't exist and couldn't be put on the spectrum because "something terrific" is meaninglessly vague.

So, we come back to George E.P. Box's aphorism.  Is the unidimensional policy space useful here?  Not with Trump.  Not when a reasonable voter can't put him anywhere on that liberal-conservative dimension, even if they could put themselves there (and many can't do that, but that's another story...).

Could we combine policy and valence in the 2016 election?  Well, yes.  But it's complicated.  And the subject of the next post!

No comments:

Post a Comment