Wednesday, June 15, 2016

Political effects of Orlando

Now that I've chased off all my readers, I'll get back to what I really do:  analysis of electoral politics.  Nobody should turn to me for moralizing anyway.

As a general rule, I discount the effects of individual events on election results.  I'm what we call a "fundamentalist."  Elections come down to the fundamentals:  the state of the economy, underlying partisan divisions, etc.  Everything else is just noise.  Back during the primaries, I picked on Sides and Vavreck's The Gamble, but I think they have it pretty much right on the general election side of things.  Everything that commentators call a "game changer" winds up changing nothing.  Sides & Vavreck call these events "game-samers."  Even Romney's 47% video did roughly bupkis.

A shooting in June?  I don't see why that breaks the game-changer/game-samer mold.  And even if it did have an effect, in which direction?  There is an old idea in political science called "issue ownership," which is that there are certain issues that intrinsically benefit one party or another.  National security and crime are "owned" by Republicans, for various reasons, so anything that brings those issues to the forefront can benefit them.  Then again, the Orlando shootings involved an attack on the LGBT community, and public opinion has turned very much in favor of at least the LGB part.  That could play to the Democrats' advantage.

Obviously, there's the Islam thing, and that could play to the Republican's advantage.  Then again, the Democrats tend to be very strategic about the gun control measures they propose, keeping them to the extremely popular measures, which could work to the Democrats' advantage.

Basically, this is an odd hodgepodge of elements, and predicting the net effect is hard.  And that brings me back to my fundamentalist leanings.  The economy.  Then again, my favorite model has always been the Abramowitz "Time for a Change Model."  That model currently predicts a Republican victory, and Abramowitz himself doesn't buy it.  Read for yourself here.

Right now, PredictWise gives HRC a 74% chance of beating Trump.  The RealClearPolitics polling average has HRC up by an average of 5.5 points, with the latest poll-- post-Orlando-- putting her up by 12 points.

In 2015, the polls and the prediction markets diverged in the Republican contest, with the polls giving Trump the advantage, and the prediction markets discounting it.  Now, both the polls and prediction markets agree.  And they disagree with my favorite prediction model.

This is a tie election.  The Orlando shooting is random noise in a tie election.  In a tie election, in a sense, everything can matter, but in another sense, nothing matters because everything is just canceled out.

No comments:

Post a Comment