Explaining Republican unity on the impeachment vote (and why my latest book fails here)

Some background first.  For three years or so, I have been listening to everyone from journalists to scholars utter some variation of, "surely this will be the thing that does it.  Republicans can't keep supporting Trump after this."  Fill in the this for yourself.  The Access Hollywood tape.  Charlottesville.  Take your pick.  I heard plenty of variations when the Ukraine story broke.  That the applied to the story, not the Ukraine.  Pedantry, thy name is Justin Buchler.  Anywho, here's how the tale went this time, as I overuse italicization.  Mike Pence wouldn't be so bad from a Republican perspective, and notice how plenty of Republicans have stayed quiet about the whole affair.  Maybe the Democrats will actually get some votes in the House!  Surely this will be the thing!

Aaaand, no.  Unless you count Justin Amash as a Republican, but he was effectively driven out of the GOP because of his support for impeachment.  In fact, that rather demonstrates the point, doesn't it?

I've been saying all along, both on The Unmutual Political Blog, and here, that Republicans will maintain their unified support for Trump.  That break from Trump is not going to come.  And it didn't.  Amash, as a reminder, had already been on the outs with party leadership when Boehner stripped him of committee assignments years ago, so he was already an outlier anyway.  Aside from him-- and he isn't technically a Republican anymore-- the Republican Party voted in complete unity against the House impeachment process last week.

As I have said, and will continue to say, the Republican Party will maintain complete, unified support for Donald Trump-- or near enough that the difference doesn't matter (e.g. Justin Amash).  Might Susan Collins, or Lisa Murkowski vote to convict?  Maybe.  I can't rule it out completely, but they don't matter.  There is zero chance of conviction, because the GOP will maintain its unified support.

Let's explain this, differently from how I have in the past.  And... somewhat against my latest book!

First, though, a brief digression on Mike Pence.  One of the variations of the surely-this-will-be-the-thing argument is that the Republican Party at the elite level likes Mike Pence.  That's why I knew before everyone else that Mike Pence would be the VP nominee!  That doesn't matter.  Removing Trump from office doesn't give the party President Pence.  It gives them President Warren.  (Or... whoever, but more likely Warren than anyone else given the state of the Democratic contest as I write this morning.)  That's why they can't flip on Trump, no matter what they think of him.  This is why I keep saying that it doesn't matter what happens, what evidence comes out, what charges there are... convicting Trump means inaugurating President Warren.  They will never choose that.

Moving on, though.  The bigger issue here is that I wrote a book!  And, much as I'd like to say it explains everything that's going on, it... kind of... doesn't.  So I suck.

Here's the deal.  My last book was Incremental Polarization: A Unified Spatial Theory of Legislative Elections, Parties and Roll Call Voting and verbosity in titles.  Essentially, it was about how parties constrain legislators to cast electorally damaging roll call votes, turning them into ideological extremists, one vote at a time.  It was about post-WWII politics, but there was a separate chapter about how the 2010 election turned the GOP into a different kind of party, more (unjustifiably) afraid of primary threats than general election threats.

This all seems tailor-made to explain what's going on, right?  Except, I don't think my model works very well here.  The better model is the one proposed by, well, at least a guy who blurbed my book-- Mathew McCubbins (along with Gary Cox).  Elaboration to follow.

Here's the basic argument from my book.  Before I explain why it doesn't work here.  Let's say you have a party in which Members of Congress are "cross-pressured."  They have electoral incentives to vote one way, and policy beliefs that run in the other direction.  The more unified the party is in its underlying policy beliefs, the stronger they want party leadership to be.  They give party leadership disciplinary tools to cancel out electoral pressure so that the party votes in a unified way.  That way, they achieve their policy goals.  Representation?  Not so much, because the whole point is to cancel out electoral pressure.  The result?  Extremism.

Why doesn't that model work here?  Well, two things.  First, a lot of the GOPers in the House really don't like Trump.  Give them the power to pull a Trump-Pence switcharoo without the concern that it would hand the White House to Warren and how many would do it?  Probably... a lot.  Not all.  There are some real Trumpists in the House, but there are a lot of fakers too, who are just going through the motions because they are terrified of those mean tweets.  So, on this matter, the party is divided in underlying preferences.

I kind of dealt with that topic in the book, but the way I dealt with it doesn't work when we factor in the second issue.  Thursday's vote on the rules for impeachment proceedings wasn't a policy vote.  It doesn't change policy at all.  So, its implications are really more about branding.  Put that together with internal party divisions, and my model doesn't work.

What model does?

"The Cartel Model," which was published in Legislative Leviathan, by Gary Cox and Mathew McCubbins.  Here's the deal with that model.  Each party is basically trying to construct an electorally beneficial "brand," and they try to maintain party unity around the brand that will help them collectively.  Towards that end, the majority party can construct an agenda that will help them, which the minority party can't do, so, you know, sucks to be them, but maintaining party discipline can still help with their electoral branding.

On the impeachment proceedings vote, Cox & McCubbins have a model that works much better.  Why?  Underlying preferences within the GOP are heterogenous, but they need to maintain a unified front for branding purposes, and it's not directly about policy.  Meanwhile the majority party is constructing an agenda that the minority party can't influence, with an eye on how that agenda will help its electoral prospects.  The minority party is complaining about its limited capacity to call witnesses without oversight by Schiff and Nadler, but, you know, Cartel!

So, who wins here?  Cox & McCubbins.  When the vote happened on Thursday, I figured I'd do a post today on how my book would explain the process, but... it doesn't.  Intellectual honesty matters, and Cox & McCubbins beat me here.

I'm not throwin' out my book, which I think explains a lot about votes on policy matters-- that's not what's happening here-- but it's important not to become so attached to your own arguments that you lose sight of their own limitations.

Comments