Why do law professors hate the attitudinal model so much?

 The Supreme Court continues to make news, and I have no real comments on any new rulings.  Rather, let's step back and consider some interdisciplinary disputes among scholars over how to interpret and analyze the behavior of these bizarre critters known as "Supreme Court Justices."  Are they, or are they not merely ideologically motivated politicians in stupid costumes?  Let us take the sartorial criticism as a given, and focus on the analytic substance.

The attitudinal model is a dead horse, which we shall beat, because beating a dead horse is better than beating a live horse.  I like horses, and really, animals in general.  (People?  No.  People suck, but animals are cool.)  Anyway, recall the aphorism from George E.P. Box:  "all models are wrong.  Some are useful."  The attitudinal model is a political science model which we use to explain the voting decisions and formal opinions of judges and Justices.  It falls into the "useful" box.  (Spare me your groans.)  The question is, how useful?  If we're honest, very useful.  Here is how it works.

You know that left-right thing?  Some political scientists and other scholars are intentionally obtuse about the word, "liberal," but in American politics, we describe the left-right spectrum as the liberal-conservative dimension, and I won't belabor the history of the development of that terminology yet again, particularly since those who are intentionally obtuse about the terminology, by definition, don't want to know why the word means something different in America from the UK.  Such people are as unreachable as Trump voters with respect to "the Big Lie."  Anywho, the description of politics in terms of left and right actually goes back to the French National Assembly, a few centuries ago.  Parliamentarians would sit, physically, on either the left or the right side of the chamber, depending on their opinions on how powerful the executive should be.  That really is the start of thinking about politics in left-right terms.  Fast-forward a few centuries and you get a one-off paper in 1929 by Harold Hotelling, which proposes a model of businesses locating along a street in a market in which consumers are lazy, and don't want to walk any more than they must, and then blah, blah, a few more decades gets you to Anthony Downs and his 1957 book, An Economic Theory of Democracy, and by that point, political scientists and economists just cannot think about ideology except in terms of space and distance.  Politicians and voters and everyone?  Stick'em all along a line, representing the liberal-conservative dimension, everyone seeks proximity, and glorious, glorious math follows, yeehaw!  Also, I get to publish, tenure follows, so yee-fuckin'-haw!

But none of this has anything to do with law!  Right?  For you see, within the hallowed halls of law schools, and courtrooms, and judges' chambers, everyone is above mere ideology, right?  Right?

Have you not listened to all of those enlightened people during their confirmation hearings?  They tell you!  Forthrightly!  They are not ideologues!  They have no personal opinions, or at least, none that would influence their rulings!  They are the superior... nay, the master race, when it comes to anything approaching governance.  They read the Constitution!  The law!  They follow the Constitution!  And the law!  No ideology here!

Um... OK, so us political scientists?  We call bullshit on this.  We take those models of left-right politics, which we use to explain elections, voting in Congress, and all that?  And we ask, do these models work to explain the Supreme Court?  When applied to the Supreme Court, we call it, "the attitudinal model."  Judges and Justices can be put along the same liberal-conservative dimension that we put voters, Members of Congress, presidents... everyone else, and just as other political actors seek proximity, so do Supreme Court Justices.  The difference is, everyone else is honest, and they're fucking shitbag liars.  Why was Susan Collins stupid enough to believe Kavanaugh?  Like I told you at the time:  she is the dumbest person in the history of the U.S. Senate.  We were all screaming at her, "he's lying!  What, are you fucking stupid?!"  Yes, she is stupid.  Yes, they are lying.  The gist of the attitudinal model is that they're all liars.

What is the nature of the lie?  The nature of the lie is as follows.  Kavanaugh, for example, is nothing more than a basic conservative frat-bro, complete with some drunken rape attempts.  In the party of Trump, rape is no problem, and Donnie-boy himself is happy to bond and compare notes.  For the purposes of this post, the point is more the conservatism.  Yes, Kavanaugh was obviously a conservative.

But here's the lie.  The lie is, no, I'm not a conservative.  I interpret the Constitution, and here's the philosophy of interpretation I take.

And that's the shit that law theorists discuss.  Originalism!  No, it's a living document!  You have heard these terms, I am sure, and probably laughed at them.  The thing is, they are actually just simplified versions of very real ideas bandied about by law professors, who, to varying degrees, take the no-I'm-not-an-ideologue thing seriously.

Does the Federalist Society know what they're all about?  Yeah.  But you can also find some law professors who disagree with, and even get angry about the attitudinal model.

No model can ever be sacrosanct!  And of course, the attitudinal model is definitionally wrong.  It is a model.

What happens if you take a set of judicial decisions and apply the mathematical algorithm that we use to estimate congressional ideology-- NOMINATE, originally developed by Poole & Rosenthal, and incrementally modified by others over time?  You'll get a bunch of scores.  But here's the thing.  There is more noise in the data, because a lot of the shit that goes to SCOTUS, and in fact, more than what Congress addresses, has no ideological component.

So if you observe the coalitions without paying any attention to the substance of the votes, in Congress, you still see basically nothing but left/right divides these days.  At the SCOTUS level?  You can find a lot of weird coalitions.  Remember, the attitudinal model is a model.  WRONG!  It's WRONG!

But... also, useful.  How useful?  The scores revealed by the application of something like Poole & Rosenthal to SCOTUS rulings give you what we call, in technical terms, a "fuckton" of predictive power on any issue with ideological content.  Keeping in mind, also, that Justices have a bit more freedom than lowly House members.  For example, Roberts tried to split the difference on Dobbs by opining that the specific restrictions were hunky-dory, but he didn't want to reverse Roe and sign onto that sweeping ruling that Alito blew threw the entire Western world's supply of Kleenex and lotion writing.

Predictive power.  It's all about predictive power.  How do we assess "usefulness?"  That's it.  Right there.  Does the attitudinal model have predictive power?  Yup.  Lots and lots of it.  I can tell you, right now, how Ketanji Brown Jackson will vote on every single case with an ideological dimension.  Why?  because while she was sworn in yesterday, I was not born yesterday.  Do you want to pretend I'm wrong?  I know exactly how Kagan and Sotomayor will vote.

If there is ideology at stake, is Thomas going to cast a... fine, sarcasm-quotes, "liberal" vote?  Wanna bet?  What kind of odds would you need, if money were at stake?  Thinking about it in monetary terms makes it clear, doesn't it?

Roberts is as incrementalist, but a conservative.  Those two characteristics are in tension, so sometimes, he'll piss off the conservatives.  Like I said... a model.  But in terms of predictive power, does the model have it?  Yup.  I can predict a fuckload with it.

Now.  How's about all that originalism stuff.  Or "living document," or any of the other stuff that... yeah.  Anyway.  Where is the predictive power, and how many assumptions must be added?  How many variables?  How much parsimony is abandoned?  Do you see why we, political scientists, tend to like the attitudinal model?  On the stuff that one would expect it to predict, it has a lot of predictive power with parsimony.

But wow can it piss off a lot of people in law schools.

And it isn't necessary for it to do so!  There are two ways for this to go.  One is for the law side of academia to look at the predictive power of the attitudinal model and say, woe!  Sorrow and woe, the practitioners have fallen from the true faith!  And in fact, we see such dynamics elsewhere.  If a nominee were selected from academia, and venerated within academia, the disincentives would be obvious, but the norm these days is to select a sitting judge, in which case we could envision a theorist/practitioner division.

Yet we observe no such thing.  Instead, the theorists claim that the practitioners are as above everything as the theorists themselves.

Do they do so by contesting the science?  No, law and science are different.  They argue on different terms, so rejection of the attitudinal model does not come from disputes over the econometrics.

Rather, it is interpretive, and anti-parsimonious, and contains no predictive model.  Which, to a scientist, is just very strange.  The core of how we work, in scientific analysis, is the quest for a parsimonious predictive model.  The rejection of parsimony, and the refusal to accept the responsibility for prediction shifts the terms of discussion to something that makes no sense to a scientist.

So what's the point?

The point is this.  If Supreme Court Justices are all just liars... then what's the point?  If law schools and law journals and all of that stuff... if all of their highfalutin theorizin' about how to interpret the Constitution and the laws without ideology falls by the wayside and at the end of the day, the most eminent graduates of the most eminent law schools in all the land, nominated by the president and confirmed by the Senate, throw in the towel instead of using that most useful of items as a thinkin' cap and just say, fuck it, what policy do I want, ideologically-motivated policy-maker that I am... then what's the point?

What's the point of those law journals?  (He writes, having published a few law review articles...)  What's the point of the bar exam, and all those hoops?  If it's all just kabuki theater, then...

Then the first point is there is a fuckload of money being wasted, because law school costs a fuckload of money, but the broader point is to look at this from the perspective of a law scholar.  It would call into question the entire endeavor.

Do you know what "quantizing" is?  As long as we are thinking in quantitative terms, let's throw this one out there.  It actually has nothing to do with social science.  This is about music.

Your music sucks.  This is part of why.  Part of the production process in modern music is "quantizing," which means taking the recording, and snapping everything to a grid, so that everything hits at exactly the beat it is supposed to hit.  Sounds great, right?  Actually, no.  It sounds fucking terrible.  Part of what makes music sound good is a combination of the imperfection and the creative element of hitting a bit ahead of or behind the beat.  The surprise.  Without that, it sounds like a machine because it quite literally is a machine.  I can list a bunch of the great jazz drummers, and you probably won't recognize them, but here goes.  Art Blakey.  Elvin Jones.  Max Roach.  Philly Joe Jones.  Tony Williams.  Roy Haynes, yeah, fine, let's go back to Buddy Rich and Gene Krupa and that crowd but obviously my tastes lean a bit later.  Fuck it.  You want some names you do know?  Fine.  Ginger Baker.  Serious jazz drummer.  Yes, there are good rock drummers, but the point is, they actually played, and it wasn't just technique.  It was a sense of when to hit just ahead of or behind the beat.  These days, producers for any pop genre (which is anything you listen to) use software to move that drum hit, and everything else, to a specific beat within a measure, which does two things.  First, it fixes any true errors, but the flip side is that it takes away the intentionality that can come when a truly sophisticated instrumentalist-- drummer or otherwise-- moves the beat just a hair for a specific effect.

So what happens if you are a music teacher.  Do you pretend that quantizing doesn't happen?  That it doesn't exist?  If you are teaching someone who wants to learn jazz, then great.  Go forth into a life of art and poverty!  The producers over at Blue Note probably aren't going to demand that they quantize everything.

But literally every recording that comes from a major pop studio (broadly defining pop) will quantize and autotune the shit out of everything.  And yet somehow, the shit will remain.  If you are teaching people, and pretending that this isn't happening... you're deluding yourself.

Perhaps this is a poor analogy.  I don't think Julliard is pumping out grads for the pop factory, so the analogy gets a little strained.  Yet if quantizing infected Blue Note and the other jazz labels, the professors at Julliard, Berklee and the other music schools would have to ask some serious questions about what the fuck they'd be doing.

It's kind of like that.  To the extent that judges and Justices are nothing more than ideologically-motivated politicians in stupid costumes, a lot of what gets taught in law schools is just bullshit.  Meaningless games played only in law schools and in law journals.

One can see how such an idea would discomfit the players.

That doesn't make the model wrong.  The fact that the model is a model is what makes it wrong, but the fact that it has predictive power is what makes it useful.

So how about a really awesome drummer?  It's Friday, so let's go with jazz.  Kahil El'Zabar.  Here's the title track from Ethnic Heritage Ensemble's The Continuum.