A Proposal for a New Federal Clerkship Hiring Plan (When The Current One Collapses)

It's time to stop trying to cartelize the market for law clerks

Jack Goldsmith writes on Twitter about the latest iteration of the federal clerkship hiring plan, which was supposed to involve a widespread pact among judges not to hire students until the summer after their 2L year. Many judges did not agree to follow the plan, but among those who did, it seemed like there might be a stable equilibrium.

Not so, Jack argues. When 2Ls apply this summer, he says, they will discover that "many on-plan judges went ahead and filled some (and for some judges all) of the 2021 clerkship slots with other students (3Ls, grads, etc.)" This includes, says Jack, some of the judges who organized the plan in the first place. (Read the whole thread for more.)

If all of this is true, and I trust Jack very much, though I don't have direct testimony myself, it will make the plan seem quite unfair and it seems unlikely to be sustainable. This would mark a much faster collapse than the last round of the clerkship hiring plan.

On the other hand, some off-plan judges are hiring students as soon as the January and February of their 1L year. And off-plan clerkships generally leak out by word of mouth, with students and professors relying on personal relationships and networks to find out who might be hiring when.

If and when the plan collapses, I have a proposal for a new plan, which I mentioned parenthetically in an older post on this blog. This new plan would impose a modest restriction on clerkship hiring in order to have a fairer and more efficient competition for these positions.

1. Judges can hire law clerks any time they like, and law clerks can apply any time they like.

2. All judges agree to publicly post, on Oscar or the equivalent, when they are hiring law clerks and how many positions they are trying to fill. They also agree to post when the clerkships are full.

That's it. You could embellish this plan with some of the other rules like the rule that applicants must get at least 48 hours to decide, or with a rule that the posted positions must be kept open for X days, but I see those as best practices that need not be part of the core plan for it to succeed.

Cartels are hard to organize, and incomplete cartels are destined to fail. That is the lesson of the previous two clerkship plans. And worse, the current clerkship plan has generated a significant loss in transparency, as those who circumvent it feel various pressures to go outside of the usual channels. It may be time to focus on transparency, and forget the strange dream of the cartel.

[Cross posted from Summary, Judgment]

Comments

[Bob from Ohio]

Oh dear, those poor privileged Ivy Leaguers. Denied a clerkship. My heart breaks.

[y81]

January of 1L year? That seems like another strategy to help the well-connected and well-counseled (which are basically the same people). Someone who doesn't know very much about legal hiring is supposed to get their first semester grades, immediately think, "Wow, I'm a viable candidate for a federal clerkship," and "That is obviously my best career path option," and start submitting applications? And how could a person have professorial or judicial connections that soon in their law school career, unless said connections were familial in origin?

[Beldar]

January of 1L year? That seems like another strategy to help the well-connected and well-counseled (which are basically the same people).

One doesn't have to be well-connected to be well-counseled. The pool of the former may be static, or at least beyond the power of rules to change. But I'm in favor of enlarging the pool of the latter, and I think that's eminently doable.

In fact, as compared to the state of things when I was making clerkship applications as a 1980 graduate, I would respectfully submit that the information now available on the internet — thanks to folks like the Volokh Conspirators and like-minded people with relevant, recent, and deep experience — makes it possible for essentially anyone who's adequately motivated to become well-counseled.

[Peter Gerdes]

Seems to me that plan still results in a super bad outcome where all the judges have to make hiring decisions far before they have sufficent information to be really sure (later grades) which disadvantages those without connections etc..

Couldn't the chief just order federal judges to abide by certain rules? If not get congress to pass a law.

[Lee Moore]

This touches on my question which was "what is the perceived evil of not having a plan ?"

You pre-answer me with : "all the judges have to make hiring decisions far before they have sufficent information to be really sure"

But if they make their decisions "too early" this implies that they will select some dummies along with stars. Meaning that a judge who waits will still have some unfound stars to scoop up later. Shrewd early judges who are good at spotting budding stars and avoiding dummies will benefit from their skill. Judges who are poor pickers on slim evidence, but who still insist on rushing in, will suffer the consequences of their lack of talent in this matter.

So it's still not obvious why the judges need to be regulated at all. This is - like most - a market that regulates itself.

[Peter Gerdes]

I agree if we assume that judges are all homogeneous and we don't care who gets assigned where that wouldnt be such a problem.

However, what I expect to happen is the same kind of unraveling we've seen in various other markets (good examples on market design blog) where the less prestigious judges try to send out exploding offers to the most promising students super early in hope of landing the caliber of clerk who would be otherwise take a clerkship with an appealete judge or even a justice.

So now we end up with a situation where the most promising students have to decide to commit early on to take a less prestigious clerkship or gamble on a more prestigious one. In actual fact this same pressure has resulted in similar markets making binding selections years in advance. Even if that doesn't happen you end up with a much less than ideal assignment (best applicants should go to best clerkships).

[Lee Moore]

I agree if we assume that judges are all homogeneous and we don’t care who gets assigned where that wouldnt be such a problem.

We who we ? Each judge and each clerk is trying to do the best for himself or herself. That the aggregate allocation pattern of clerks to judges fails to meet the approval of some third party observer, like you or me, is not an example of "market failure."

However, what I expect to happen is the same kind of unraveling we’ve seen in various other markets (good examples on market design blog) where the less prestigious judges try to send out exploding offers to the most promising students super early in hope of landing the caliber of clerk who would be otherwise take a clerkship with an appealete judge or even a justice.

And the problem is what ? The less prestigious judge is offering both a position and the early resolution of painful uncertainty to the proising student. Sure, after the promising student has passed top of the class a year or two later, he or she might conclude that it would have been better to wait. But that is just because the uncertainty has been resolved by the pasage of time. But while uncertainty reigns, eliminating it is valuable. Why do you think farmers sell crops forward to lock in a price ?

So now we end up with a situation where the most promising students have to decide to commit early on to take a less prestigious clerkship or gamble on a more prestigious one.

Excellent practice for the marriage market, neh ?

In actual fact this same pressure has resulted in similar markets making binding selections years in advance.

So, cartels will cartel and they find it difficult to make it stick. So what ?

Even if that doesn’t happen you end up with a much less than ideal assignment (best applicants should go to best clerkships).

We return full circle. Ideal for whom ?

Moreover, "best applicants" and "best clerkships" is in the eye of the beholder, not you or me. There are all sorts of reasons why Judge A and Judge B might disagree on the qualities they want in an applicant. And though most applicants might agree that the prestgiousness of a clerkship is the most important thing, it won't be the only thing. Location, the judge's talent at and commitment to mentoring, the judge's connections, the personality of the judge and other clerks, und so weiter.

[Peter Gerdes]

Also remember that applicants often have other constraints like wanting to live in cities with significant others etc. The more you allow some judges to lock in some applicants super early the less ability the applicants have to plan their lives because they have less information.

[Peter Gerdes]

Whether or not my theoretical arguments convince you there is ample empirical evidence that offers start getting sent earlier and earlier and that suggests that the advantage to waiting isn't sufficent.

http://marketdesigner.blogspot.com/2009/06/unraveling.html?m=1

[Lee Moore]

The blogster seems to be a self referencing economics Prof who has coined "unraveling" to mean "not in accordance with what I like."

There are hordes of good reasons why people in commercial transactions (and indeed in non commercial transactions*) like to pay to minimise uncertainty. If lots of people like to do this, locking in transactions - whether that be earlier transacting or buying insurance, or futures or options - is not market failure. it's part of the market for managing risk.

* Mrs Moore insists on getting to all appointments, airports whatever, an unconscionably long time in advance of the appointment time. Plus an extra allowance for "once in a decade" traffic conditions. This obviously wastes a ludicrous amount of my time. She is content to pay (and have me also pay) this vast cost to resolve her uncertainty. Is this deranged ? IMHO yes, but IHO no.
We all value things differently.

[Vandalia]

Well, the medical residency match has operated as a successful "cartel" for over sixty years. Medical students and residents have complained about many things related to training, but the match itself is rarely one of them. If the "match" works for medical residencies and fellowships - and has a nearly 60 year track record - there is no reason it would not work for judicial clerkships.

[Peter Gerdes]

The medical residency program works by having everyone rank canidates and then running a global algorithm the results of which the hiring parties commit to honor.

I’m skeptical judges will be particularly inclined to accept that deal especially given that it forces the hiring parties to rank a whole bunch of canidates and canidates to rank many employers. Something which is harder the more personalized the process.

[Beldar]

Well, the medical residency match has operated as a successful “cartel” for over sixty years.

I have no expectation that one could ever get a majority, or even a good-sized plurality, of federal judges to voluntarily commit to such a thing. It'd take an act of Congress to ram it down all the necessary throats.

BUT: I agree with you that the legal profession — beyond just judges and judicial clerks — very badly needs to emulate the medical profession when it comes to post-graduate training and specialization. In particular, we desperately need a national board certification program comparable to the medical profession's national boards; except for the hardy few ultra-generalists who aspire hang a shingle outside school, virtually every lawyer with any aspiration to be good at any particular kind of practice needs something comparable to the internship (PGY-1), residency, and fellowship programs. Our profession is badly failing the public welfare both because we don't provide systematic and measurable post-graduate training in specialties and because we aren't helping the lawyer-hiring public choose the lawyers best suited to their individual needs.

[David Bremer]

I'm not intimately familiar with the medical-residency match, but isn't the obvious difference that medical residents are selecting institutions versus selecting individual doctors? This sort of system might work if you could rank courts (e.g., 7th Circuit, 9th Circuit, N.D. Cal., 11th Circuit, and on down). But judges are going to resist a system where they don't have a chance to talk through the personalities and views of potential candidates.

[Robert Beckman]

Exactly right, but even that’s largely fixable (where fix means to improve from the perspective of the users).

You’ll have to have a large registry, and in it a fixed set of characteristic, so that applicants can indicate that they want a judge that is (purposivist, over 70, on the east coast , in an urban location, does travel to further courthouses, etc), and a judge can indicate that they want a student who is (female, ethnic minority, only from Yale, not an originalist, over 30, etc). These can then be matched for round robin screening calls - the ones who both swipe right can then arrange longer discussions.

This largely (though imperfectly) solves the information gap due to the larger scale - a student who wanted Supreme Court first, or anything at all if not that, could make that clear in her priorities. A judge who needs clerks, but is ambivalent on which (perhaps because they have a huge paper backlog and just need hands rather than minds) can make that clear too.

[Beldar]

This new plan would impose a modest restriction on clerkship hiring in order to have a fairer and more efficient competition for these positions.

1. Judges can hire law clerks any time they like, and law clerks can apply any time they like.

2. All judges agree to publicly post, on Oscar or the equivalent, when they are hiring law clerks and how many positions they are trying to fill. They also agree to post when the clerkships are full.

Number 1 seems to be a rule saying there aren't any rules regarding timing, which seems unnecessary.

Regarding Number 2, I suspect that for at least some other sorts of civil service jobs, that's required by law or regulation. Regardless, it seems like something the Chief and the Judicial Conference ought to be able to impose by fiat without too much backlash. Like any other market, this one will work best — most efficiently and competitively — when there is real-time, accessible information flowing continuously. It's certainly less onerous than, say, requiring judges to file their financial information for public disclosure each year!