Bundesverwaltungsgericht: Michael Greve Replies to His Critics

Thanks to Philp Wallach, Ilan Wurman, and Michael Rappaport for their thoughtful comments about my Liberty Forum essay, and thanks to Richard Reinsch for instigating this friendly sparring match. This fifth round features a few more thoughts on what prompted me to launch this trial balloon, on what I take to be common or disputed ground, and how we might proceed.

My bottom line is that the administrative courts must be generalist, independent of any government line agency, and not reversible by the executive. They must act as a systematic check on the executive’s interferences with citizens’ private affairs. And there must be enough of them to perform that function. On most anything else I’m open to debate.

For example, I’m open to building an Administrative Judiciary over time rather than in one fell swoop, and equally open to Ilan Wurman’s idea of administrative courts that operate (like magistrate judges) as adjuncts to Article III courts—just so long as they are adjuncts to courts, not to agencies. But I take it to be common ground that in whatever form, tribunals of this description would be a serious break with the appellate review model.

Empirics and Principle

Philip Wallach wants more empirics before we embark on such a bold institutional experiment. I concur. (I’ll even wear the “functionalist” label that that Wurman guy is trying to pin on me.) Anecdotally, administrative lawyers in Germany win cases on a fairly regular basis. (And “win” means a binding verdict, not a lame remand to the agency.) That’s largely because the Verwaltungsgerichtsordnung (the VwGO, or Fau-vay-gay-oh) is organized around substantive rights, not (like the Administrative Procedure Act) proceduralized forms of begging for mercy.

I haven’t had much success in finding empirical studies on Germany’s system to back that up. Happily, though, we don’t have to travel that far. The U.S. Tax Court—a plausible model for the administrative courts I have in mind—adjudicates the IRS deficiency determinations de novo. It reverses the IRS in a remarkable number of cases (far higher than what you see in administrative cases in Article III courts), and by all accounts it is working remarkably well. Also, three of our 50 states (Louisiana, South Carolina, and North Carolina) have an administrative judiciary of the sort I envision: generalist; independently appointed; de novo review, or close to it; not reversible by administrative or executive agencies. Deep down in Louisiana, citizens—but, crucially, not the government—can appeal adverse tribunal rulings to the regular courts. Wow. Well worth a look (more below).

At the risk of sounding dogmatic, what I do resist is the suggestion that our current administrative tribunals might be working well enough to leave them alone, save perhaps for some limited, agency-specific reforms. The AdLaw profession thinks that way; constitutionalists can’t.

The British Empire worked tolerably well in 1776. The Bostonians nonetheless dumped tea into the harbor—not because the tax was the final straw but because they thought a critical line had been crossed. The line that’s being crossed here is that prosecution and adjudication belong in truly separate and independent institutions. Once you accept that some executive-quasi-judicial hybrid may be good enough, you’re playing in Adrian Vermeule’s and Jerry Mashaw’s and Richard Fallon’s universe, where the only argument that counts is the comparative competence of courts and agencies to render rough justice. I won’t play in that sandbox. The better course is to recognize the principle, and then to think through the institutional consequences and implications.

Maybe (I’m guessing) that insistence underlies some of the differences between myself and my critics. For example, I disagree with Mike Rappaport’s focus on expertise and decision costs. As a practical matter I don’t see how his proposed institutional division among courts—science, medicine, economics—would work. Environmental or health-insurance cases involve all three disciplines (and alas, once you meet an “expert” in all three disciplines he’s Jonathan Gruber, the architect of the Affordable Care Act’s three-legged stool; and what a fitting word that is). Far more important, though, is the point of principle: It’s okay to put expertise and decision costs front and center when it comes to veterans’ benefits, disability benefits, food stamps, government employment disputes, and the like. Parts of the existing adjudicative machinery for these sorts of benefits seem to work reasonably well, others not so well. But it’s a technocratic problem, not a rule-of-law problem.

In the regulatory controversies I have in mind, administrative adjudication is a rule-of-law problem; and the calculus should shift. We put up with monstrous decision costs in criminal cases because we think that something important is at stake. Well, I think that’s also true when big gummint messes with citizens’ livelihood in a non-criminal, civil fashion. Committing those matters to independent judges is of course going to increase decision costs, especially for the government. We should not make the proceedings needlessly cumbersome—for example, I’m no fan of having juries on the premises. But with that limited proviso, let the costs be what they may.

Institutions and Law

I’m struck that none of my critics has taken issue with the idea of an administrative judiciary, alien though it is to our legal tradition. (Even Ilan Wurman, if I understand him correctly, is open to a “formalist” version of that program.) We all agree, then, that Hayek was on to something? In other words let’s hear it for the Rechtsstaat? By all means.

As Philip Wallach duly notes, Verwaltungsgerichtsbarkeit is too long to emblazon on a MAGA hat, and even the “grass tops” professoriate might be too invested in doctrinal fights to attend to institutional questions and concerns. But that’s just my point. The conservative-libertarian legal intelligentsia has spent 34 years choreographing the angels on Chevron’s pinhead. (Zero step?  Double step? Triple step? A King v. Burwell jump to the left?) Good fun, but we shouldn’t mistake marathon dancing for law reform. If you’re serious about judicial control of the executive, you need courts that are built for that purpose. Ours aren’t. Thus, I want different institutions—an Administrative Judiciary—not just for reasons of institutional capacity but also because I believe we need a very different body of law. (Gabe flagged that point in one of his kind comments on my original piece.) And yes, I am willing to gamble that an Administrative Judiciary would, over time, develop a sensible, rights-centered body of law.

Why Believe That?

That’s what Mike Rappaport and Philip Wallach are asking, each in his own way. I’ll dispense with facile answers. “It’s worth trying because things can’t get any worse”? Nope, things can always get worse. “It works in Germany, so why not here”? Also wrong, for the most part. That was the attitude of Ernst Freund, Felix Frankfurter, and other sages who imported pieces of German AdLaw—not the real thing, mind you, but the absolute worst traditions of that discipline—into this blessed country. Contemporary German administrative law’s successes, I’ve suggested, are a product of legal traditions, historical contingencies, and institutional dynamics that we can’t replicate. We can learn a great deal from the German model but we can’t just import it. Doing so would cause another transatlantic shipwreck.

What then is my answer? It’s in the nature of institutional reform proposals that you can never make a wholly convincing, “yes-this-will-work” case. After all, most institutional experiments fail. Then again, our exchange seems to be informed by a Buchananite spirit that says: Think about institutions, because you must. Be careful and realistic, though, and be clear about what you’re trying to accomplish.

In that spirit, I’ve tried to identify the institutional and constitutional problems to which the new Administrative Judiciary is supposed to be the answer. I have articulated non-negotiable institutional design features, and I have emphasized the importance of getting jurisdiction right. You don’t want benefit claims in these courts—the whole point is to re-articulate a lost distinction between food stamps and a professional license, or patent. You’ll want an amount-in-controversy requirement, to prevent the Administrative Judiciary from becoming a small claims court. Contra Rappaportem, you exclude rulemakings from these courts, lest they start thinking like ersatz regulators or legislators, or for that matter like the D.C. Circuit.

I’ve further tried to explain why this Administrative Judiciary would be likely (likelier, at any rate, than a general Article III judiciary) to build a constituency and an institutional culture that would develop—not start with, but develop—better administrative law doctrines over time. By “better,” I mean doctrines that are likely to impose meaningful constraints on executive interferences with private conduct. Granted, that’s not a certainty. And granted, I don’t know—beyond the examples I provided—what those doctrines might look like in detail. And granted, we do not understand very well why some institutions manage to build a mission-conforming culture and others don’t. But in large measure, it seems to be a matter of suitable design, and we have examples of successful experiments in that vein.

The federal judiciary was designed for certain purposes, by way of appointment provisions, tenure protection, and jurisdiction, and it worked admirably well for 150 years. The Office of Information and Regulatory Affairs (OIRA) was designed for an institutional mission and has developed a culture (“We’re the only ones who can say no to anything in this town”) that has held up remarkably well throughout partisan turmoil. The U.S. Tax Court is another example, as is the Court of Claims, as is the Congressional Budget Office. We can study and learn from those experiments (as well as from failed experiments like the over-specialized Federal Circuit), and it’s in that sense that I endorse Philip Wallach’s demand for more empirics. Comparative institutional analysis is needed. I hope to write more about administrative courts in other countries. One of these days I may head for a train to Baton Rouge, and see how that system works.

I cheerfully concede, too, that quite a bit more needs to be said about the kind of law one hopes to get out of a new Administrative Judiciary. I’ll even volunteer an observation that Ilan and Mike were kind enough to suppress: The proposal would compel a rethinking of much of what we teach in ConLaw I, and AdLaw, and Federal Courts. In that large-ish enterprise, though, I deem myself in the splendid company of an entire cohort of “anti-administrativist” jurisprudes. What I hope to contribute to the enterprise is a plea is for more institutionalism and a bit less abstract theorizing. I continue to think that that’s directionally right, and to hope it might produce fruitful thought about the pathologies of the administrative state.