Friday, January 27, 2017

Law's Abnegation: A Reply to the Symposiasts

Guest Blogger

Adrian Vermeule

For the Symposium on Adrian Vermeule, Law's Abnegation

            I'm grateful to Jack Balkin for organizing the symposium on Law's Abnegation, and to all of the symposiasts for their generous remarks and for their time. It seems to me that no single common theme emerges; their points, while interesting, do not converge on any single line of conversation. So I will mention a few disparate issues, attempting to touch on all of the symposiasts except Sandy Levinson, whose cogent and (overly) charitable entry speaks for itself and calls for no response.

Law's Sovereignty

            Two of the symposiasts, Andy Koppelman and David Dyzenhaus, are primarily interested in law's sovereignty. Koppelman makes one central point: "It isn’t abnegation for the law to stand by and let someone do what they’re entitled to do.  It’s just the ordinary operation of law." Well, yes and no, as I try to explain at several places in the book. The "yes" part is that Koppelman is repeating Henry Monahan's formal argument for reconciling judicial supremacy with deference to agencies, as to questions of law: judges exercising their supremacy may decide that the law itself just is that agencies decide what the law is (within the bounds of silence or ambiguity). The "no" part is that this reconciliation is in essence a convenient legal fiction that lightly disguises the long-term retreat of law's empire. A measure of how far we have come, from my baseline of Crowell v. Benson (1932), is to imagine asking Chief Justice Hughes whether he would be satisfied with Monahan's and Koppleman's reasoning. He would certainly be appalled, seeing quite rightly that the "government of a bureaucratic character, alien to our system" that he attempted to prevent in Crowell is not any the less triumphant, even if its official rationale is that law itself chooses to abnegate its office to the bureaucracy. It's a bit like going to London, seeing the Crown on all the mailboxes, learning that the Queen asks the Prime Minister to form a government, and then concluding that the Queen (in Parliament) is still sovereign. De jure this is so, but it is a sovereignty that has (in virtue of polite conventions) ceded most of its real power to other actors in the system.

            Let me turn now to David Dyzenhaus, whose contribution is keyed to a distinction between "law" and "legality." I have to say I find the distinction puzzling in the extreme. Indeed language itself rebels; one can say that something is "lawful but not legitimate" or vice-versa, but one cannot, in English anyway, say that something is "lawful but not legal." Sometimes, however, Dyzenhaus does seem to intend the distinction between legality and legitimacy after all, as when he attributes to the book "the claim that anything produced in conformity with constitutionally mandated principles is for that reason constitutionally legitimate." To read him this way, would, however do excessive violence to his repeated references to law and "legality."

            The best I can do is read the distinction as intended to track a more familiar distinction between the "thin rule of law" (formal law) and the "thick rule of law" (legality) -- with the latter read to mean standard Lon Fullerite criteria for law-like-ness, and to derive from what Dyzenhaus calls a "constitutional value structure," embodied in official action through processes of legalistic deliberation. The reconstructed point, in other words, is that legality is more than formal law. When Dyzenhaus says that "law regulates is own production" through legality, even in the administrative state, I take him to mean that thin or formal allocation of legal authority is one thing, whereas the actual operation of value-inflected legal reasoning in courts and other deliberative legal institutions is another. Again, I may have misapprehended all this, but will have to forge ahead.

            Conditional on that understanding, I don't know what in the book seems to suggest the sufficiency of thin formal lawmaking as an account of law's abnegation; that is certainly not my picture. The internal legal process of abnegation I have attempted to describe, extended over decades, is as thick as thick can be; it is not as though some bare statute or single precedential fiat, read literally and woodenly, brought about law's abnegation. Rather, the book claims, the process of abnegation was -- at least on the judicial side -- one of sustained argument within, through, and by means of all the apparatus of deliberative and value-laden common-law legality, argument conducted under a sustained commitment to principled consistency. On the judicial side, in two of my major examples from the crucial years of World War II, NLRB v. Hearst and Yakus v. United States, we see judges engaged in principled deliberation under a governing ethos of rational consistency with precedents, and working-pure the principles underlying those precedents, in a way that would have made Dworkin proud -- if he had read them. And the paradoxical result, I suggest, was the ever-accelerating abnegation of the power of courts, a continual shrinking of what Dworkin called the "heartland" of law's empire.

            Putting the same point another way, I entirely agree with the claim that Dyzenhaus sees as critical to legality: the premise that even in the administrative state "law regulates its own production." But what I do claim is that law decided to regulate its own production largely by making a well-deliberated, value-laden decision to outsource that production to administrative agencies -- to step aside. Like Koppelman, Dyzenhaus insists on the sovereignty of law, overlooking that law, exercising that very sovereignty, itself decided, by means thick deliberative processes, that its own sovereignty should become nominal and conventional, like the sovereignty of the Queen in Parliament, or the sovereignty of the Frankish Kings.

            An excellent illustration involves Dyzenhaus' own example -- "privative clauses" in legislation that expressly bar judicial review, what American lawyers would be more likely to call "preclusion of review" clauses. Dyzenhaus seems to think that if courts interpret such clauses narrowly and with resistance to their obvious purposes, thereby retaining a substantial role in "saying what the law is" on the merits, it counts as an objection to the book's argument -- as counter-evidence. But this is precisely the set-up for the book's claim, the stage on which the play unfolds. That is because courts, exercising their power to say what the law is, on the merits, may decide to say -- and I claim that this is in fact what they have, for what most part, decided to say -- that law is what agencies decide it will be, within broad boundaries. That is, Dyzenhaus, like Koppelman, doesn't come to grips with Monahan's critical point about Chevron, which I emphasize throughout: courts, acting in thickly legal and highly deliberative ways, may decide that the law just is that agencies, not courts, will say what the law is. The sovereignty of law is not an alternative to abnegation, but -- in my account -- the precondition for it. (I note here but cannot discuss Nick Bagley's superb demonstration that in U.S. administrative law anyway, the presumption in favor of judicial review is much less robust than conventional wisdom among administrative lawyers would have it; that argument fits seamlessly with my own claims about how administrative law has developed on the merits).

Constitutional Theory

            Let me turn now from these threshold issues to some issues of constitutional theory. Jeff Pojanowski offers two critiques, alongside two larger points that I found illuminating. Let me begin with the critiques.

            First, Pojanowski discerns in the book a "premise of the radical inseparability of law and policy," which he also sees as widely shared in the legal academy. Let us call this "strong realism." The premise of strong realism is, however, not one the book assumes or requires. I have no doubt that there exists a domain of legal questions that are clearly separable from policy questions in a first-order sense (bracketing the second-order point that it is good policy to have such a domain of autonomous legal questions). What the book does argue is much more limited: that the sorts of complex questions at the frontiers of the administrative state, questions that, over time, increasingly occupy the attention of agencies and courts, tend disproportionately to have the feature that legal questions and policy questions, and indeed also procedural questions, are inextricably intertwined (whether or not conceptually separable). The consequence is that lines between law, fact, policy and procedure have not been rationally defensible over time; and the judges's own commitment to principled consistency has caused all these types of questions to be treated with increasing deference. A central strand in my story of the development of doctrine, for example, is that Crowell's line between judicial deference on factual questions and de novo review of legal questions proved untenable. More on this shortly.

            Pojanowski also critiques one of the book's central claims: that the originalist call for a return (insofar as possible) to the Constitution of 1789 and the classical separation of powers is self-defeating and incoherent, because the classical institutions, operating as they were designed to do, themselves quite deliberately generated the administrative state. Pojanowski says that the argument is "nonresponsive" --  the thought apparently being that if abnegation was (is) "unconstitutional," apparently on originalist premises, that's the end of the story, and it is irrelevant that some or all of the constitutional institutions have endorsed it.

            Yet this response itself seems entirely nonresponsive from my perspective. The argument at hand is that the original Constitution, functioning as it was designed to do through the thickly-legal processes of separated powers, itself brought about law's abnegation. Suppose with Pojanowski that the end state of this process is indeed unconstitutional on an originalist theory. That would merely underscore the book's point. It would show, in other words, that the original Constitution was incoherent at a deep level, because its unfolding according to its intended operation brought about an invalid state of affairs. That is not to offer an originalist rejoinder to my claims; if anything it indicts the coherence of originalism, and indeed of the Constitution itself. As Mark Graber puts it in his review, which got all this exactly right, "constitutional institutions in their pristine form provided political actors and judges with reasons and incentives to abandon originalism.... Law’s Abnegation convincingly asserts that if institutional arrangement A in the normal course of operation produces institutional arrangement B, and institutional arrangement A is legitimate, then institutional arrangement B is legitimate." Graber might have added the contrapositive: if one denies that B (the administrative state) is legitimate, then A (the original Constitution) isn't either.

            In all this there is a separate issue, of course, about my assertion that the original Constitution operated as it was supposed to in the course of generating the administrative state. The alternative view is that the administrative state came about through some sort of "abdication" of constitutional functions. I argue against that position at length in the book (see, for example, pp. 43-47), in part on historical grounds by drawing upon Jerry Mashaw's exhaustive treatment, in part on logical grounds by showing that the view of "abdication" needed to make this argument run would end up proving far too much; it would prove that the relevant institutions never functioned properly, thereby impeaching the original Constitution all over again. Let me again quote Graber, who gives a laudably precise formulation of the idea:

If Merlin by magic returned the American constitutional order to its pristine form, historical institutionalism and Vermeule’s work suggest that the same reasons and incentives that produced contemporary judicial review and contemporary administrative law will reproduce those practices in the very near future. The problem, if a problem exists, is one of constitutional design rather than a problem of individual decision makers betraying the commitments of a constitutional order. (My emphasis).

            All that said, I found Pojanowski's larger analysis illuminating. It rests on two points. The first is that Law's Abnegation explicates essentially unwritten constitutive principles of the administrative state. That's true, in the dual sense that to a surprising degree our administrative "law" is actually constituted by unwritten conventions (as explored here, and in important work by Emily S. Bremer), and that the long-run process of abnegation I describe is given forward thrust by unwritten principles -- rational consistency chief among them -- that have caused judges to collapse the unstable Crowell v. Benson framework by moving, over time, toward ever-greater deference. There have been and will be backward eddies and countercurrents, but the trend over time is unmistakable.

            The second illuminating point is just that the doctrinal instability of the Crowell compromise is indeed the crux of the issue. If I have a global complaint about the symposiasts, it is that they did not take the internal legal argument of the book seriously enough on its own terms. The key moments in the book are doctrinal moments -- as when, in NLRB v. Hearst, the Court came to believe, that the reasons or principles that counseled deference to agencies on questions of "fact" also counseled deference to agencies on "mixed questions of law and fact," so long as the agency had "a reasonable basis in law." (Here the specter of Chevron begins to hover over the scene). Pojanowski, however, does underscore the importance of the doctrinal issue. He stakes out the view that the Crowell compromise is not inherently unstable and can be shored up, and that deference need not overrun the levy between questions of law and questions of fact (as it did in Hearst). Quite appropriately, he doesn't explain his view at length; a symposium response was not the right occasion for doing so. I fully agree that the question is critical and look forward to hearing the developed version.

            Jack Balkin's entry makes a great many points about constitutional theory, too many for me to do justice to all of them. Some of them I have more or less addressed already. Let me single out one for further discussion here, and discuss another in the next section.  In a version of Pojanowski's objection, Balkin thinks that I mix up "causation and justification" by arguing that it is incoherent to damn the administrative state and call for a return to the original Constitution, when the original Constitution, operating as it was supposed to, itself produced the administrative state. But that argument is itself a point about justification, albeit a negative point; it is an argument that the justifications for that common stance -- administrative state illegal and bad, original Constitution binding and good -- are internally inconsistent. Note well that critics of the administrative state who call for a return to the original separation of powers, and the original Madisonian schema generally, almost invariably give reasons why the original scheme was well-designed, better-designed than the administrative state, to which they attribute a host of pathologies. If, however, the operation of the original constitutional scheme itself generated the administrative state, as a chrysalis gives rise to the butterfly, then that combination of views is fighting a civil war with itself -- precisely at the level of justification. Arguments in praise of the original Constitution are themselves, by a kind of transitivity, arguments in praise of the administrative state, which the original institutions generated in the proper exercise of their functions, or so I argued.

External vs. Internal Perspectives; Politics; Law Outside the Courts

            Finally, let me turn to some threads in the responses that question the shape and limits of the book's project. That project is to examine (1) law within courts from (2) an internal legal perspective. Two of the symposiasts, Balkin and Gillian Metzger, chafe at those limitations. Balkin questions the second, and indicates a wish that the book were written (entirely?) from the external perspective of politics and economics, rather than from the internal perspective of legal doctrine. But I like doing cases and don't see why I have to be confined to one register. Read in conjunction with other books of mine, which make all sorts of external arguments of the sort Balkin mentions, this book offers an argument in the alternative, one that adopts the Dworkinian-legalist-internal assumptions of so many administrative lawyers, and attempts to show that even given those premises, my view of the administrative state follows. It tries, whether or not successfully, to speak to the internal-legalists in their native language.

            Incidentally, that language includes and has always included internal arguments about the comparative competence of court and agencies (which I express in my own local dialect by speaking about the marginal costs and benefits of additional judicial review of agencies). Balkin seems to think such arguments are necessarily external, when in fact they are common within courts and from the judicial and lawyerly standpoint, as well as from an external standpoint -- as Hart and Sacks and the whole Legal Process tradition illustrates. Arguments from comparative competence amount to one of the areas of common ground between internal and external perspectives, which doubtless helps to account for their enduring appeal.

            Gillian Metzger, by contrast, questions the first self-imposed limitation of the project, the decision to focus on law within courts. The motivation for that, as Metzger explains, was to bring the insurgency to the heartland of law's empire, which has always been the courts and the law announced and applied by judges, as Dworkin himself proclaimed. Metzger is entirely correct that there is important law applied within agencies and elsewhere, outside of courts. Yet she says nothing about the important interaction between the two domains of law, which the book emphasizes in Chapter Six. Changes in the law applied within courts themselves affect the law applied outside of courts, within agencies -- partly by the law of anticipated reactions on the part of agency lawyers forecasting litigation outcomes, partly just because judges' view of the law is always influential for lawyers, wherever they are placed. Thus the advent of Chevron, for example, seems to have reduced the overall importance of lawyers as a profession in the ecology of the agencies. So it is also true that one cannot focus on the law outside courts without attending to the law within courts. I suspect Metzger and I agree on all this.

            Metzger's second point resembles Balkin's wish for a more external version of the project; she asks about the effect of politics on the developing legal logic of abnegation. What if, for example, President Trump were to appoint a critical mass of anti-deference Justices? I have given my reasons elsewhere for being skeptical that this will happen; but what if it does? What if, in the limiting case, the Court were to overturn Chevron, for example? Graber, likewise, asks whether Trump signals a change of political regimes that might change administrative law as well.

            No one would be mad enough to claim that the arc of law's development is immune to external political currents. What I do claim is something far more modest: after an initial period of legalist triumphalism, and a further period of confusion and adjustment, abnegation and deference will reassert themselves insofar as feasible, within whatever political constraints and externally imposed boundaries are set on judicial decisionmaking. As sure as the sun will rise tomorrow in the East, federal judges will defer de facto to EPA's construction of ambitious statutes in complex rulemakings, even if the judges are constrained to speak as though deference is impermissible; they are on average too busy, too modest, too  concerned about making a hash of environmental policy, and too aware of their own limitations to do otherwise. That sort of deference long predates Chevron and would live on after it. A sort of reverse analogy here is to strict scrutiny, often described as "fatal in fact"; in the sequence I describe, judicial-review doctrine would be strictly de novo in theory, deferential in fact. The internal legalist process of law's abnegation is hardly indefensible, or inevitable; its course can be turned by main force and with the right technology, just as a river can be redirected with enough work. Yet, like a river, it will always display a strong tendency to return to its original channel.

Adrian Vermeule is the Ralph S. Tyler, Jr. Professor of Constitutional Law at Harvard Law School. You can reach him by e-mail at adrian.vermeule at

Older Posts
Newer Posts