In her Foreword,1 Professor Gillian Metzger portrays the administrative state as laid under siege by an array of judicial, political, and academic attackers. Expertly curating and deftly dissecting a century’s circus of intellectual debate and political conflict, the Foreword demonstrates the myriad ways in which today’s struggles over administrative government reprise the turmoil of the New Deal period.
Indeed, the parallels between the present moment and the 1930s may extend further than she draws them. The history of that era suggests how the “rhetorical antipathy”2 towards the administrative state that Metzger carefully documents and critiques may yet cross over from the realm of rhetoric to the realm of reality. That, of course, only makes it that much more urgent to answer the central question addressed by the Foreword — the question of how to respond to the “anti-administrativist”3 complaint that the federal bureaucracy is extralegal, unconstitutional, and tyrannical.
Metzger’s response is the provocative rejoinder that the administrative state is not merely constitutionally permissible and not merely constitutionally beneficial, but also constitutionally obligatory.4 This argument diverges in critical respects from long-held conceptions of the administrative state’s constitutional status and role. It is bold in its premises and startling in its possible implications. It aims to break the siege — to quell, at once and en masse, the renascent attacks upon administrative government. But her argument for a constitutional obligation of administrative government pivots upon the threshold assumption that the Supreme Court will continue to regard broad delegations as constitutionally permissible — a point about which I do not feel as sanguine. And even if delegation doctrine persists in its present form, the full contours of the contingent constitutional obligation posited by Metzger seem to me to be both potentially enormous and — at the same time — hard to trace with precision. At the brass-tacks level, it is difficult to map out what exactly honoring the constitutional obligation of administrative government would require in the many and varied contexts in which it might be pitted against countervailing targeted arguments that regulatory power ought to be restrained. Politicians, scholars, lawyers, and judges gave us the modern administrative state; whether we can keep it remains to be seen.
I. Rhetoric and Reality
Metzger situates the current mood of anti-administrativism in its early twentieth-century roots, in the struggle over the New Deal.5 During the 1930s, the constitutionality of the burgeoning administrative state was hotly contested, in particular by the Liberty League, a coalition of businesses funded by the du Pont brothers.6 The landslide 1936 reelection of President Franklin Delano Roosevelt, the unpopularity of the League, and the Court’s change of constitutional course at the close of the 1930s ended that fight for a time.7 But that hiatus is now over. Politicians are now launching multipronged campaigns against the apparatus of administrative government, while an accumulating mass of academic writing challenges the constitutionality of fundamental features of the regulatory state.8 Perhaps most importantly, the Justices seem to be reentering the fray too. Today, flights of “rhetorical excess[]”9 and “strong rhetorical condemnation of administrative government”10 have featured in concurring or dissenting opinions by no fewer than four members of the current Court.11
What should we make of such judicial opinions “decry[ing] the dangers of the ever-expanding administrative state,”12 when — as Metzger stresses — their “bottom-line impact does not match their polarizing rhetoric”?13 Such opinions seem to bash the bureaucracy for no evident “practical gain”;14 but does that fully capture what these opinions may signify?
Like Metzger, I believe that much light can be cast on our present situation by looking back to the landmark battles between progressive and conservative Justices in the early part of the twentieth century.15 The dissenters of that period — Justice Holmes chief among them — famously argued that the extant jurisprudence of the Lochner era16 was anachronistic, unprincipled, and illegitimate. In opinions that set the bar for future dissents, they systematically hacked away at the philosophical underpinnings of that era’s jurisprudence,17 whilst also writing extramurally so as to leave no doubt as to their discontents.18 These Justices may not always have been consistent,19 and they may not always have been correct.20 But what they most certainly were was in deadly earnest.
Today’s dissenters21 should be seen as the mirror image and foil of those Lochner-era dissenters, who toiled so long and so assiduously until their point was won by a later and more fatal majority. That today’s Justices have not fully embraced every logical consequence of their views does not mean they do not hold the views.22 The litigants’ arguments and the rationales reached by the lower courts may hem in what a particular Justice deems proper to say in a given case. The modern dissenters, in other words, are applying their principles in the context of particular cases, and they are embracing the logical consequences of their views to the extent they feel it appropriate to do so as a matter of judicial craft.23 On top of that, anti-administrativist opinions play other important roles. For one, they send signals to litigants, who then bring lawsuits that do place more fundamental questions squarely at issue, and to judges on lower courts, who may then start to “percolate” particular questions more aggressively.24 For another, on those occasions when an anti-administrativist opinion is able to command a majority — perhaps precisely because of its modest “bottom-line impact”25 — then that opinion may begin the process of a death by a thousand cuts, which is a time-honored way to kill off a disfavored doctrine.26
There is not much in any of this to reassure supporters of administrative government. Conversely, to opponents of administrative government, the timeline of the decline of Lochner-era jurisprudence might teach the lessons of patience, resolve, and staying the course. When one begins the work of undermining an entire body of jurisprudential thought, one may not live to see the whole thing come crashing down.27
Viewed in this historical light, the Roberts Court’s recent expressions of judicial anti-administrativism may appear less a rhetorical threat — and more a real one — than their immediate lack of a “bottom-line impact” would otherwise indicate. Viewed in the light of the present day, two additional factors external to the Court may give added impetus to the Justices’ anti-administrativist impulses. For one, this happens to be a time when many (more) Americans are experiencing an instinctive or “almost visceral”28 resistance to claims to strong executive branch power. President Trump’s Administration may prove to be a great boon to those who would portray the administrative state as tyrannical,29 unaccountable,30 unconcerned with individual rights,31 or capable of destroying Congress’s handiwork.32 Some of the underlying themes of anti-administrativism, if not all of its precise recommendations, may find voice on the left as well as on the right. For another, the currently sitting President may not prove to be very effective at accomplishing the project of sweeping regulatory reform to which he has committed. Earlier Presidents have signally failed at that task.33 While one might construe that failure as a symptom of the absence of an underlying widespread political resolve to “deconstruct[] . . . the administrative state,”34 a President’s unwillingness (or inability35) to translate anti-administrativist rhetoric into reality may just deepen some anti-administrativists’ conviction that the administrative state cannot be politically “tamed”36 and that a constitutional remedy from the bench is therefore required. Either or both of these dynamics might tend to push judicial anti-administrativism out of the periphery of concurrence and dissent and onto the main stage.
If, however, Metzger carries the day with her most intriguing claim — that the administrative state is constitutionally obligatory — then all these quickening crosscurrents of judicial anti-administrativist sentiment, whether rhetorical or sincere, won’t matter much; the Constitution, if it is understood as Metzger argues it ought to be, will curb and contain what anti-administrativists can accomplish.
II. The Administrative State’s Connection to the Constitution
In order to understand Metzger’s constitutional argument, it is useful to begin by identifying how her account departs from another idea that connects the administrative state with constitutional law — the idea of the “administrative constitution.”37 The basic notion of the “administrative constitution” is that core aspects of administrative law — including the Administrative Procedure Act38 (APA), the key precedents and conventions that shape agency action, and other transsubstantive doctrines and rules such as open-government laws and presumptions of judicial review of agency action — have come to supply the administrative state with an internal structure and a framework of control that operate essentially as a constitution for regulatory government.39 By “legitimating, through controlling rules and procedures, the exercise of power over private interests by officials not otherwise formally accountable,”40 the administrative constitution performs the function of an “unwritten”41 or quasi-constitution for the regulatory state.42
The important thing to remember about this type of quasi-constitutional law is that the chief answer to the questions “Where does the administrative constitution come from? Who made it?” is Congress.43 The administrative constitution is akin to the kind of quasi-constitutional law most famously presented by Professors William Eskridge and John Ferejohn in their book, A Republic of Statutes,44 which explains how framework or “super-statutes,” elaborated by courts and agencies, have come to play a constitutive role in the fabric of American law and government despite the fact that the statutes are not embodied in the text of the Constitution.45 The administrative constitution is rooted in the enactment of such statutes — most importantly the APA46 — and in the meaning and conventions encrusted around those statutes by agency practice and judicial elaboration, including by watershed holdings such as Chevron47 and State Farm.48 Needless to say, the administrative constitution is nowhere codified as such, nor is it “formally entrenched against change.”49 But rather than rendering it vulnerable, those very features have seemed to make the administrative constitution that much more resilient and adaptable.50
The notion of the administrative constitution has been a helpful organizing framework. But the fragility of that model has become increasingly evident. For one thing, it leaves the administrative state vulnerable to shifting political coalitions and erosion of legislative support for administrative government. If Congress made the administrative constitution, then Congress can break it.51 For another thing, the now-evident permeability of the administrative constitution has made it seem not as robust a framework as it was long assumed to be. The administrative constitution is riddled with “outs” and “loopholes.”52 Many techniques — including waiver, delay, nonenforcement, and dealmaking — allow the executive branch to lawfully avoid the constraints of the administrative constitution, to act unilaterally, without advance public input and without judicial review, even though the resulting policies affect millions of people.53 The growth of OIRA review — an executive branch procedural innovation54 — has made it harder to credit the notion that Congress is the ultimate steward of important administrative procedure.55 And Congress itself has tinkered with the ground rules at times, by enacting on an ad hoc basis broadly worded preclusions of judicial review to insulate consequential executive branch policymaking from being examined by courts.56 As unconventional uses of administrative power have expanded in scope and consequence, the exceptions to the administrative constitution have seemed to swallow the rule;57 ordinary administrative law seems to have become a “lost world.”58 In short, because the administrative constitution has come to appear both politically vulnerable and outmoded by administrative practice, its utility as an organizing framework has seemed increasingly questionable.
By breaking with this framework, Metzger’s account might offer remedies for these worries. She derives the backbone of the constitutional case for the administrative state from Article II as well as Article I. Her emphasis is upon how the Constitution secures the executive branch’s capacity to govern in an effective and accountable way, not just upon how the Constitution constrains administrative government.59
With this one stroke, Metzger’s account would counter both frailties of the model of the administrative constitution. First, by anchoring the administrative state to the executive branch instead of only to Congress, Metzger’s account would help to shield the administrative state against swings in support in Congress for administrative government. If the administrative state is “constitutionally mandated,”60 if it is the “constitutional string” attached to delegation,61 then Congress may make administrative power as an initial matter — by delegating — but it might not then be able to unmake it — by withdrawing or cabining the personnel and resources that the executive branch requires to carry that delegation out.62
Second, Metzger’s account might rationalize the more exotic variations of administrative power that the model of the administrative constitution has struggled to defend. Indeed, Metzger’s account — important aspects of which she has also defended elsewhere63 — has the potential to flip the fundamental terms of the debate over such actions. Consider, for example, one of the Obama Administration’s deferred-action immigration programs and the dangling question the Supreme Court left undecided concerning its legality64 — the question, posed by the Court itself, “[w]hether the Guidance violates the Take Care Clause of the Constitution.”65 While the lower court and some scholars have been engrossed in debating whether the deferred-action policy ought to have been promulgated in compliance with the ordinary procedures for notice-and-comment rulemaking,66 or whether it comported with the values that the “small-c” administrative constitution seeks to shield,67 Metzger has swept above that skirmish by placing this policy on the plane of the “large-C” Constitution. Such a prospective nonenforcement policy, Metzger has argued, has “strong constitutional roots”68 because it enables the President to perform his Article II–based duty to supervise the execution of the laws.69 Given the de facto delegation of immigration lawmaking power to the executive,70 and the President’s obligation to supervise how lower-level officials enforce that law,71 Metzger has contended, it would raise constitutional problems to forbid the adoption of categorical and prospective programs of non-enforcement in the immigration context.72 In other words, the question that the Court posed — whether the Guidance violated the Take Care Clause — had it almost backwards; instead, the correct question to ask was whether the Take Care Clause required that the President be able to issue the Guidance.
III. The Constitutional Obligation of Administrative Government
Had the Court posed the question thus, the challenge would still remain how to answer it — and how to answer the analogous questions that will be raised in myriad contexts across the regulatory state. The Foreword’s answer is simple — and radical. The project of constitutional analysis must be “reorient[ed],” Metzger urges, to encompass not merely “constitutional constraints on government but also constitutional obligations to govern.”73 That obligation to govern, she contends, means that we must have an administrative state, “with its bureaucracy, expert and professional personnel, and internal institutional complexity.”74 The obligation to ensure “accountable, constrained, and effective” administrative government, she contends, may require that the executive branch be able to promulgate regulations, conduct policymaking, adjudicate cases, and oversee the combined exercise of legislative, adjudicative, and executive functions.75 Giving due weight to the duty to govern may require judicial deference as well, both because that deference gives effect to a congressional decision to delegate power to an agency,76 and also because courts must defer to agencies if courts are to “reach accurate, coherent and consistent determinations” in cases requiring “expert elucidation” of specialized statutory schemes.77 Honoring the duty to govern may also impose limits on congressional power; given that the Environmental Protection Agency (EPA) has to be able to faithfully execute its delegated powers, she contends, it may “begin to look constitutionally suspect” for Congress to “massively underfund[] the EPA” or to “repeal[] environmental rules necessary to implement delegated authority without adopting an alternative enforcement regime.”78
In sum, the Foreword claims that a proper understanding of the Constitution makes the administrative state “constitutionally obligatory”;79 it contends that given “the phenomenon of broad delegation,” the constitutionality of which is “not at risk of judicial invalidation,”80 the administrative state must also follow as the “constitutional consequence[] of delegation.”81 The executive branch may not always get what it wants; but — at least once it is delegated power — it must get what it needs.
On two scores, these claims give pause for thought. First, to conclude that the administrative state is constitutionally obligatory in nearly any respect would entail a sharp break with constitutional practice and historical understandings.82 The Foreword emphasizes that there is scant support on the current Court for holding delegation unconstitutional,83 but no support seems to have been voiced for the view that when Congress chooses to delegate, the Take Care Clause then places an affirmative duty on Congress to supply the executive branch with the financial resources needed to govern effectively or to enforce the laws adequately.84 Nor has Chevron deference been treated as a “necessary consequence”85 of delegation, if we take that claim to mean that it is constitutionally obligatory for courts to presume that when Congress delegates, it intends to delegate to agencies the authority to resolve ambiguities in the statute;86 clearly, the Court regards itself to be empowered to deny deference or to cut back on the boundaries of Chevron’s domain, and — as evidenced by the existing system of enforcing the federal criminal code — even broad delegations of authority to implement statutory provisions have not been treated as mandating judicial deference to the agency charged with implementing the scheme.87 Nor has the principle of separation of powers yet been treated as requiring either the creation or the continued existence of internal administrative constraints,88 such as expert government employees or an independent civil service, even though these elements of administrative government may help to prevent presidential unilateralism and ensure that the executive branch will exercise delegated power in a legally or politically accountable manner.89 In advancing these propositions, the Foreword urges shifts in constitutional law that could both greatly increase executive branch power and also sharply rein in unilateral presidential power. But it refrains from supplying a meta-theory for evaluating when changes in the constitutional regime of such potential scope and consequence are warranted. Nor does it expressly address why one should feel confident that, as an empirical matter, these cross-cutting effects would, on net, leave us with an adequately constrained executive branch and President.
Second, the claim that the administrative state is constitutionally obligatory pivots on a threshold premise that some will find hard to stomach: the proposition that delegation is both constitutional and “such a fundamental and necessary feature of contemporary government that it is mandatory in practice.”90 After all, the conviction that delegation is unconstitutional is the igniting spark, if not the entire engine, of anti-administrativism; the idea that delegation is “fundamental,” or “necessary,” or “mandatory in practice,” is precisely the kind of idea that many anti-administrativists most fiercely contest. From the early twentieth-century history of political contestation about the administrative state, many have gleaned the message that Metzger draws — that those who do battle against delegation’s constitutionality are fighting the last war.91 But the history of the 1930s equally may be read to demonstrate the huge scale of the transformations in constitutional law that can occur because of a shift in the Court’s stance following the election of a President bent on making a radical break with extant constitutional understandings.92 Today, the sitting President was placed into office in part by campaigning on an anti-administrativist platform,93 and he may have the opportunity to appoint hundreds of judges that share his anti-administrativist aims.94 To assume that the status quo on delegation will persist is to elide a key lesson of the 1930s — that constitutional revolutions happen.
Let us set these points aside, and stipulate that the courts will continue to regard broad delegations as constitutional,95 in order to probe the Foreword’s contention that, given delegation, the administrative state comes clothed with affirmative constitutional protections.96 Put another way, if we take it as a given that delegation is constitutional and a necessary feature of government, what are the kinds of arrangements that are also constitutionally necessary to ensure that the executive branch has the necessary tools to carry out its duty to govern accountably and effectively? For future executive branch lawyers who want to argue for the validity of a given type of regulatory power, or for future litigants who wish to attack some perceived agency overreach, what will have to be demonstrated to establish (or to defeat) the claim that a given aspect of administrative law or practice is a constitutionally required “consequence” of delegation?97
Perhaps the broadest way to implement Metzger’s account (and this is my own formulation, not one proposed by Metzger) would be to view Article II as implicitly giving the executive branch something like its own “necessary and proper” power.98 On this view, anything that the executive branch deemed necessary and proper to carrying out a delegation would be deemed constitutionally protected,99 and the executive branch would receive deference in determining any “factual predicates necessary to conclude that a given action is appropriate.”100 This approach would effectively extend an umbrella of constitutional protection to anything that the executive branch did that was plausibly connected to implementing its delegated powers — whether that was rulemaking, guidance, waiver, prospective nonenforcement policy, OIRA review, or in-house adjudication. This approach would, as well, mandate deferring to the executive branch’s interpretation of statutes and to its own regulations, because such deference would promote the value of democratically accountable executive branch control over government.101
Adopting that approach would have the benefit of simplicity. It would, however, also have the uncomfortable consequence of ousting Congress almost entirely from the role of specifying how administrative government ought to operate — a result that sits uneasily with historical understandings and democratic principles. It would also carry risks; as Professor Jack Goldsmith and Dean John Manning have pointed out in reference to James Madison’s cautionary remarks on Congress’s necessary-and-proper power, “one can always spin out ways in which remoter and remoter means can be related to some broadly framed end.”102
I doubt that Metzger would find appealing that implementation of her account,103 and I know that I do not.104 If, however, one attempts to construct a narrower formulation of what the executive branch requires in order to govern, then some difficult line-drawing problems arise. Begin with rulemaking. Certainly, to implement a statute as complex as the Affordable Care Act (ACA), many regulations must be promulgated. But what if the Department of Health and Human Services were to use that rulemaking power to create exemptions to statutory requirements for entities who had philosophical objections to the ACA’s rules — say, to the rule requiring the provision of contraceptives without cost sharing?105 Such an exemption might reflect the executive branch’s considered conclusion that effectively carrying out its regulatory mission requires it to draw the rules with large holes in them, in order “to reduce and relieve regulatory burdens and promote freedom in the health care market.”106 Is the power to craft such an exemption, for such a reason, a part and parcel of — a “constitutional string” attached to107 — the initial delegation?
Turning to adjudication, consider the SEC’s recent shift away from courts and toward administrative settlement, which Professor Urska Velikonja has described.108 As Velikonja explains, rather than face the scant handful of federal judges willing to ask rude questions about its settlement choices,109 the SEC moved a large proportion of its administrative settlements in-house, with the consequence that courts did not review them until after they were finalized.110 For the SEC to have the capacity to settle cases is surely valuable;111 and dealing with such settlements in-house may enhance the ability of the SEC’s leadership to steward and supervise the administration of the securities laws. But does the SEC also have to have the power to reach so many of its settlements in-house, without meaningful judicial review before the settlements are finalized? Is such a power constitutionally obligatory, too, as a “string” attached to the SEC’s delegated authority?
One or two of the regulatory reform ideas that are now under consideration would also pose tough questions. Consider, for example, the proposal that Congress enact legislation requiring all rules to sunset after fifteen years.112 Would such a law be permissible in a world in which the “administrative capacity” necessary to carry out delegations was treated as constitutionally mandated?113 If the law is viewed as a post hoc limit upon the scope of an earlier delegation — akin to a statutory restriction on whether an agency can adjudicate cases or not, or issue binding rules or not — then perhaps it would be entirely permissible.114 On the other hand, a regulatory sunset law would surely consume agency resources by forcing ongoing and costly reconsiderations of slews of earlier-promulgated regulations. Holding constant the resources that Congress supplied to the agency, such a law would hinder the agency from executing the other functions that Congress has also delegated to it — a consequence that, on Metzger’s logic, may be “constitutionally suspect.”115 If that level of hindrance were sizeable, would this new string on delegation snap, or would the older strings give way?
The point is just that recognizing a constitutional obligation to govern contingent upon delegation still leaves us with the challenge of tethering particular administrative procedures, or particular degrees of executive branch latitude, to the initial anchor of a delegation. Many things might be consistent with carrying out a delegation; but saying that any are obligatory is far more difficult. Judges (or politicians or scholars), even those wholly convinced of the benefits of the modern administrative state, may hesitate to regard any particular facet of administrative government as a necessary consequence of delegation, as a feature that Congress cannot defease or that courts must ratify. That, in turn, would leave ample room for the accretion of various anti-administrativist challenges and legislative cutbacks, the sum total of which may eventually resemble the wholesale rejection of administrative government for which some voices plead.116
To state the obvious, the questions just raised have largely been questions of application and implementation. They have sought to specify, rather than to deny, the Foreword’s highly original claim: the proposition that the obligation to govern should carry affirmative constitutional protections, not just constitutional constraints. Constitutional principles, new and old, gain their substance from the type of iterative line-drawing that comes from posing and answering such questions. And there is no escaping the urgency of the Foreword’s core message — that supporters of administrative government would do well to devote their energies to the task of making an affirmative case for the regulatory state, instead of just defensively responding to attacks on it.
Conclusion: Of Pennies, Pounds, and Pence
The Foreword portrays the current moment as a “redux” of the 1930s. I agree; and I think, too, that adopting that viewpoint places into its proper perspective Metzger’s own intervention. In it, one can discern echoes of the words spoken by President Franklin Delano Roosevelt in a barn-burner of a speech he delivered at Madison Square Garden in New York, in the fraught days just before his landslide reelection in 1936. For a dozen years before he took office, President Roosevelt said, the nation had been “afflicted with hear-nothing, see-nothing, do-nothing Government.”117 “Powerful influences,” he warned, were “striv[ing] today to restore that kind of government with its doctrine that that Government is best which is most indifferent.”118
In its reframing of the constitutional stakes of the debate over the administrative state, the Foreword seeks to accomplish for administrative law something akin to what President Roosevelt sought to accomplish for the idea of government and all it might accomplish. Instead of seeing that administrative state as best which is most enfeebled and most externally limited, Metzger urges us to see that administrative state as best which is most energetic, most internally accountable, and most effective at governing. Today, as when President Roosevelt spoke in 1936, giving full effect to Metzger’s vision would require a sharp shift in the “reigning constitutional discourse” — now, a shift away from questions of constitutional constraints on administrative government and toward the “scope and nature of constitutional obligations to govern.”119
It is there, then, that I end, very nearly where I began, with a reluctant word of caution on the potential shifts in “reigning constitutional discourse” we may now face — a word on pennies, pounds, and Pence. The Foreword’s contention is that if you are in for a penny — with delegation — then you are in for a pound — the administrative state. But if delegation does come with strings attached, and even if one is fully persuaded that these are the strings, the fact would still remain that the administrative state is only constitutionally obligatory if delegation is constitutionally permissible. Future Justices in the mold of Justice Thomas or Justice Gorsuch may rather take back the penny than pay the pound.120 There is, after all, no Franklin Delano Roosevelt sitting in the White House today. And we may be (at most) one President Pence away from seeing the Supreme Court controlled by a majority of such Justices.
* Professor of Law, University of San Diego School of Law. Many thanks to Adam Cox, Chris Egleson, Richard Fallon, Daryl Levinson, David Pozen, and Adam Zimmerman for helpful comments and conversations, and to the editors of the Harvard Law Review for their excellent editorial work.
Recommended Citation: Mila Sohoni, Response, A Bureaucracy — If You Can Keep It, 131 Harv. L. Rev. F. 13 (2017), https://harvardlawreview.org/2017/11/bureaucracy-if-you-can-keep-it/.