What sort of political system do we have in America? Why, republican self-government, of course. Government of the people, by the people, and for the people, as mediated by our hallowed constitutional institutions. Doing our thing for 227 years now.

A great many people aren’t buying that story these days. They look at our actual manner of governing and see bitterly contested symbolic elections awkwardly joined to a powerful technocratic administrative state that makes most of the important decisions. The disconnect between the ideal and the reality has brought us a legitimacy crisis that has been often touched upon in these virtual pages lately.

Following in the long tradition of the original progressives, some people observe how far our administrative practice has diverged from our constitutional theory and then conclude that the quaint old theory must go. Democracy, taken at its word, doesn’t look so great for anyone, and so it is better to put our trust in those who know the most, imperfect though they may be. Some are willing to argue that the best thing would be to make a clean break, chucking out such obsolete baggage as the Congress.[1] More often, those who see the administrative branch as our most knowledgeable and healthy branch offer elaborate reconciliations of technocracy and our democratic heritage and urge citizens to take comfort in the ways in which our current dispensation is formally open to their plaints.[2]

Given the mood of 2016, both the naked and democratically clothed justifications of our technocratic administrative state come off as rather willfully oblivious: angry citizens who want government by the people are unlikely to be placated by being told by regulations.gov that they can “Make a difference. Submit your comments and let your voice be heard.” There is a profound lack of trust, a sense that in practice the administrative state is insular and uninterested in the concerns of ordinary people, even if it has to check the box of procedural openness. Instead, many on both the left and the right see agencies as nested in an elaborate system of crony capitalism they lack the perspective to overcome. People are angry, and they want their government back.

Populist Escapism

Unfortunately, this often-justified anger is mostly channeled into unproductive outlets that, ironically, are likely to worsen our current dysfunctions. This is especially so for the populist impulse that has surged to the forefront of our politics. The common thread uniting populists left and right is the idea that if the people could shake off the yoke of iniquitous elites, they would naturally govern effectively. Self-serving interests prevent that from happening, entrenching the status quo by colonizing the mediating institutions, which by design are meant to serve the people.

There is more than a little truth to the populists’ criticisms, but their ideas about the natural state of things if only they could strip out those infected intermediating layers are utterly fantastic. In their mind, these institutional encumbrances will recede from the path of the people’s righteous march forward and the public good will be realized immediately: without delay and without being corrupted.

This Rousseauian dream has been with us a long time, but its central flaw can never be fixed: there simply is no general will from which a government can faithfully take dictation. Government cannot simply be a portrait-in-miniature of society, especially when its work extends into areas with which most people are completely unfamiliar. Even when citizens do have views that could be mirrored, there is no perfect way for government to ascertain their presence and intensity. And there is no chance at all that simply taking on the people’s will through osmosis could result in an internally coherent (or even non-contradictory) web of laws. Politics may well have the potential to be corrupting, but it does effectively allow collective decisionmaking about what our government should do and how.

Populists have no substitute, and so they are likely to create a vacuum in which technocrats prevail. Democratic idealism goes hand in hand with cynicism about demo­cratic practice: people desire “a democracy too good for politics,” and when reality disappoints “the demos turns to fugitive escape.”[3] Today, as was the case a century ago, populism ends up bundled with empowerment of “apolitical” technocrats precisely because they are willing to pretend that they can neutrally channel the public’s interests through the exercise of purely technical expertise.[4] Of course, in practice, politics will merely take different, less legible forms, and people will resent the subterfuge; we will be right back at the legitimacy crisis with which we began.

Libertarian Escapism

Libertarians, second to none in their hatred of politics, offer their own distinctive forms of escapism even as they congratulate themselves on their unusual sense of realism. Starting from the undeniable premise that voters are rationally ignorant and/or irrationally passionate about nearly all political matters, they jump to an end-point in which government keeps its nose out of all but the most crucial functions, thereby leaving the bulk of human actions free of the pernicious influence of politics.

But how to get from here to there? We live in a democracy full of fellow-citizens who think that “crucial functions” are very extensive indeed. One libertarian tradition is simply to flip these fellow-citizens the bird and insist that a cadre of the right kind of rulers—anti-technocratic technocrats, more or less—will somehow gain power and make our government best by governing least. That is undoubtedly an applause line in many rooms, but it is the equivalent of not playing the game at all.

Sophisticated anti-political classical liberals go beyond this hand-waving to offer mechanisms by which our current mixture of low-brow electoral politics and high-powered administrative decisionmaking could be dislodged, most often putting their faith in the courts. The rights-based vocabulary of constitutional law offers them an ostensibly apolitical sword to wield against government actions, one that can be said to serve a higher master than normal democratic politics. Taken up within the context of specific legal battles, this has proven to be one of libertarians’ most impactful entrées into active political disputes, whether or not they would characterize it that way.

It is less clear whether this way of thinking offers a path to wholesale change, but there are now a number of serious efforts in that direction. Notably, Charles Murray’s latest book waves the banner of government “By the People” and calls on the public to combine widespread civil disobedience to the administrative state’s edicts with maximalist legal opposition. The goal would be to overwhelm the federal government’s enforcement capacities, exposing it as a paper tiger and forcing a new age of genuine self-government.[5] He is confident a silent majority can be called upon to denounce the obviously harmful scourge of overregulation; I am dubious. In envisioning a revolutionary transcendence of the administrative state, Murray attempts to fuse a libertarian agenda with the current anti-political, populist mood, but in doing so he partakes of that populism’s weaknesses.

Accepting Politics

The deepest recent attempt to dislodge the administration-heavy status quo with an appeal to first principles is Philip Hamburger’s “Is Administrative Law Unlawful?”[6] A distinguished legal historian, Hamburger tries to reanimate ancient Anglo-American constitutional principles for a modern audience and show why they are so profoundly incompatible with the prerogative-like tendencies of our contemporary administrative state. By sensitizing the public to the ways in which our current arrangements effectively represent a return to our “preconstitutional past,” he hopes to shift the paradigm of thinking about the administrative state and lay the groundwork for a massive reorientation of our government. To his credit, he acknowledges that this is sufficiently revolutionary as to feel somewhat quixotic.[7]

But Hamburger can’t be accused of being anti-politics; indeed, the force of his argument relies on people understanding why democratic legislatures are superior bodies for decisionmaking relative to expert-led administrative agencies. He is worth quoting at length:

The administrative power substitutes the specialized knowledge of administrators for the specialized authority of the branches of government. In a strange way, therefore, the administrative combination of expert knowledge and consolidated power really does hark back to the medieval monarchical vision of a wise ruler, who knows what is best for his people, and who therefore must have the full range of unspecialized power to impose justice. The result is a government ill equipped to handle modern life. Rather than take advantage of the diversity, freedom, and epistemological openness of modern science and society, administrative law responds to these conditions with epistemological arrogance and consolidated power.[8]

This is a deep point. Because administrators style themselves as above-the-fray solvers of collective problems, it makes it very difficult for them to question their assumptions or open themselves to exchanges with critics on equal terms. Doing so would threaten their pose as “wise rulers” and show the political choices embedded in their thinking, and so their natural tendency is to hunker down—the very thing that has led to legitimacy problems over the long run, even as it fends off some short-term headaches by branding critics as “political” or as working for “special interests.” Legislatures, on the other hand, institutionally specialize in openness and the ability to find compromises even among people who remain in open political disagreement.

What is needed, once again, is a way to get from here to there. Hamburger often allows his historically informed contempt for the administrative state to blind him to the solidity of its political support in the present, and to overlook the ways in which our current legislature fails at precisely the functions he assumes are its specialty. This is particularly clear in his discussion of consent. To his mind, the legislature possesses legitimacy because people have meaningfully consented to follow its choices through voting; and this rationale is even more powerful today than in the distant past because of the greater inclusiveness of the franchise. I’m not sure whether the formalism of that account fits with people’s perceptions of reality, though. Consent is everywhere and always a legal fiction, but if there is a strong bond between the people and their legislature, it may be a sacred one that offers an enduring foundation for government. If that bond is weakened because people think of legislators as hapless or corrupt or both, the idea that legislators’ deals represent the public’s consent will come to seem grotesque.

Accepting Incrementalism

To move our government’s center of gravity out of the administrative state will require several steps. One is to challenge the pieties of modern administrative law scholars, who are as a rule staunch defenders of current arrangements. Hamburger’s book does valuable work in this regard. Another is to effectively catalog the status quo’s shortcomings when it comes to the insularity and high-handedness of the administrative sate. This must be done in a way that transcends shallow partisanship, as opportunistic arguments implausibly blaming the Obama administration discredit deeper attempts at understanding and should be vigorously contested.

Probably most important, though, is to show that Congress can actually do the work of an intermediary rather than just telling people it is meant to play this role. That means doing the work of incremental legislating, even when it is unglamorous and even when it requires working through suspicions that interests are working the system. Inevitably, they will be; everyone is an interest of some sort. Not performing this role successfully is the signal failure of Republican congresses over the past two decades, and especially since 2011. If Congress actively demonstrates itself to be a body devoted to zero-sum posturing, that is how people will think of it, making a technocratic administrative state look relatively appealing.

Not one of these efforts is aided by imagining that the administrative state will vaporize at some point if only its opponents get their ducks in a row and bring the right constitutional challenges or finally elect the right president. Indeed, pining for such abrupt transformation is the surest recipe for continued dominance of the administrative state.

Should we have allowed the administrative state into our common law home, so long ago? That question makes for fascinating counterfactual history, but its answer has only limited bearing on our present situation. The struggle for modern self-government, whatever that can realistically mean, entails disciplining and humbling the administrative state we have, not re-running more than a century of state development. The sooner the administrative state’s critics realize this, the sooner they can begin to restore the legitimacy of our government.

For an extended version of this argument, see Wallach’s recent Brookings paper, “The administrative state’s legitimacy crisis.”

I remember when I first became a libertarian. There was no Internet, just countless enthusiast flyers in the mail. One of them showed a bureaucratic overlord warning a subject, “The line in the road never shines brightly: I tell you where the line is.”

As we reckon with the administrative state’s “legitimacy crisis,” for which Philip Wallach’s “Farewell to the Administrative State?” advises “middle way” solutions, I must make a foundational observation: We are dealing with the fruits of the assumption of sweeping powers by Congress in the first instance, and the over-delegation of power to administrators atop that. Whatever the roots of the expansive administrative state (progressivism, neglect, populism, envy, cronyism, power lust; you can find books arguing each), lawmaking power left the house of classical liberalism long ago.

Granted, most want more government than I do; my own personal “Article I” limits Congress to picking a national bird. OK, I exaggerate; however a “national mammal” was just selected this week.[1] I’m the sort who appreciates Henry Hazlitt’s framing in “The Torrent of Laws”: [I]f the government confined itself to enacting a code of laws simply intended to prevent mutual aggression and to maintain peace and order, it is hard to see how such a code would run into any great number of laws.”[2]

That’s an alien idea today when legislative, executive and judicial one-upsmanship seemingly comprise an unbounded societal game of rock-paper-scissors. We would worry less about the “profound lack of trust” in our governing institutions if government remained confined to specific functions.

So what to do? Philip Wallach seems reconciled to the legitimacy of the administrative state. He seems to hold that its problems are not innate and that incremental reforms can make it efficient and responsive. We aren’t told exactly what solutions would be in his brief Cato Unbound essay, but his larger paper “The Administrative State’s Legitimacy Crisis” defines his “middle way” as one “in which statesmen lead with the guidance of experts and are made accountable to those parts of the public capable of meaningfully judging their results.”[3] That reads a tad like there are some members of the public from whom Wallach isn’t interested in hearing.

The Yes Minister joke goes, “If the right people don’t get power do you know what happens? The wrong people get it.”[4] But I am concerned less about the wrong people, and more about the existence of the power in the first place. James Madison wrote that if men were angels they wouldn’t need government;[5] and since they clearly are not angels, “statesmen” is not the label defining the kind of men that emerge in an unlimited power scenario. Nor, for that matter, do they emerge in today’s “pen and phone” scenario of unilateral executive action.[6]

Wallach asks, “Should we have allowed the administrative state into our common law home, so long ago?” He answers that we must settle for “disciplining and humbling the administrative state we have, not rerunning more than a century” of progressive rule by experts. I’m in favor of such incremental reforms, but I see the administrative state as authoritarian and irredeemable, and so I do want a “rerun.” Perhaps we can find a different “middle way” that doesn’t celebrate centrally managed health care at one extreme, and imposing dribble EPA gas cans and FDA serving sizes for breath mints at the other.

Rule of the Unelected

The abuses of the democratic “legitimacy crisis” cannot be described as accidental; they are conscious policy. President Barack Obama’s 2014 State of the Union pledge to use his “pen” and “phone” to implement a “year of action” with or without Congress became a taunt. Problems with a polluter? The EPA administrator said to “crucify them” to make them “really easy to manage.”[7]

The unelected do the bulk of lawmaking. Congress passed 114 bills last year that President Barack Obama signed into law. Outside the normal legislative process, though, agency regulations from the allegedly “most knowledgeable and healthy branch” numbered 3,410.

I call the large annual multiple the “unconstitutionality index.” Notionally, those rules undergo Administrative Procedure Act scrutiny, and the public gets ample time to comment. But over a third of agency rules are issued without a Notice of Proposed Rulemaking,[8] exploiting the “good cause” exemption by claiming notice is “impracticable, unnecessary or contrary to the public interest.”[9] Agency and Office of Management and Budget (OMB) cost-benefit analysis of the costliest subset does not actually exist; seven rules out of thousands had cost-benefit analyses in 2015, and 14 the year before.

Atop all this is what I’ve taken to calling “regulatory dark matter.” Agency memoranda, guidance documents, bulletins, circulars, notices (thousands annually), administrative interpretations, letters—a veritable galaxy of words—has become almost intractable and dispenses with democratic accountability altogether.[10] A Department of Labor blog post and “Administrative Interpretation” made many independent contractors employees;[11] New Housing and Urban Development guidance made landlord and home seller denial of those with criminal records a likely violation of the Fair Housing Act.[12] Some agencies have become self-contained little national governments, raising their own funds outside congressional oversight,[13] with bosses that even the president cannot remove.[14] Countless more examples could be added.

The Pretense of Expertise

Wallach also notes that some “argue that the best thing would be to make a clean break, chucking out such obsolete baggage as the Congress.” In short, the unification of executive and legislative power.

Shall we have an elected king? The technocrats are already willfully oblivious to the public.

The founders erected a different system, one recognizing the simple idea that most things are not public policy questions. Progressivism and the administrative state, the rule by experts, whatever one calls it, disagrees. As more of the economy falls to federal government rather than private sector management—retirement, medical choices, finance, insurance, energy, science funding and “manufacturing hubs,” critical infrastructure, control of lands and so on—the gap will grow. And progressives have done a good job entrenching themselves. Regulators have a gigantic head start in domineering frontier sectors like drones and driverless vehicles, since the government controls the airwaves and owns the roads, and will use that oh-so convenient fact to pander to emergent cronies.

Even as this “absolute imperious style”[15] threatens liberties and prosperity, administrators display a cavalier technocratic attitude. It’s not just that the FCC ought not turn the Internet into a utility via net neutrality, nor that public utilities enable rent-seeking or cronyism; rather it’s that the utility model—guaranteed profits and the prohibition of competition to replace competing networks—was the product of rent-seeking in the first place. (There may have been aesthetic problems with crowded wires, but not natural monopoly problems.)

Precision has never been the object of our minute federal regulations. It has been deployed chiefly to build the appearance of credibility. Few know that Competitive Enterprise Institute president Fred L. Smith Jr. was once a starry-eyed “planner” at EPA. Relating to me a joke symbolizing his transformation, he described the “young precise analyst” who proudly finished calculating results to four decimal places, as the older analyst looked over his shoulder and pointed to the estimate and proclaimed: “Son, I don’t know about the last decimal place, but your first one is wrong!”

Toward Solutions: Most Things Are Not Public Policy Questions

Wallach calls for iterative solutions. I like that, and I guarantee there are places we can meet. Valid, important approaches suggested in his broader work such as localism, justification, more agency scrutiny, and building congressional capacity relative to the executive branch are vital. But emphasis belongs too on what the Congress ought not be doing, not just what it should “do.”

Wallach hears and is sensitive to those frustrated with institutions and the disconnect between government and the citizen,  as for example when populists want to throw off the yoke but can worsen matters by creating a vacuum filled by still worse experts. Yet some “populists” on the right I’d argue do not wish to dismantle institutions. Rather they admire the institutions of limited government, and they voted Tea Party precisely to have them reestablished, a promise they have yet to see fulfilled. Wallach is right that imperfect politics can easily best imperfect populism.

Addressing a perceived libertarian “escapist” tendency, Wallach asks “how to get from here to there?” I have an answer: We start by not expanding government at each new frontier as is current practice. Again, government controls the airways but it doesn’t have to regulate drones under 1950s air traffic control; government owns roads but doesn’t have to regulate driverless cars or communications among them. Both are aggressively underway. Wallach also notes, skeptically, the libertarian wish for “anti-technocratic technocrats” to take charge and start streamlining. Yet the OMB’s Office of Information and Regulatory Affairs under Reagan was that, kinda sorta, and I bet Wallach and I could agree on a strengthened role for it in regulatory review and rule sunsetting.

Wallach respects the ideas of Professor Philip Hamburger, who accepts politics and a governing apparatus as a value rather than yielding to the populist vacuum. Hamburger would replace the administrative state with, well, Congress. When Hamburger calls today’s system a “return to our ‘preconstitutional past’” it’s a way of saying lawlessness. The Constitution was a deliberate repudiation of that, more than made clear by Hamburger when he said that the Constitution actually bans delegated power.

Wallach continues, “Indeed, the force of [Hamburger’s] argument relies on people understanding why democratic legislatures are superior bodies for decisionmaking relative to expert-led agencies.” But then: “To [Hamburger’s] credit, he acknowledges that this is sufficiently revolutionary as to feel somewhat quixotic”

But why? Why are classical liberal values quixotic? Enshrining constitutional values beyond just those that remain acceptable to modern liberalism shouldn’t be impractical. That voting to take property of a fellow citizen is today’s non-quixotic baseline speaks volumes. Even our hosts at the Cato Institute (my former haunt, by the way) called voter ignorance a threat to American democracy.[16] Of course, a remedy for excess and for voter ignorance is to make sure that government can’t do all that much damage, and that’s what the Constitution did.

Again, getting from here to there is a problem. But assume intervention is warranted; Wallach rightly proposes to “move our government’s center of gravity out of the administrative state” by challenging academic pieties and the arrogance of the ministers. He warns against partisanship in chiding us not to “blame Obama” and presumably thereby derail compromise. Fair enough, but one need only observe Obama’s own word about going around Congress, something alarming not merely to Obama’s detractors.[17] Congress should directly legislate wherever possible rather than tolerate open-ended agency regulation, or today’s dishonorable “informal” guidance. Confronting the obsolescence of decades-old statutes is a necessary, fundamental task, as David Schoenbrod has argued with respect to environmental regulation.[18]

Wallach wants evidence that Congress can “actually do the work of an intermediary.” If that’s the case, he might support the Regulations from the Executive In Need of Scrutiny (REINS Act).[19] Nothing in play now would go so far to make Congress directly accountable for what agencies do, in this instance by requiring a vote on “major” regulations. REINS would require owning up to major and controversial government mandates, and more importantly would change the calculus of future lawmaking. President Obama promises a veto, but I submit that you’ll stop hearing about REINS if Republicans keep congressional majorities and gain the presidency; entrenched regulatory rents may be too powerful for them.

Wallach observes that “If Congress actively demonstrates itself to be a body devoted to zero-sum posturing…a technocratic state [will] look relatively appealing.”  Yet anger is not directed only at the administrative state, but also at a Congress that cannot or will not stop it. The populist right’s perception is that Congress enabled the administration’s pursuits (spending, Obamacare), which is a big part of the “legitimacy crisis.” Such people aren’t tempted by agency technocrats; they want to see Congress get things undone, not get things done.

Wallach urges libertarians like me to join the call for “disciplining and humbling” the administrative state rather than “rerunning” some libertarian, non-technocratic past. I’m partly there, but the administrative state doesn’t even follow its own rules, and it’s getting worse with the rise of dark matter.

Still, I am certain we in this symposium shall find common ground on middle way, so to speak. Some of the past’s best proposals have required bipartisanship. Sen. Ted Kennedy helped drive transportation deregulation.[20] It was Lloyd Bentsen who proposed a regulatory budget,[21] an idea now in play again in the House Budget Resolution. It was Sen. Mark Warner who proposed a modest “one-in, one-out” for regulations, something being tried in Canada, the UK and the Netherlands. Whatever today’s disconnects, cooperation will occur again with urgency if the economy recovers more robustly and interest rates rise.

As we push Congress to take on its proper role, we might also attempt to enlist the executive. It certainly must be the case that if executive power is acceptable to progressives in 2016, it shall be so again in 2017. Reagan’s executive order 12291 brought rule counts and Federal Register page counts down by a third. We had a pen and phone that curtailed liberty, we can have “liberty’s meataxe” to expand it (within bounds). In 2012, for example, Mitt Romney urged passage of REINS, but vowed that if Congress didn’t he’d still forbid agencies to put rules in effect without Congress’s approval.[22]

Real “technocratic expertise” entails not curtailing liberty and substituting political choices for private ones, but discovering and extending institutions of liberty and voluntarism, and patrolling for ways to lessen the scope of the state in private concerns. Just for example, allocation systems for private property rights remain in their infancy (airsheds, watersheds, drone corridors, roadways, and intangible property) and require extension, not the obliteration they get at the hands of administrative bureaucracy. Such is the challenge of expertise in complex modern society: not to pretend to run things efficiently, but to help make sure healthy institutions enable the rest of us to do so.

I admire the spirit of Philip Wallach’s essay. Excepting his rather defensive one-liners about the professors of administrative law, he tries to put a range of views about the administrative state in their best light, mining truth from wherever it lies and proposing a middle way of incrementalist legislation to discipline the bureaucracy. It therefore feels almost churlish to argue, as I will, that there is no need for even a moderate solution because there is no demonstrated problem to begin with.

Although the spirit of the essay is admirable, the substance is weak. Three concepts are indispensable to any discussion of a putative “legitimacy crisis” in the administrative state: delegation, the presidency, and welfare, in the sense of well-being. None of these essential concepts appear in Wallach’s essay (and although there is some discussion of some of them in the longer companion paper, that discussion does not squarely address the points I will make here). When we do consider them, talk of crisis doesn’t look evidence-based; it is hardly obvious that there is any widespread illegitimacy afflicting the administrative state to begin with, whether we understand legitimacy in normative or positive terms or both. Wallach observes that “[w]e live in a democracy full of fellow-citizens who think that ‘crucial functions’ [of government] are very extensive indeed,” an observation that strikes me as entirely correct, but that also sits very uneasily with his initial premise that a legitimacy crisis even exists. Given Wallach’s observation, it seems that the administrative state is pretty much what our republic wants.

Let me take up the three concepts in turn.

Delegation.

From a lawyer’s perspective, many critics of the administrative state are oddly hazy about its constitutional origins. It is somehow just there, looming ominously over the constitutional order. From the more feverish of the critics one gets the impression that the federal bureaucracy was imposed by a Stuart monarch, or at best by Franklin Roosevelt, acting extra-constitutionally. The haze contributes to the suggestion of illegitimacy, as though the administrative state were an alien construct.

            In fact the administrative state is almost entirely the creature of Congress, acting through the constitutionally prescribed processes of lawmaking. This is conventionally called “delegation” to agencies, but it is delegation of statutory authority, not delegation of Congress’s own legislative powers, as I will explain shortly. Federal agencies, with a few exceptions related to the President’s core constitutional powers, are entirely creatures of statute; they are brought into being by legislation, given their powers by legislation, and constrained by legislation – including constraints that appear not only in the agency’s organic statute, but in the Administrative Procedure Act. (Not to mention constraints stemming from relevant constitutional provisions, such as the Due Process Clauses of the 5th and 14th Amendments.)

            One sometimes encounters the suggestion that Congress has unconstitutionally “abdicated” to the administrative state; and if the author is especially overheated, references to the Reichstag’s Enabling Act of 1933 will start flying about. Nothing could be farther from the truth. As Jerry Mashaw has shown, the creation of the administrative state has been a long-term, consistent, bipartisan project of the Congress as an institution, initiated more or less co-terminously with the birth of our Republic (the First Congress delegated wide powers to the President). That project has included an enormous amount of shaping and constraining of agencies, abolition of agencies that have outlived their usefulness, and oversight of agencies to check their excesses. This is no abdication of Congress’s functions, but a deliberate, sustained, and nuanced exercise of those functions. If the administrative state were somehow abolished tomorrow, Congress would in all probability start laboring to re-create it, in a cycle of eternal recurrence.

            In a more formalist vein, a bad argument circulates to the effect that by creating and empowering agencies, Congress has delegated away its own legislative powers, in violation of the principle that delegata potestas non potest delegare: a delegate (Congress, given its powers by the sovereign people) cannot redelegate (to the President and the agencies). But it has never been the theory of American law that Congress gives away legislative power whenever it grants statutory authority to agencies, although if the grant is excessively broad or vague, and thus lacks an “intelligible principle” to guide administrative discretion, a genuine problem of unconstitutional delegation may arise. In the normal case, agencies acting within the boundaries of their statutory authority exercise executive power, not legislative power. As the rabid liberal and statist Antonin Scalia wrote on behalf of the Supreme Court, “agencies make rules and conduct adjudications and have done so since the beginning of the Republic. These activities take ‘legislative’ and ‘judicial’ forms, but they are exercises of — indeed, under our constitutional structure they must be exercises of — the ‘executive Power.’” Scalia’s opinion here is no outlier; with the arguable exception of Justice Thomas, no modern Justice has fundamentally contested the legitimacy of delegation, whatever their complaints as to particulars. Its basic validity commands assent from Ginsburg to Alito. And the Court has invalidated delegations only twice in its history, both in 1935.

            The larger significance of delegation is that arguments in praise of Congress and classical lawmaking are themselves arguments in praise of the administrative state. It is not as though the administrative state was created against Congress’s wishes (how? by whom?). Whatever legitimacy Congress possesses transfers to the agencies. If we think that Congress possesses unique deliberative capacities, or uniquely representative properties, we should also think that Congress will delegate when it makes good sense to do so, after well-conducted deliberation and in an appropriately representative way. It is the strangest thing in the world to see critics of the administrative state turn around and praise Congress’s special qualities as a lawmaker, and call for a reinvigoration of Congress’s constitutional role – apparently not realizing that delegation is itself just one more species of lawmaking, one more tool in Congress’s toolkit, so that any argument for Congress’s special capacities is an argument that it should be entrusted with the power to delegate. When Wallach works around to his main positive suggestion for disciplining the bureaucracy – “show that Congress can actually do the work of an intermediary rather than just telling people it is meant to play this role [which] means doing the work of incremental legislating” – he doesn’t seem to realize that delegation just is a common form of incremental and disciplinary legislation. Very few or no statutory delegations are blank checks; as the political scientists have discussed, Congress almost always delegates under substantive and procedural constraints, expanding or contracting agency authority incrementally.

The presidency.

Wallach says essentially nothing about the presidency and its role in the administrative state. Again this is, I think, a symptom of haziness and abstraction about what exactly the subject of discussion really is. The executive establishment of the United States is internally complex. It includes line agencies in the executive hierarchy, either within or without the Cabinet departments; independent agencies; the multifarious Executive Office of the President; and the vast shadowland of quasi-public bodies and government corporations. In this world, the Presidency plays an essential coordinating and supervisory role, sometimes as overseer, sometimes as decider. Even as to the independent agencies, the President’s views are influential, and his power to appoint agency heads is eventually decisive. Most broadly, as Max Weber argued long ago, an independently elected President is one of the great constitutional checks on the bureaucracy. By shaping and constraining the behavior of the administrative state, the President contributes to and helps to ensure its legitimacy, transferring to it his own legitimacy as the sole elected official with a colorable claim to represent the nation as a whole.

We have, then, an administrative state that has been created and limited by the sustained and bipartisan action of Congress and the President over time; that is supervised and checked by the President as it operates; and that has been blessed by an enduring bipartisan consensus on the Supreme Court. The classical Constitution of separated powers, cooperating in joint lawmaking across all three branches, itself gave rise to the administrative state. When critics of the administrative state call for a return to the classical Constitution, they do not seem to realize they are asking for the butterfly to return to its own chrysalis. If political legitimacy is not to be found in this long-sustained and judicially-approved joint action of Congress and the President, the premier democratically elected and democratically legitimate bodies in our constitutional system, then legitimacy resides nowhere in that system, and the real complaint of the critics is not that the administrative state is illegitimate, but that our whole constitutional order is intrinsically misguided.

Welfare.

Happily, however, things are not so bleak. What, after all, is the evidence that a legitimacy crisis afflicts the administrative state? Legitimacy is a protean concept, but unless we are willing to subscribe to a strictly normative conception of legitimacy – which would have the odd consequence that the administrative state might be stamped “illegitimate” even if almost everyone likes it – we will have to admit that deep, widespread and sustained popular approval of the administrative state contributes to legitimacy. And then it becomes clear that whether we look at a short time-scale or a long one, the broad mass of the citizenry seems quite pleased indeed to live in an administrative state. (It is thus no accident that Congress and the President are as well).

In the short run, the probable nominees of both major parties are strong advocates of the administrative state; indeed a decent description of the Trump phenomenon is that a large fraction of Republican voters have rebelled against the libertarian, doctrinaire hostility to the welfare state propounded by Republican elites. A nation that twice elected Barack Obama by clear margins is a nation comfortable with technocratic governance. In a longer-term perspective, both parties, including presidents of both parties, including Ronald Reagan, have been enthusiastic promoters and protectors of the administrative state’s major accomplishments, including clean air, clean water, Social Security, and product safety. Internationally, the evidence suggests, happiness is strongly correlated with a robust welfare state, one that constrains the role of markets and that alleviates the main sources of anxiety in the citizenry’s lives – health care, educating their children, safety in both physical and financial senses. Well-being, including citizens’ subjective sense of well-being, evidently inclines them in favor of robust social welfare. Far from facing a crisis of legitimacy, the administrative state goes from strength to strength.

In the companion paper to his essay here, Wallach cites survey evidence suggesting that the American citizenry is dissatisfied with (1) “government” generally, (2) Congress, and (3) democracy. Putting aside the inherent limitations of such surveys, which can capture gripes and venting that would never result in support for tangible action, none of these findings bear directly on the legitimacy of the administrative state. The first is too general and abstract; the second actually works against Wallach’s own argument, insofar as his proposal is for a return to Congressional lawmaking of a certain incremental sort; and the third is entirely consistent with a preference for technocracy, and thus is no evidence of opposition to the administrative state. What “legitimacy crisis” is there, anyway? Before we have a long conversation about solutions, let’s figure out whether there is a problem.

In his lead essay, Philip Wallach decries the transformation of the American political system from one of representative government into one in which governmental power and decisionmaking resides with unelected “experts” within the executive bureaucracy–the administrative state. He also paints a fairly bleak picture of our ability to reverse the direction of this transformation. I share his distaste for the administrative state, but perhaps not his level of pessimism. But then, since I am a libertarian, this may simply be due to my tendency to indulge in impractical escapism, my willingness, as he says, to uncaringly “flip [my] fellow citizens the bird,” and my willingness to “not play[] the game at all.”

Of these three accusations, I am willing to plead guilty only to the last. Against the first two, I would try to mount a defense. So, let us first consider the charge of impractical escapism.

I believe that this charge and Wallach’s pessimism derive from his frame of reference, which is limited to the political. If I were a supporter of representative government looking through a lens focused exclusively on the political realm, I, too, would be depressed. For I would see no political mechanism that could transform our current administrative state into a representative government.

Happily for my emotional well-being, I am unwilling to narrow my focus exclusively to the political realm. I see the world as divided between the realms of individual and collective choice. The realm of individual choice is the realm in which individuals are free to decide how to act for themselves and personally reap the rewards and suffer the detriments of those decisions. The realm of collective choice is the realm in which a single choice binds all members of society. What color shirt I wear today is a matter of individual choice. Whether the federal government of the United States should build a wall along the Mexican border is a matter of collective choice. The political realm is the realm of collective choice.

If the political realm is limited to the formation of general rules that regulate interpersonal relationships, perhaps it could function without much of an administrative state. Rules that simply tell people how they are permitted to behave, such as laws prohibiting murder, theft, and fraud, or laws authorizing (or prohibiting) marriage between members of the same sex, could perhaps be implemented without a great deal of bureaucracy. But this is not true for collective decisions involving the regulation of the economy, the redistribution of wealth, or, generally, the expenditure of public funds to realize particular ends.

Elected representatives have neither the expertise nor enough hours in the day to oversee the implementation of such complex regulatory matters. The more areas of human life that move into the realm of collective choice, the more necessary non-elected economic, scientific, and other experts become to implement any such decisions.

Further, to the extent collective choices involve the expenditure of resources, they require first assembling the pool of resources and then designating people to make the necessary distributions. But as Ludwig von Mises pointed out as long ago as 1944,[1] in the absence of market pressure to use resources efficiently enough to generate a profit, there must be some means of controlling the behavior of those authorized to make the expenditures. As he explained,

The only alternative to profit-seeking business is bureaucratic management. It would be utterly impracticable to delegate to any individual or group of individuals the power to draw freely on public funds. It is necessary to curb the power of the managers of nationalized or municipalized systems by bureaucratic makeshifts if they are not to be made irresponsible spenders of public money and if their management is not to disorganize the whole budget.[2]

If one restricts one’s focus to the political realm, pessimism over the possibility of representative government without the administrative state is not only justified, it is mandatory. This is because it is a mistake to see representative government and the administrative state as alternative mechanisms of political organization. They are two sides of the same coin, inextricably linked by the requirements of collective decisionmaking. Under these circumstances, it is the belief that representative government could function without the administrative state, rather than the exhortation to try to shift areas of collective choice back into the realm of individual choice, that constitutes impractical escapism.

I would also resist the charge that libertarians are uncaringly dismissive of the interests of their fellow citizens in advocating the reduction of the realm of collective choice. Wallach may well be correct that “[w]e live in a democracy full of fellow-citizens who think that ‘crucial functions’ are very extensive indeed.” But it is not disrespectful to my fellow citizens to point out that they are unlikely to realize the benefits of these functions if the functions are performed by political agencies. Politicians are excellent at passing legislation guaranteeing justice that is never realized and promising services that are never delivered. It is worth noting that Democrats are currently campaigning against the evils of income inequality after eight years of pursuing policies designed to prop up Wall Street at the expense of savers and those on fixed incomes.

Further, it is not inappropriately dismissive to oppose public demands for extensive government services. I am not “flipping the bird” at my daughter when I resist her ardent demands that I buy her a new toy or do her homework for her. Taking such an attitude is problematic only if one is functioning within a democratic political framework in which the desires of large numbers of people magically become invested with moral authority. But here is where I must plead guilty to the third of Wallach’s charges, that of not playing the game at all.

Like him, I am skeptical of the prospects of rolling back the administrative state through political action. There is no way to oppose the administrative state by working within the system, if by the system we mean the democratic political system. The idea that supporters of small government (what Wallach derisively calls “anti-technocratic technocrats”) will ever come to power as a result of the knowledgeable support of a majority of the electorate is as much a fantasy as is a large representative government without the administrative state. As evidence, one need only point to the current presidential nominating campaign.

Indeed, my attitude toward politics is nicely captured by the denouement of the 1983 movie WarGames, in which a high school hacker inadvertently causes the supercomputer controlling the U. S. nuclear arsenal to play a game of global thermonuclear war with him. In order to prevent the computer from destroying the world, the hacker has it play tic-tac-toe with itself billions of times. The computer then stops playing the global thermonuclear war game and declares, “An interesting game. The only way to win is not to play.” In my opinion, this is a perfect description of the game of politics.[3]

I remain optimistic because I do not believe that it is necessary for the majority of the public to understand the benefits of free markets for there to be positive change. The best way to counteract the administrative state is to outrun it. Technological developments that create new ways of doing things that outpace government regulation–think PayPal, Uber, Bitcoin; think how long it took the feds to get around to regulating the internet–may be all that is needed.

Returning to a freer, less bureaucratically regulated society would be nice. But if that really is impossible, I would be satisfied with one in which the realm of individual choice–the market–was expanding slightly more rapidly than the realm of collective choice–the administrative state.

From the bottom of my heart, I hope that Adrian Vermeule’s breezy confidence in the remoteness of a legitimacy crisis is right, and that my own significant trepidation is just so much needless chin-stroking, as he suggests in his response to my original essay. But Vermeule’s arguments fail to persuade, and I remain seriously worried.

Vermeule says that “three concepts are indispensable to any discussion of a putative ‘legitimacy crisis’ in the administrative state: delegation, the presidency, and welfare, in the sense of well-being.” I address them in turn.

Delegation

Vermeule argues that “the administrative state is almost entirely the creature of Congress,” which has delegated significant powers to the executive, all the while “shaping and constraining” those delegations and providing oversight to prevent abuses. Delegating is a species of normal lawmaking—maybe the predominant one—and as such enjoys all of the legitimacy of normal lawmaking. Vermeule thinks that when I and others criticize the administrative state and offer incremental Congressional lawmaking as a preferred alternative, we are being utterly incoherent.

Vermeule errs here in revealing ways. Although he is unquestionably correct in describing delegation as an ancient and venerable tradition in American policymaking, he shows no awareness of any sense of degree. From the fair assertion that statutory delegations aren’t “blank checks,” he somehow leaps to the conclusion that they are equally unproblematic in coupling executive empowerment with needed limitations and accountability mechanisms. Obviously they are not.

On one end of a spectrum are delegations that charge agencies with well-defined, clearly bounded policy objectives; on the other, those that create reservoirs of ill-defined powers that can be taken up to address problems quite unrelated to those targeted by the statute’s authors. At this point in the administrative state’s history, delegations on the problematic side of that spectrum have piled up and worryingly empower an increasingly insular administrative state to do as it pleases.

If Congress nevertheless continually adjusts the scope of delegations as it learns from experience, that should tend to produce arrangements that are both substantively effective and imbued with fresh democratic legitimacy. But this kind of reorienting happens with distressing infrequency. In Vermeule’s way of thinking, the fact of congressional inaction allows us to infer the basic acceptability of the status quo, but, like many others, I suspect profound legislative and political dysfunction is more often the cause in recent years. The persistence of the status quo for any given policy is thus no reassurance that it enjoys public support or legitimacy. Administrative state actions are surely backed up with chapter and verse citations of existing statutes, but this formalistic justification loses its ability to confer legitimacy if the body of statutory law is being manipulated rather than refreshed. The key point here is that the legitimacy crisis of the administrative state and the steady deterioration of Congress are two sides of the same coin, and they must be reversed together.

I’m afraid I cannot pass from this subject without briefly noting that Professor Vermeule is an exceedingly strange messenger for this argument about the legitimacy that Congress can confer. In The Executive Unbound (coauthored with Eric Posner), he denounces the conventional “Madisonian” constitutional understanding, which sees the legislature as basically controlling the executive. He claims that this model cannot be reconciled to the inevitable reality of a dominant administrative state largely uncontrollable by law. More recently, he mused that Congress’s irrelevance has advanced so far that we should seriously consider just sacking it. So I am a bit perplexed to find him arguing that the administrative state does and will continue to draw meaningful legitimacy from the sanction of an institution he claims is has passed its sell-by date.

The presidency

Vermeule then suggests that the president can discipline and lead the administrative state and thus cover it in his own legitimacy; here he channels Elena Kagan and makes a reasonable-sounding case. The president’s election by the whole country does seem like it should confer an unmatched sense of legitimacy that can be spread around much of the federal government’s work.

But does it, in practice? In the modern hyperpolarized age, the president seems to be a focal point for partisan anxieties as much as a source of legitimacy; the White House’s next occupant seems likely to begin office facing a level of animosity not seen since the election of 1860. Nor is it clear that even a reasonably well-liked president can take ownership of the administrative state’s activities enough to make the public feel that its actions are basically emanations of the public will. My concerns are, to a significant degree, rooted in worries of ungovernability and incoherence, and come from studying policy areas where presidents seem to by playing catch-up to ad hoc decisionmakers rather than providers of political direction.

Welfare and the Appropriate Standard of Judgment

Vermeule asserts that “whether we look at a short time-scale or a long one, the broad mass of the citizenry seems quite pleased indeed to live in an administrative state.” He says living in “a robust welfare state” causes people to be happier (though offers only some international correlations in support), and dismisses polling data that I cite showing widespread distrust of American government as not “bear[ing] directly on the legitimacy of the administrative state.”

Vermeule is not the first to react to my work in this way, and I will admit that it is hard to produce empirical data that squarely and definitively show the administrative state has a legitimacy crisis. But, especially in this political season, I find it hard to understand the unflappable complacency that allows him to think “the administrative state goes from strength to strength” in terms of legitimacy. I don’t doubt that Medicare and federal worker protections have broad support among both parties, but I have no idea why Vermeule thinks that means people must not be alienated from the workings of the federal government. Nor do I see why he should be so indifferent to polling data showing plummeting levels of trust in our government institutions; perhaps he thinks that if the people are not revolting, there must not be a crisis.

I may be living in a different world than Vermeule, but recent political developments convince me that a crisis is indeed upon us, if only in simmering form for now. When three of the four leading vote-getters in the presidential primaries call for dramatic departures from our current way of governing, we should be able to agree that something is afoot.

John Hasnas writes: “ ‘The only way to win is not to play.’ In my opinion, this is a perfect description of the game of politics.”

But is politics an optional game? Isn’t it rather the sort of game such that, if someone with values or beliefs or interests antithetical to yours decides to play, you have to play too, or lose? And maybe politics has no outside anyway; maybe there is no bench one can sit on while others are playing politics. Liberty, and the free market, and property rights in emerging technologies, are just another set of political arrangements that will be contested and defended through and by means of politics. “Not playing” is just one sort of stance within the game - and probably a poor one that leaves the initiative, and the decisions, to others.

John Hasnas says that libertarians shouldn’t delude themselves into thinking that “representative government could function without the administrative state” or that “supporters of small government…will ever come to power as a result of the knowledgeable support of a majority of the electorate.” Likewise, Clyde Wayne Crews argues that the only path to legitimacy is to remove most areas of life from public policy altogether. The only good choice is to stop playing the game of politics entirely, they say.

Well, you may be through with the game, but the game ain’t through with you.

Libertarians are famously ambivalent about democracy, and my current interlocutors put that conflictedness on display in their responses to my original essay. On the one hand, Crews bristles at my suggestion that administrators be “made accountable to those parts of the public capable of meaningfully judging their results,” because it sounds like I am itching to exclude some voices that ought to legitimately be heard. The story is that if only widespread common sense triumphed instead of elite-held dogmas about the need for a protector, we would be better off.

On the other hand, there is a sense that we need to get a massive departure from the status quo, in which the very nature of our state is profoundly transformed and narrowed, and that it isn’t terribly important who supports such a change. Hasnas says “it is not disrespectful to my fellow citizens to point out that they are unlikely to realize the benefits of these functions if the functions are performed by political agencies.” Educating others may well be harmless, but following this logic further requires assuming a conclusion that others (myself included) do not reach. To override that fact of disagreement—by saying that it doesn’t matter so much what citizens or political actors understand as long as we know the truth—is surely not a sign of respect. Neither is flippantly dismissing the idea that the process of democratic choice could make “the desires of large numbers of people magically become invested with moral authority.”

I really don’t know what kind of country we would have if our political leaders, alone among the nations of the world, had long ago decided to limit legal regulations to those evolved through the common law. In the early years of our republic, that possibility may have been on the table; the First Congress considered wording what is now the 10^th Amendment to say that any power not “expressly” granted to Congress is reserved to the states or the people. But they did not choose that path. We do not live in that country.

Nor is it clear to me that we really could, human nature being what it is. Although libertarians prefer to exclude anything involving state coercion from the category of “spontaneous order,” there seems to be a near-universal human tendency toward handling certain collective problems through government. Figuring out how that government can be a beneficial (or at least a benign) agent of social cohesion, and do so while retaining legitimacy, strikes me as the inevitable central question for all who dare to dabble in social theory, libertarians included.

That doesn’t at all mean getting starry-eyed about the virtues of democracy, or its metaphysical foundation in the consent of the governed. I share Schumpeter’s skepticism on that front.[1] Representation is imperfect; the idea that it can merely transmit the public’s pre-existing will is usually incoherent; and there is no doubt that collective choices will often be oppressive. But recognizing representative democracy’s cons should not blind us to its pros. Often it can do a tolerably good job of learning about social problems and offering meliorative solutions; the need to court electoral support ought to drive socially beneficial choices, just as the profit motive drives capitalists to produce consumer surplus; and there is a long history of representative government achieving legitimacy. That’s not a bad package deal—even if, as I very much agree with Hasnas, the idea that it can somehow do without an administrative state of some kind is probably fantastic.

But what kind of administrative state? Well, that brings me back to my original arguments. It seems to me that Crews and Hasnas, as much as Vermeule, have arguments that are too powerful to allow them to draw any distinctions. Vermeule says “nothing to see here,” and they say “let’s call the whole thing off.” My argument is precisely a plea to avoid this kind of dichotomous thinking, such that we can think about how our current form of government might pursue a renewal of its legitimacy, and respond in an incremental fashion. Not all instantiations of the administrative state are equally offensive to liberty, the rule of law, or the Constitution; not all attempts to create accountability, both real and perceived, are equally fruitless. Libertarians do themselves a real disservice if they refuse to shape the outcomes of the reform game, because the larger political game is going to go on either way.

And there it is likely, as Crews says, that there is some common ground to find. I have cautiously supported some version of REINS before, but only if it is accompanied by an attempt to make Congress equal to the task of judging regulations. To the extent that people want Congress to be capable of getting things “undone,” as he says, they need to build capacity that they currently lack; unfortunately, things won’t get undone on their own.

And I agree that we should seek to allow new social developments to “outrun” the administrative state when possible. A silver lining to our cloudy legitimacy situation is that there are some real opportunities for this. In the case of self-driving cars, for example, the Googles of the world are often able to proceed with no more government oversight than an informal meeting with state and local leaders. Because our tech giants often possess greater legitimacy than the state at the present juncture, they find themselves with some socially beneficial room to maneuver. Undoubtedly there will be problems that lead to much more complicated politics down the road (for autonomous vehicles and other analogous areas of disruption), quite possibly because of the companies’ own desires to regulate their potential competitors into oblivion, but I agree with both Crews and Hasnas that for now we should enjoy the ride.

In preceding replies, Adrian Vermeule and Philip Wallach questioned libertarians over their relationship to democracy and their presumed withdrawal from politics. Before another post on the administrative state’s legitimacy or lack thereof (because we still seem not to be getting to the essence of what government may not do, whether administrators or Congress do it), I wanted to address that just a bit.

I very much appreciated when Wallach discussed “[f]iguring out how government can be a beneficial (or at least a benign) agent of social cohesion, and do so while retaining legitimacy.” I agree with him that it’s “the inevitable central question.”

He and I come at this from different perspectives, but I was saying the same in my first essay when I argued that, in a complex society of free individuals, the true task of the “experts” managing a limited administrative state apparatus is to make what was public business into private, not to try to expertly run things.

Yet while I do believe there are areas to remove from public life altogether, I do not believe “[T]he only good choice is to stop playing the game of politics entirely.” We can’t. And I don’t think either libertarians or Libertarians, little “l” or big “L,” are withdrawn; quite the opposite. I can see why some might think that, since I myself have cycled in and out of the “abstain” or “none of the above” mindset in frustration. But that’s transitory, not defining for the activist bent in the movement as a whole.

There is of course most obviously the political party with former New Mexico governor Gary Johnson as its presidential nominee in the past cycle; and there are local party operatives. Almost upon my own introduction to market liberal ideas in college, I became a candidate for state senate in South Carolina (and like all Libertarians, got a pitiful share of the vote!).

In any event, small-“l” libertarians often are what they are because they want to see change.  Often they are activists. Even if they are not in electoral politics, policy groups are involved in coalitions, regulatory filings, testimony, and so forth. Even when I say that I’d like to see areas of life removed from public policy, I think that has to be done via politics in the sense of legislators willing to roll back the state and the bureaucracy they spawned (we also seem to be having a non-argument over culpability for the administrative state; yes: Congress created it). Just as, alas, to remove an administrative state rule, one must go through the notice and comment cycle that adding a rule (allegedly, but does not) go through, Congress must act to roll back over-reach, and that still requires elections. Or convulsions like an Article V convention or even Charles Murray’s civil disobedience.

Practically, we don’t elect libertarians. But what does happen through activism, policy marketing, and outreach is that (some) policies libertarians support get adopted by one of the major parties. Libertarians favored school choice, something easily grabbed by Republicans, and long ago we championed Social Security reforms and privatization. But now it’s in the Republican platform. There is much overlap in activist work on left and right on issues like privacy and encryption and civil liberties.

I earlier jibed Wallach about being a tad elitist, and he cleverly turned that back on libertarians, stating that libertarians impart the “sense that we need to get a massive departure from the status quo… and that it isn’t terribly important who supports such a change.”

I do think it is “terribly important who supports such a change.” Those supposedly silent voices are making noises, and partly at issue is whether they have been getting heard. There were elements of grassroots libertarianism in the Tea Party (and that name came from somewhere, after all), and in the past election cycles that handed power to congressional Republicans, and in the current presidential cycle in which mainstream, establishment (pick your term) Republicans have been rejected by that public (the delegate count seemed to be a different tale there for a while.) So in my kind of libertarianism, I most certainly do appreciate those voices. They’re the folks I come from, too.

(On the other hand, the presidential campaign has demonstrated that a very large swath of young people are interested in and seduced by socialism, which cuts against my optimism, and given the decades of work by my beloved Cato, it alarms me!)

Libertarians, just like other political and/or party persuasions, aren’t withdrawing from democracy. This matters a lot given the topic of this forum. We libertarians surely must not be charged with reluctance to play the democracy game by advocates of a massive administrative state, populated by the unelected and the unaccountable, as untethered as anything can be from democracy.

Wallach rightly says that “Not all instantiations of the administrative state are equally offensive to liberty, the rule of law, or the Constitution.” But some are, and part of what we need to do yet is to better distinguish between them, not just insist upon or deny a blanket legitimacy. It is again statements like that one that make me see much overlap between Wallach and libertarians. We’re getting there.

Professor Vermeule raises a very interesting question when he asks whether politics is an optional game. I will do my best to respond briefly to his points without sidetracking the conversation too far from its focus on the administrative state.

First, I should clarify terms. Being subject to the political power of the state is not optional. The state will apply its political power to me regardless of what I do. I cannot opt out of that. However, whether I participate in the political process–specifically, the democratic process that decides who will wield the power of the state–is optional. This means that in this respect, I must decide what I should do.

Now, if the reader will forgive my use of philosophical terminology, we can examine this question from a either a consequentialist or a deontological perspective. Let’s take the consequentialist perspective first.

With regard to the political participation game, Professor Vermeule points out that “if someone with values or beliefs or interests antithetical to yours decides to play, you have to play too, or lose.” This is true. But it is a significant observation only under the assumption that if one played there would be a non-negligible possibility of winning. In my opinion, one who believes that peaceful individuals should be free to lead their lives as they see fit has no possibility of winning when the game being played is the struggle for political power. Space is too limited in this format for me to expound on this point at length, so let me just refer the reader to Chapter 10 of The Road to Serfdom entitled Why the Worst Get on Top, or perhaps merely gesture toward the current presidential campaign. In such a case, I believe it is better to work outside the system rather than toil in futility within it.

Now for the deontological perspective. Sometimes one is called upon to do the right thing regardless of consequences. If I am conscripted into a one hundred person firing squad that is going to execute an innocent person, my pulling the trigger or not can have no effect on the outcome. That does not mean that I should pull the trigger, even if everyone else does. If I am told that there is an organization that will extract and spend 30% of the wealth of the United States no matter what I do, and if I am offered a choice as to whether that wealth should be spent building a wall along the Mexican border or throwing sand in the gears of economic growth, the right thing to do may simply be not to participate in the system.

In our business ethics courses at Georgetown, we sometimes explore situations in the global marketplace in which there is no way to compete successfully without engaging in unethical or corrupt practices. Our highly motivated and competitive students often find it difficult to accept that in such situations the right thing to do is to go out of business. I find myself in a similar situation when contemplating participating in the political system.

My apologies for taking the conversation so far from the question of reining in the administrative state.

I am always happy to participate in a Cato Unbound discussion, but in this case I think I owe the readers an apology for being a poor interlocutor. If I understand the lead essay correctly, it assumes that democratic government is justified, but complains that the modern administrative state is so far removed from accountability to the people that it does not partake of the legitimacy conferred upon representative government. I am a poor interlocutor because I do not share the underlying assumption.

Mr. Wallach writes that “[l]ibertarians are famously ambivalent about democracy.” I cannot speak for other libertarians, but speaking for myself, I am not ambivalent about democracy at all. I am opposed to it. I do not believe that having more people on one’s side confers any moral legitimacy on coercive action.

There is nothing inconceivable about a government endowed with only a small number of particularly described or “enumerated” powers in which the officials charged with exercising these powers are selected by majority vote; that is, with there being a small island of collective choice embedded in a large sea of individual choice. This is entirely distinct from a democracy in which government is empowered to enact wide-ranging “public policies” that are chosen by representatives elected by majority vote; that is, where there is a sea of collective choice upon which one is required to sail in a ship captained by whoever is currently most popular. What would be inconceivable would be to believe that by participating in the political process under the latter form of government–that by engaging in the struggle for the captaincy of the ship–one could transmute the government into the former–that one could dry up the sea and trade the ship in for a rowboat. This was the source of my original suggestion that it would be better to look for protection against the deprivations of the administrative state outside of the political process.

I understand that none of this is helpful. I originally characterized myself as less pessimistic than Mr. Wallach because I chose not to see acting through the political process as the only means of counteracting the administrative state. However, if to advance the topic under discussion, I must adopt that frame of reference, then I am considerably more pessimistic than he is. I certainly agree with his assessment that “[r]epresentation is imperfect; the idea that it can merely transmit the public’s pre-existing will is usually incoherent; and there is no doubt that collective choices will often be oppressive.” However, I am unable to share his hope that the benefits of representative democracy–at least on its current national scale with virtually no restriction on the realm of activities that can be subjected to collective choice–can outweigh these drawbacks; that it “can do a tolerably good job of learning about social problems and offering meliorative solutions;” or that “the need to court electoral support [is likely] to drive socially beneficial choices, just as the profit motive drives capitalists to produce consumer surplus.” I’m afraid I have seen very little evidence of this over the course of my lifetime. And I have no idea what means by saying that “there is a long history of representative government achieving legitimacy.” If what is being referred to is moral legitimacy, then representative government achieves legitimacy in direct proportion to the extent to which large areas of individual activity are placed outside of its coercive control.

We should oppose the administrative state to the extent that its actions impose unconsented restrictions on individuals that prevent them from peacefully pursuing the goals that make their lives meaningful. We should oppose representative government to precisely the same extent. My inability to see how having more numbers on one’s side makes it legitimate to do what would be illegitimate if done by a single individual renders me unable to share the underlying assumption upon which the lead essay rests. This places me outside the bounds of the discussion, and renders me a poor participant in the larger conversation. If I make another contribution, I will try to do a better job of engaging with the issue.

Philip Wallach’s reply (“Our Simmering Crisis”) to my earlier defense of the administrative state (“What Legitimacy Crisis?”) helpfully clarifies some points of disagreement. I will focus on two. But I will ignore Wallach’s suggestion of an inconsistency on my part, insofar as I elsewhere defend expansive executive government. The suggestion misunderstands the logic of our conversation. While I don’t share his premises about the desirability of congressional government, I am willing to argue within his framework. I argue, in other words, that conditional on accepting Wallach’s premise that congressional government is in fact desirable, he should actually admire the administrative state, for the administrative state is Congress’s own beloved offspring, with whom Congress is well-pleased. There is no inconsistency between “Given your premises X, you should think Y” and “My premises are -X.”

Now to the merits. The first point of disagreement, I think, is that Wallach doesn’t seem clear in his own mind whether there is (1) a public consensus in favor of the legitimacy of the administrative state; (2) a (growing) public consensus against the legitimacy of the administrative state; or (3) a polarized public dissensus on the question. His first essay (“Farewell to the Administrative State”) hovered between (1) and (2), suggesting both that there is a growing legitimacy crisis, and yet also that “we live in a democracy full of fellow-citizens who think that ‘crucial functions’ [of government] are very extensive indeed.” His reply hovers between (2) and (3), suggesting both that there is a widespread albeit “simmering” sense of illegitimacy, and also that Congress is polarized on the question due to partisan disagreement, and thus is unable to legislate with greater force and clarity to check and monitor the executive. These possibilities are in severe tension with one another, because one thing Congress is tolerably good at is responding with alacrity to a widespread sense of anything. If Congress is paralyzed, it is often evidence, though not ironclad evidence, of deep public disagreement. So I wish Wallach would spell out his descriptive assumptions more clearly, although I suppose it is too late in the dialogue for that. My own view, of course, is (1). I think the broad mass of the public is quite accepting of the administrative state, largely because administrative government has raised public welfare dramatically. Indeed the most striking feature of politics 2016 is that the winning candidates in both parties are clearly comfortable with a robust administrative state. The opposition to it stems mostly from a small set of more-or-less libertarian policy intellectuals and lawyers, and an even smaller set of judges they have influenced.

The second point of disagreement is whether congressional delegation of authority to agencies, over time, has been “too broad” or “too discretionary” in some sense. I never understand what this means, especially in the mouth of those who otherwise favor congressional lawmaking as the putative alternative to administrative government. What’s the baseline for thinking that delegation is “too” broad? I suggest the following criterion: if the constitutional scheme of separated powers and its unique institutional attributes make Congress the preferred lawmaker (and again, this is a strictly conditional point for the sake of argument - I do not think those things), then the presumption ought to be that Congress delegates authority when and to the extent that it is wise to do so. There is no independent constitutional criterion of desirable breadth or clarity or what-have-you.

Beneath and behind these tensions there lies a central unresolved tension in Wallach’s view, one that runs through many critiques of the administrative state. It is a kind of love-hate relationship with Congress. On the one hand, Congress has failed systematically in its duties, leaving the administrative state to grow unchecked. On the other, the remedy is for … Congress to do better. But if the diagnosis is correct, why should we have any faith in the prescription? If Congress fails, not randomly but systematically, why should we think it will be capable of correcting its own pathologies, any more than a person could pull herself up by her own shoelaces?

“Who better than overreaching bureaucrats to decide when the bureaucrats are overreaching?”

That was one respondent’s characterization of the mindset that provoked the recently introduced Separation of Powers Restoration Act.

What can the administrative state not do? What are its bounds? Reducing the influence of the coercive state rather than reconciling to it still hasn’t fully registered in this Cato Unbound exchange over the regulatory state’s legitimacy.

Instead the central concern in this exchange has been what counts as legitimacy. To me this starting point means recognizing there are some things that we cannot vote on, and therefore there are powers over our economic and social lives that even Congress, the most accountable branch, cannot claim. We individually do not have such powers to delegate to representatives.

In turn, the delegated-to, unelected administrators would have to be even more restricted. That appears to be something with which John Hasnas would agree; our lead essayist Philip Wallach might partly agree, but Adrian Vermeule sees no problem with the administrative state at all. 

Congress assumed vast power and then delegated not just wiggle room but acreage to agencies. But Vermeule is fine with that: “If we think that Congress possesses unique deliberative capacities, or uniquely representative properties,” we should trust its delegation.

I don’t recall granting such concession; I’ve never blindly “praise[d] Congress’s special qualities” as an untethered lawmaker, and don’t recall anyone doing it at Cato when I as there either. For a Congress that doesn’t even read its own laws, delegation cannot be “just one more species of lawmaking, one more tool in Congress’s toolkit.” I’m against deference to agencies, but I’m against blind deference to Congress, too. I see the role of public policy and the statesman as continuing to expand deference to free citizens in recognition of their ability to solve problems. But that’s not happening, and it’s a sign of the state we’re in when an upcoming tell-all confessional by (apparently) a democratic representative will expound upon the prevailing “screw the next generation” attitude in our Congress.

I strongly agree with John Hasnas that democracy as construed today is not a virtue, that just because large numbers desire something does not render it “magically…invested with moral authority.” Democracy is picking personnel not masters.

Where Hasnas properly warns that numbers do not confer legitimacy, the administrative state doesn’t even bother with that. We don’t have large numbers calling the shots. We have relatively small numbers of the unelected, the self-interested. Yet Vermeule argues agencies don’t even legislate, calling what agencies do a delegation of authority, not an exercise of Congress’s own legislative powers. Writing about constraints that don’t apply, he says “agencies acting within the boundaries of their statutory authority exercise executive power, not legislative power.”

Vermeule continued, there are “very few…blank checks,” but rather “substantive and procedural constraints.” In my first entry in this exchange I noted the way the administrative state works, or rather, doesn’t work, in the sense of dispensing the Administrative Procedure Act that allegedly governs it, for rules, let alone the guidance, bulletins, memoranda, notices and the other material that I’ve taken to calling “regulatory dark matter.” Perhaps in an upcoming post I’ll note some examples of how “lawmaking,” for want of a new more descriptive term, happens now. Despite traditions going back to the ancients that inspired our framers, maybe nothing gets written down at all by today’s rulers.

Not only are there blank checks, some agencies run their own finances without congressional oversight, as little independent national governments. Even a change in presidential administrations may render little or no control over agency heads. As a nonlawyer hearing talk of Auer and Chevron deference and the enabling dominoes that tumbled before them, I regard the “intelligible principle” test for “legitimate” delegation as a diversion. Kind of a sandbox on a beach.

Vermeule further contends that “A nation that twice elected Barack Obama by clear margins is a nation comfortable with technocratic governance.” Apart from the numbers-rendering-legitimacy misconstruction, I may have more to say on such alleged “expertise” later as well. The administrative state does different things: part of it is wealth transfer, other parts are certain kinds of economic regulation, other parts range from health and safety protection to nannyism, and we have yet to distinguish between them in this forum. But the fact that officialdom neither does so nor recognizes the limits of competence achievable among them is another strike against legitimacy.

Making a law needs to be a risk to someone if done wrong. If you’re a legislator, perhaps you’re voted out. But today we have blank checks and the pretense of expertise instead of ironclad accountability for rent-seeking and cronyism. That allows the administrative state to say to those who wish for and want to protect traditions of limited government—in the country that largely created the phenomenon—to give it up. Indeed what anymore is regarded as not the business of the state, and championed by the bulk of academics? We are to believe that not only is it right for the administrative state to grow, but there is nothing to do about it. No thanks.

Vermeule did acknowledge a “vast shadowland of quasi-public bodies and government corporations.” Yet again we are assured there is no legitimacy crisis: “It seems the administrative state is pretty much what our Republic wants.” Alas, if we have an “administrative state,” then what we have is not a “Republic.”

The Constitution is one of humanity’s signature creations. Perhaps it could have been stronger and even more explicit than it is in restraining the state. But how sad to see the inadequacies of the branches to limit one another exploited so as to argue that the non-limitations were intended. What a disappointment to see the all-too-natural human failings of this precious, unique creation  get turned against the very country it created and used to justify blatantly unintended expansions of the state.

The question of legitimacy of the administrative state hinges on what government is, and can be, in the evolution of human liberty. America may well governed by the view that the Constitution smiles upon the administrative state. But even if the Constitution didn’t ban delegated power by design (I side with Philip Hamburger that it did), or if it failed to do so in practice (which is obvious), we remain able to envision, and enable, the ideal of limited government.

My previous posting seemed to distract our conversation from its focus on the administrative state due to my unwillingness play in the political ballpark. Let me see whether I can do better this time.

I will assume that we must live under the present governmental configuration in which the realm of collective choice is extremely wide-ranging, that much of the collective choice decisionmaking takes place in executive bureaucracies–i.e., the administrative state, and that the only action we may consider to change this situation is political action–action that can be taken within the current governmental structure.

Working within these confines, I would favor a result that shifted decisionmaking authority away from the executive bureaucracy and toward the members of Congress. The problem is that in my first posting I suggested that this was not feasible. There I pointed out that “[t]he more areas of human life that move into the realm of collective choice, the more necessary non-elected economic, scientific, and other experts become to implement any such decisions.” I now would soften that assertion a bit. What I should have said was that the more areas of human life that move into the realm of collective choice, the more necessary non-elected economic, scientific, and other experts become to implement any such decisions effectively.

For present purposes, let’s ignore the entire field of public choice economics. Let’s ignore the problems of regulatory capture, of the divergence between the agendas of elected officials and professional bureaucrats, and the lack of any price structure to rationalize the use of public resources. Let’s merely observe that there are only 535 federal legislators. Even with all their staffs, there is a limit to how much they can do. The human capital available to Congress could never approach that available to the executive bureaucracy. Further, because Senators and Representatives must address the entire range of issues that come before Congress, none of them can specialize, and there is a limit to how much they can allow their staffs to specialize. Finally, in a government with partisan divisions, the efforts of the members of Congress and their staffs will often pull in opposite directions, significantly reducing the effectiveness of any policymaking.

In other words, if policymaking must reside in the legislative branch, there is a limit to how much can get done. In the absence of any meaningful constitutional restraints, the physical limitation on the size of Congress and the inefficiencies inherent in democratic politics place an outer bound on the government’s regulatory power. This provides some limit to the realm of collective choice, which in my opinion is a good thing.

This may appear to be mere idle speculation, given my previous statement that I see no way to roll back the administrative state through electoral politics. But electoral politics is not the only form of political action. One can operate within the confines of the current governmental structure by advocating judicial action as well. And all that is really required to roll back the administrative state would be for the Supreme Court to reinvigorate the non-delegation doctrine last seen in  A.L.A. Schechter Poultry Corp. v. United States, 295 U.S. 495 (1935).

One of the protections that was supposed to be built into the structure of the Federal government was the separation of powers. The idea was that individual liberty would be better preserved if government power were divided into three separate and distinct functions–the legislative, executive, and judicial–that were administered by different branches of government. Thus Article 1, section 1 states, “All legislative powers herein granted shall be vested in the Congress of the United States, which shall consist of a Senate and House of Representatives.” Until 1935, the Supreme Court interpreted this to mean that Congress could not delegate its lawmaking authority to the executive branch of government. Unfortunately, the Court abandoned enforcement of the non-delegation doctrine, which enabled the massive shift in rulemaking authority into the executive bureaucracy that we are discussing today.

So one way to rein in the administrative state would be to convince the Court to start enforcing the non-delegation doctrine again. Now, one objection to this proposal might be that it is unrealistic. There is no way the contemporary Court is going to revive this long dormant legal doctrine. But such an objection would be unfair. By hypothesis, we are playing in the ballpark of political action. And, as far as I can tell, in this ballpark wishful thinking is standard operating procedure–e.g., raise the minimum wage and end poverty, enact Obamacare and bend the cost curve, invade Iraq and spread democracy, build a wall and make America great again. Calling a proposal for political change unrealistic could not distinguish it from any other proposed political action.

Further, it certainly has to be easier to change a few minds on a nine-member judicial panel than it would be to educate the public about the difference between legislative and executive rulemaking. The Obama administration has made it clear that it believes that it is empowered to act whenever Congress will not. If the current Presidential campaign is any indication, the public loves this idea. There certainly is no groundswell of sentiment for putting more power into the hands of the endlessly squabbling and endlessly grid locked legislature.

I had intended to write a bit more, but I just read Clyde Wayne Crews’s most recent posting. Because I do not think I can do better than that, I will subside.

Loving/Hating Congress

Adrian Vermeule expresses perplexity about my combination of diagnosis and prescription: “On the one hand, Congress has failed systematically… On the other, the remedy is for … Congress to do better. But if the diagnosis is correct, why should we have any faith in the prescription?” I don’t deny that, as a practical matter, getting Congress to do the right thing is likely to be difficult (see the closings of these two pieces).

But conceptually, what’s the confusion? I say the edifice of the state is starting to crumble, and that the only ones with the power to restore it are falling down on the job; Vermeule says that the interiors are kept in rather good condition and that the building is still standing, and so what’s the problem? For me, given our constitutional structure, Congress is the necessary legitimizer of our political order, administrative state very much included, difficulties or no. Vermeule sharply disagrees, arguing that Congress is necessary for precisely nothing; it’s a pity he insists that view isn’t relevant to our current discussion.

Then there is Vermeule’s strange logic when it comes to judging congressional output. If you think Congress is “the preferred lawmaker,” he says, how can you think its outputs, including the delegations underlying the administrative state, are so bad? This is little more than wordplay. Saying that the institution of Congress is given a “preferred,” central place in our system—and doubting that anything can effectively compensate for its failings—in no way commits me to believe that the actual occupants of the institution have done well. Indeed, it makes me more sensitive to their failings. Vermeule says “there is no independent constitutional criterion of desirable breadth or clarity or what-have-you” for congressional delegations, but I’m not seeking to advance any sort of doctrinal non-delegation point about what is constitutionally permitted. Instead, I’m saying that Congress’s combination of intentional responsibility avoidance and unintentional, dysfunction-induced neglect is leading to growing legitimacy problems for our constitutional system.

What sort of legitimacy?

That brings us to the question of what legitimacy means. Employing diametrically opposed understandings of legitimacy, John Hasnas and Adrian Vermeule are both frustrated with my lack of clarity on the subject. Elsewhere, I’ve tried to lay out my position: I agree with Weber (and Vermeule, I think) that we ought to think of legitimacy as an emergent social fact—positive, and not normative (or “moral”) legitimacy. When I say that representative governments have often achieved legitimacy, I do not mean to imply that they meet Hasnas’s (or Crews’s) exacting standards, but that they have employed processes and produced results that most people find broadly acceptable.

Given that posture, Vermeule usefully demands clarification as to whether I see the administrative state facing: (1) a supportive public consensus; (2) a critical public consensus; or (3) a polarized division of opinion. He thinks (1) is obviously correct, with the public correctly apprehending that the administrative state has increased public welfare, and the fact of congressional inaction testifying to the lack of a widespread public sense of outrage.

What I think is the following: support for some kind of administrative state with significant regulatory and welfare functions is very broad and deep. Hence my divergence from libertarians who hope for a redo. But that does not mean that people accept the current instantiation of the administrative state as legitimate. As I explain in my longer paper, people have a vague but growing sense that their government is insular and irresponsible. Of course regular folks don’t usually throw around the terms “administrative state” or “technocracy,” but they are nevertheless worried about the government’s hubris and dismissiveness of critics.

True, these concerns are most strongly felt on the libertarian right, and most often energetically rejected by those on the center-left—but my sense is that, across the political spectrum, concern about the future of representative government is at its highest point in decades. Again, I admit that it’s hard to substantiate that with polling data, but I’m still a bit bowled over by Vermeule’s willingness to look at the 2016 primary campaigns and conclude that the American people are giving a ringing endorsement of the status quo.

When I look at current developments, it makes me think that intellectual discussions like this one need to help Americans get from “something about the system is so wrong that we need to consider blowing it up” to “here’s what needs to change and how our representatives ought to be doing it.” No doubt, peddling a message of “Make Congress Great Again” will strike Vermeule as misguided and many libertarians as hopeless. I’m still game.

A brief thought for all the participants, but especially provoked by Clyde Wayne Crews’ latest contribution: let us not muddle everything up. More specifically, we ought to recognize three distinct questions and keep them firmly separate:

(1) The scope of government generally, as against “private” ordering and markets. Crews starts out as though he means to argue against the administrative state, but much of his quarrel turns out to be with the state, period. Many of his objections would hold no matter what the form of government, for they are objections to collective authority as such. (By the way, I put “private” ordering in scare quotes because I believe with Robert Hale that there is no such thing, that in fact “private” ordering is entirely parasitic on an old-fashioned form of governmental regulation, conducted through common-law courts prescribing rules of property, contract, and tort. Those courts are just an odd premodern type of administrative agency, with special tenure rules and procedures. However I need not argue this issue here; it is inessential to the distinctions I draw).

(2) The scope of national governmental power as such, vis-a-vis other actors. Even if the scope of Congress’ powers has expanded over the course of our history (and that is a controversial proposition), that does not yet have anything to do with the administrative state. Sometimes seeming objections to the administrative state are really objections to the content of regulatory policies, whether created by bureaucracies exercising delegated authority or instead created directly by substantive legislation itself. (I think that some of Crews’ objections are like this, but I can’t really tell). The scope of the national government’s total powers may or may not be “too” great, but the question about the administrative state involves the allocation of those powers within the government, between legislature and executive, not their scope as a whole.

(3) The scope of executive and administrative power. Finally we come to the administrative state proper. Here we encounter the paradox of arguments against delegation, my repeated theme in these comments. Why do critics of the administrative state (like John Hasnas in his latest) want more lawmaking to be accomplished directly by Congress, rather than indirectly by delegation to the administrative state? If it is because they believe that the Constitution makes Congress the preferred lawmaker, or that Congress has unique institutional properties (representativeness, deliberation, etc.), why does that rationale stop with direct substantive lawmaking, rather than including indirect delegated lawmaking as well? If they trust Congress to make laws, why don’t they trust it to make delegating laws? If they don’t trust it to determine when delegation is appropriate, why should they prefer that Congress decide the content of the laws for itself? Or is the putative objection to delegation really a disguised objection to the relevant law being made at all, by anyone? If so, let us have a conversation about the state, not the administrative state.

In my last post I suggested that I may point out a few prominent examples of agencies regulating with “guidance” alone rather than the notice-and-comment regulatory process. I will do so here.

Guidance—and notices, memoranda, circulars, letters, bulletins, and the rest of the regulatory dark matter “word cloud”—dispense with even the Administrative Procedure Act’s sub-constitutional requirement for formal advance public notice. The APA conveniently doesn’t apply to “interpretative rules, general statements of policy, or rules of agency organization, procedure, or practice” (P.L. 79-404. Section 553).

Now, I’m glad to partly agree with Adrian Vermeule’s characterization that my “seeming objections to the administrative state are really objections to the content of regulatory policies,” whatever their source. But it’s one thing to say Congress delegates legitimate administrative power, and what’s happening now is quite another. The difference was apparent for example in a Senate Hearing last year over agency guidance documents. Even under questioning, the process for how agencies decide whether some new initiative will be a formal rule or informal guidance was a “black box” to exasperated lawmakers. Democratic Senator Heidi Heitkamp questioned, “How some of this became guidance when it seems to us that some of it might have been better done in a substantive rulemaking, or better done in the bullying process by Congress.”

Such comes on top of my noting earlier how over a third of agency rules issue without a Notice of Proposed Rulemaking, thanks to the APA’s obliging “good cause” exemption. Further, most final rules have no Office of Management and Budget (OMB)-reviewed cost-benefit analysis (13 rules in the 2015 report, out of thousands or rules that actually issue any given year).

If legislation passed by Congress and signed in plain view by the president sometimes may strain the doctrines of limited government, then sub rosa decrees and guidances of bureaus most assuredly do. While President Barack Obama’s unilateral executive actions are today’s most prominent examples of action without Congress, federal agency guidance documents, memoranda and other proclamations probably predominate.

While their costs remain largely unmeasured or debated, we can count agency proposed and final rules, and even presidential executive orders and memoranda. But agency memoranda, guidance documents, notices, bulletins and the like are more intractable. Purportedly not legally binding, they do intimidate. It has been long recognized that guidance binds “as a practical matter” given that “failure to conform will bring adverse consequences, such as an enforcement action or denial of an application.”

Prominent recent executive agency guidance includes:

• Housing and Urban Development guidance asserting that landlord and home-seller denial of those with criminal records violates the Fair Housing Act;
• A Department of Labor blog post and “Administrative Interpretation” announcing that many independent contractors are henceforth employees;
• A Treasury Department delay (first by a July 2013 blog column, then by Internal Revenue Service guidance) delaying the Patient Protection and Affordable Care Act’s (ACA) employer mandate and in turn the accompanying tax penalty for noncompliance. This happened without public feedback or the mandatory economic analysis that is required (theoretically, but not actually) for “major” rules;
• The November 2013 declaration (in a presidential press conference, and then in Department of Health and Human Services guidance) that non-ACA compliant health policies could continue to be sold to the public;
• Environmental Protection Agency Clean Water Act interpretive guidance for “Waters of the United States.” Somewhat uniquely, this directive solicited notice and comment, but the agency apparently illegally manufactured endorsement by lobbying for supportive comments; 
• A Federal Aviation Administration “Notice of Policy” on drones that temporarily outlawed commercial activity. That action was reversed by the National Transportation Safety Board as an APA violation;
• Education Department guidance affecting colleges and schools with new mandates at the rate of one issuance per business day. One media-grabbing recent presidential/Department of Justice guidance concerns toilets in public schools.

Whether or not some will disagree that these count as abuses, these are a just a handful; I’ve personally tallied up a partial inventory of 580 pieces of “significant” guidance, yet there are thousands of sub-regulatory guidances that become more difficult to sort out even according to the agencies themselves. In addition, the aftermath of reforms like Dodd-Frank and net neutrality seemingly enable regulation without writing anything down at all. Increasingly, we’re living in a mother-may-I world.

In the recent paper “Why Congress Must End Regulation by Guidance Document,” I described the rise of federal agency regulatory dark matter and proposed some options for dealing with it. In this Cato Unbound exchange over legitimacy of the administrative state, questions still remain over what counts as expertise, why we think agencies have it, and whether we can “outrun” the administrative state as John Hasnas hopes.

As it happens, congressional task forces are engaging in some of this work as well. They are looking at overlapping issues of the expansion of executive branch power, restoration of congressional authority under the Constitution, and oversight and control of the federal regulatory enterprise. I hope that our lead essayist Philip Wallach continues to contribute to those efforts. It will be interesting to see the results.

Professor Vermeule asks three entirely reasonable questions, which bring helpful clarity to our conversation.

With regard to his first point, he is entirely correct to say that Clyde Wayne Crews’ points are arguments against excessive state power generally, not against the administrative state specifically. The same is true of my contributions. I think that this is a legitimate criticism of my comments for which I am willing to apologize. I was happy to be invited to participate in this discussion, but I am trained as a philosopher, not a political scientist. My perspective is the perspective of moral philosophy, and from that perspective, it does not matter much whether the legislature or the executive wields too much power. The problem is the power, not who wields it.

With regard to his second point, I have nothing to say. My objections in this context are not to the content of any regulatory policy, but to the power to regulate itself.

His third point addresses the relative scope of Congressional and executive rulemaking authority. In my last intervention, I intentionally restricted my focus to this question. I stated that I preferred that rule making power remain vested in the legislature.

I hope that I was clear that I do not take this position because I believed either “that the Constitution makes Congress the preferred lawmaker, or that Congress has unique institutional properties (representativeness, deliberation, etc.).” I take this position because it is physically impossible for the legislature to do as much as the executive agencies are able to do. Congress has more limited human capital and more limited financial resources than the executive. Further, it is internally divided and easily distracted, which renders it much less efficient. Given that the federal government long ago slipped its constitutional bonds, I favor legislative over executive law making in order to place some restriction on how much the federal government can interfere in our lives.

If the question that is addressed to me is whether Congressional or executive rule making is more efficient, then I am clearly on Professor Vermeule’s side of the argument. However, as a normative matter, the last thing I want is efficiency in government’s ability to restrain individual liberty.

Professor Vermeule closes his last posting by asking whether “the putative objection to delegation really [is] a disguised objection to the relevant law being made at all, by anyone?” My answer to this is, “yes, but I did not intend it to be disguised.” He is entirely correct to point out that in such a case, I am having “a conversation about the state, not the administrative state,” which may make me a poor interlocutor for this topic. But I can only write about what I know.

As a final note, I may also owe an apology to Philip Wallach for not understanding what he meant by the term “legitimacy.” Mr. Wallach clarified this in his last posting, stating that he was using the term to refer to “an emergent social fact—positive, and not normative (or “moral”) legitimacy,” and that “[w]hen I say that representative governments have often achieved legitimacy, I do not mean to imply that they meet Hasnas’s (or Crews’s) exacting standards, but that they have employed processes and produced results that most people find broadly acceptable.”

My obtuseness on this point may again be due to disciplinary differences. Perhaps this use of “legitimacy” is an accepted term of art that is well understood by political scientists. But even now, to my philosopher’s ear, the concept of non-normative legitimacy sounds like an oxymoron. It seems to imply that even extremely repressive regimes can be legitimate as long as people acclimate themselves to the repression and find the resulting state of affairs “broadly acceptable.”

I will close by observing that if Mr. Wallach is using the term “legitimacy” to refer to “results that most people find broadly acceptable,” then it seems to me that Professor Vermeule clearly wins the argument. For under this understanding of the term, the administrative state is clearly legitimate.

I had thought that my last intervention would be my last on this topic. However, because today was the first sunny day in years on which I did not have to either drop off or pick up children for school or activities, I decided to go to work by bike. On the ride home, I let my mind wander, and it wandered to Professor Vermeule’s main point in his argument against Mr. Wallach, to wit, “If they trust Congress to make laws, why don’t they trust it to make delegating laws?” From there it wandered to the way I, as a professor, am often tempted to make use make use of my research assistants, and I thought I saw an answer to the question.

As a professor, part of my job is to publish scholarly articles. Georgetown University trusts me to do this sufficiently to allow me the academic freedom to pursue the truth wherever it resides, no matter how unpopular that truth may be. I do my best to fulfill this obligation, and I flatter myself to think that my scholarly output is of high enough quality to justify what the university pays me.

I also hire research assistants to help me in this endeavor. They help me assemble the factual basis of my contentions and perform literature searches to make sure that I do not overlook relevant contributions by other scholars. As Georgetown law students, my research assistants are often very bright people with good analytical ability who write well. Despite this, they do not have the experience or insight into my thinking to write my articles for me.

I think I am a fairly responsible person. However, I am also an human being. Writing publishable articles is time-consuming and often frustrating work. I confess that there are times when I think to myself that it would be easier to simply let my research assistants write my articles for me. I never indulge in this dereliction of duty, but I do feel the temptation. I suspect that others do as well, and some with less resistance to the idea than me.

In my experience, professors are often intellectually brilliant men and women, but they are frequently also people with some unusual psychological proclivities. I can easily understand how universities could trust such people enough to allow them to write whatever they wanted, and yet not trust them enough to make unchecked decisions about how much work to delegate to their research assistants.

It seems to me that something analogous could be at play in the political realm. Assume for a moment that those elected to Congress are generally responsive to their constituents’ best interests–something that I regard as a counterfactual. Assume further that like professors, they are particularly good at something–in this case, crafting legislation that is acceptable to these constituents. But like professors, they are also human beings who are engaged in time-consuming and often extremely frustrating work. And as human beings, they must be sorely tempted to pass some of their workload off to others who do not have the experience or insight to craft legislation with this characteristic. Further, like professors, politicians frequently have some distinctly unusual psychological proclivities. If Hayek is correct in Chapter 10 of The Road to Serfdom, in order to get elected, they must be individuals who are not overly restrained by commitment to principle. And so, assuming the counterfactual to be correct, as Mr. Wallach appears to want to do, there would seem to be a perfectly good reason to trust the members of Congress with the power to write legislation themselves, but not with the power to delegate such power to others.

Is this an adequate answer to Professor Vermeule’s challenging question? I leave that to my fellow discussants to decide.

Legitimacy notwithstanding, we tend to discuss the administrative state as if it is a functioning expert entity, taking expertise in its divisions for granted.

But the question of whether the administrative state can actually be “expert” at anything has particular ramifications on John Hasnas optimism that we can “outrun” it with regime-changing technologies like Uber, or Airbnb, or Bitcoin, capable of displacing legacy industries or even govenrment oversight functions themselves. If expertise in delegated-power bureaus is something we must question, that makes it all the more important to outrun it, as well as to affirm Congress’s proper role in lawmaking, a step Philip Wallach seems more than willing to achieve by meeting us libertarians half-way to achieve.

I argue that the administrative state is not to be thanked unconditionally for “clean air, clean water” and “product safety” and the like because such values are also forms of wealth that require other disciplinary pressures to advance; that expertise must evolve, and that these are not just features that someone in a central government “regulates” into being.   

The stakes are higher at this modern stage in business history given the rise of networked technologies like the internet of things, drones, autonomous vehicles, cars potentially becoming parts of networks, 3D printing, augmented/virtual reality, privacy/cybercecurity concerns and such. Innovators and innovation alike are vulnerable to political predation that could derail the open-ended potential of technologies to better our lives and create not just a richer and freer world but a fairer and safer one than the administrative state can deliver.  

One form of political predation is the fact that so-called regulatory “experts” inevitably stop at market socialism in every modern case (for example, “retail wheeling” in electricity reform, net neutrality in telecommunications). Little interest is shown in extending institutions of private property rights in transportation grids, power grids, airspace, airsheds, watersheds, or critical infrastructure. Indeed, resources that were in government hands prior to the progressive era and the rise of the administrative state largely remain there. Expertise has bogged down; it didn’t expand.

Thus a primary threat to the modern tech sector is the hundreds of clinging regulators whose once-convincing justifications for existence no longer apply, even granting that they ever did.

The other ever-present threat is crony capitalism, ranging from the government funding of science and technology that widely displaces private funding and distorts the evolution of free competitive enterprise, to billionaires with their hands out for federal subsidies and favors.

Without being utopian about it, government failure has always been a graver threat than transitory market “failures” in my view. Government doesn’t merely pick winners and losers; it influences business models, and entire industry configurations, and entire economies. In the modern world, the decades-long damage regulators are capable of doing will be worse than in the world of our forebears, who outlawed competition in the communications and electricity sectors for a century, explicitly imposing government-granted monopolies for generations.

Today, “expertise” means drone and self-driving car policy appears poised to morph into 21st century versions of 19th-century public-utility style regulation. Since roads already are primarily government-owned, and since airspace is government-controlled, rest assured that the experts are not troubling themselves with liberalizing alternatives that reduce their power, such as the extension of private property rights into airspaces.

Locational tracking (if government doesn’t meanwhile mangle cybersecurity) could make it possible to (theoretically) pack the sky with commercial and personal drones in highly complex corridors. Some on the ground would allow flyover, some not, but the system would work anyway. There’s no need for one federal answer.

But what is the Federal Aviation Administration (FAA) doing instead? Requiring the registration of drones. That self-interested move undermines the work needed to address the issue of property rights in airspace/airsheds by ignoring it. Air traffic control-style, we’re set to get a handful of licensed, dominant operators with the “rights” and “licenses” to control the likes of the national drone package-delivery market as distinct from new, complex wealth-expanding contracts to ensure property function and protect property and lives. This would be familiar, just as 100 years ago, competitive electricity and communications services were eliminated in favor of monopoly franchises and a permanent regulatory superstructure to manage it.

This basic regulatory malpractice leaves aside the legitimacy questions, particularly the fact that Congress hasn’t passed law to allow regulation of drones. Such authority is taken for granted in enabling legislation here and elsewhere; so one important step for Congress today is to disallow regulation of new technologies unless it explicitly approves such. Similar presumptiveness that calls out for congressional intervention is the FCC campaign communications among driverless cars; private companies are already engaged. Maybe the cars should not communicate, but only detect one another; there are endless options either way. Maybe there won’t be driverless cars as such later on, but husk/chassis devices of various wheel counts that we snap a mobile device into to get where we’re going. In any event, regulation can inappropriately favor particular paths, closing off others.

I’m less sure that technology can outrun the administrative state as currently construed, but I wish it to be the case. Technology, luckily, is capable of exposing prior and ongoing regulatory malpractice if we are vigilant. But the planners are way ahead at this point; they exploit technology to increase centralization and side with rent seekers. They seduce allegedly deregulatory Republicans to go along, as they did with the capitulation to FCC’s net neutrality by proposing to ban so called “throttling” and “blocking.” We may yet see Uber become more like taxis than taxis becoming like Uber. And labor unions have their sights set on the sharing economy and the gig economy as ripe new territories.

The administrative state presupposes that some know everything, or at least a lot; I am sure we do not know everything. Alongside problems with over-delegation and legitimacy, the regulatory state can derail expertise rather than advance it. That undercuts its very justification, even if demands for Congressional oversight were met. Contractual, insurance, and liability innovations, which are forms of regulation, should become easier, not harder, alongside technology itself in the normal course of events. Those are the areas in which expertise must be extended through experiment. Expertise is characterized by discovering ways to expand the ambit of individual freedom and cooperation for the good of all.

I’m very grateful to Cato Unbound for hosting such a lively, thought-provoking conversation over the last month. My interlocutors have left me with a great deal to ponder. Here are a few parting thoughts.

What kinds of delegations?

Adrian Vermeule has doggedly persisted in arguing that if we trust Congress enough to make policy decisions itself, we must also trust it when it decides that delegating some part of policymaking is the wisest course. But Congress is an institution with changing membership, and the world that legislators confront at the time of lawmaking will not be the same as the one that exists 10, 20, or 50 years later. Statutes possess legitimacy when they reflect the best attempt of plural, conflicting legislators to direct the power of the (yes, generally administrative) state to the best effect given existing circumstances.

But there is no reason that the incompletely theorized agreements that result from this process should confer legitimacy on whatever actions the administrative state tries to tie to them for the rest of history. For purposes of legitimacy, old delegations are more problematic than recent ones; delegations that resulted from bitterly contested partisan victories are more problematic than those backed by nearly the whole range of diverse legislators; and delegations accompanied with clearly articulated purposes are less problematic than those conferring unpredictable powers on administrators.

One way to approach all of these issues is through the question: delegations to the administrative state, yea or nay? If this is the issue, probably I am closer to Vermeule than to Crews or Hasnas: I do not think across-the-board rejection is feasible or desirable. Where I part ways with Vermeule is when he suggests that if we choose the affirmative, any legitimacy problems our government may face today must have nothing to do with the particulars of its administrative character. The administrative state we have is often inadequately policed by Congress, insular and dismissive of external criticism, and overly convinced of its own ability to transcend politics. As I see it, addressing these flaws is crucial to maintaining our government’s legitimacy; complacency about them is a recipe for disaster.

What can Congress do?

John Hasnas gives one cheer for congressional government, if only because there are fewer people in the legislative branch. That means that if we really forced all policy decisions to go through them, government’s capacity to widen the “realm of collective choice” would hit a “physical” limit. Hasnas admits that this is not the world we live in, but nevertheless thinks that pining for a revitalization of the non-delegation doctrine and a robust separation of powers is no more unrealistic than other “wishful thinking [that] is standard procedure” in the realm of politics. 

But if congressional incapacity is a virtue in Hasnas’ alternate world, I hope that readers will nevertheless consider whether improving congressional capacity is the better way toward disciplining the administrative state in our world. Our institutional inheritance now includes a residuum of federal statutory powers that empowers the executive branch to take on nearly any issue, albeit sometimes through quite strained interpretations. In the absence of congressional direction, the administrative state is left to find its own way, with bad consequences for policy substance and legitimacy. (Also note that the historical underpinnings of non-delegation restorationism are shaky.) Of course, even if legislators are able to check the administrative state, getting them to make use of that capacity may be hard. But worries about willingness should not make us pleased with congressional disability.

Big-L Libertarians and the Administrative State

These issues take on new urgency as we turn toward the 2016 presidential election. Democrats and Republicans will offer two candidates more reviled than any others in American history, and neither of them will be offering a vision of a constrained administrative state. The usually quixotic Libertarian Party, on the other hand, has just nominated a ticket of two fairly moderate former governors who can offer a sober message of limited government. How they fare is likely to be of enormous consequence as the platforms of the two parties resettle into some new equilibrium.

The debate we have been having here is crucial to how Libertarians approach their opportunity. Will they view 2016 as their best chance to spread the gospel of minimal government and maximum human freedom, or will they offer their party as a practical champion of limited government in the here and now? Of course these are not mutually exclusive, but I would suggest they are very much in tension.

If Libertarians choose the evangelist’s path and present themselves as implacable enemies of the administrative state—as being committed to a vision of government that most of their fellow citizens would regard as anachronistic—they may win a few converts, but they will put a low ceiling on their support. On the other hand, if Libertarians focus on making the most of the institutional tools available to discipline the administrative state we have today, without any illusions that they can deliver minarchy tomorrow, they can force a broader electorate (and the other candidates) to take up important questions about how the administrative state goes about its business.

There are a great many such questions worth asking. Clyde Wayne Crews has raised some important ones about whether the administrative state actually observes the processes that it relies on to legitimate itself; I have offered some others, and of course regulatory reform ideas are legion. From my perspective, the most important thing is that people get beyond the distraction of the “yes/no” dichotomy and ask which aspects of the administrative state’s behavior worsen our government’s legitimacy problems, and what can be done to fix them.

I’ll wrap up our conversation about the administrative state not with grand theory, but in a more mundane register: with some legal points about so-called “guidances,” which seem to have become a flash-point for critics of the administrative state. My aim is to dispel some misconceptions which seem especially prevalent among nonlawyers. Hence I limit myself to settled legal propositions.

1. Guidances are typically either agency statements of policy, for example enforcement policy, or else agency interpretations of binding statutes and binding administrative rules. Such binding rules are often (although not always) promulgated using notice-and-comment procedures, which allow any interested person to comment, and which oblige agencies to respond to material comments.
2. In either case, guidances do not bind anyone by themselves. They implement or interpret underlying statutes or binding administrative rules that went through more elaborate procedures (or that, in the case of some rules, were validly exempted therefrom). It is not legally possible for an agency to use a guidance to create new legally binding obligations on regulated parties, except to the extent that underlying statutes and valid regulations already do so.
3. Guidances are expressly exempted from the legal requirement that administrative rules must be made through (at least) the notice-and-comment process. See Section 553(b)(3)(A) of the Administrative Procedure Act of 1946. That exemption represents a judgment by Congress about appropriate procedure, a judgment embodied in express statutory text; it was not created by executive order or agency fiat.
4. If a party believes that the agency has wrongly classified an action as a guidance, when it is really a binding rule or when the agency is treating it as binding, the party may challenge the agency’s action in court on procedural grounds in an appropriate case (assuming the law otherwise allows the party to proceed in court). Although guidances are sometimes said to have an in terrorem effect that frightens away would-be challengers, such cases are the routine stuff of administrative law litigation in the federal courts. The U.S. Court of Appeals for the District of Columbia Circuit decides them all the time.
5. If and when a guidance was properly exempted from notice-and-comment, it is because the guidance lacks “legal effect.” Because guidances lack legal effect, they do not improve or otherwise change the agency’s legal position. A regulated party is always free to argue that it does not fall within the terms of the underlying statute or binding regulation that the guidance is interpreting or implementing.
6. Even if a guidance was properly exempted from notice-and-comment procedure, it may be challenged on any number of grounds. The leading ones are (1) that it represents a substantively invalid interpretation of the underlying regulation, or (2) that the agency’s policy or interpretation was inadequately reasoned or justified (“arbitrary and capricious”). As to the former challenge, the agency will receive deference under Auer v. Robbins (1997) when the underlying law is ambiguous, but the court may always decide that the agency’s interpretation is clearly wrong.
7. The function of guidances is to guide – to give regulated parties and all the world notice of the agency’s views. In their absence, the agency would be free to take the same legal position, just without offering guidance to that effect. The absence of guidances does not mean less law; it means less guidance, less legal certainty, less clarity for regulated entities.