Carol M. Highsmith / Carol M. Highsmith Archive / Library of Congress

For decades, conservative and libertarian judges and scholars—at times joined by some liberal thinkers—have taken the lead in criticizing the ever-expanding authority of the modern administrative state. These critics attack the power of the massive bureaucracy to write and enforce its own rules and regulations. They protest the enormous discretion that this power confers both on officials within the executive branch of government and on members of independent agencies. Much of the criticism is directed to the now foundational Supreme Court decisions of Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., which in 1984 codified a principle of deference to administrative agencies on questions of law, and National Cable & Telecommunications Association v. Brand X Internet Services, which in 2005 further expanded that deference to cases in which an agency has reversed its own established practices.

The administrative state, of course, is not unconstitutional in all its manifestations. The large and sophisticated corpus of 19th-century administrative law offers us a benchmark by which we can evaluate post–New Deal developments. The success of that body of law depended heavily on the limited mission that it was asked to discharge, given its deep respect for both the doctrine of federal enumerated powers and a relatively robust conception of property and contract rights. But the New Deal expansion of the constitutional order has failed, as I argue in my new book, The Dubious Morality of the Modern Administrative State. To understand the extent and character of that failure, look only to what administrative law now allows: excessive government discretion to implement vast statutory schemes, many of which impose overbroad controls in such critical areas as environmental, labor, and food and drug laws.

Before the New Deal, the bulk of administrative-law cases dealt with matters such as land and patent grants, and military and civil-service employees. In both contexts, standard contractual principles of interpretation helped constrain the operation of administration law. The deference of 19th- and early-20th-century judges was not given to all administrative decisions, as the courts would later do in Chevron and Brand X, but only to those that followed consistent patterns of custom and usage. For example, Chief Justice William Howard Taft, in J.W. Hampton, Jr. & Co. v. United States in 1928, found an “intelligible principle” behind a statute that required that tariffs imposed on imported goods be designed to eliminate any unfair differential in price between domestic and foreign goods. The task there was manageable precisely because it was mathematical, so that the level of administrative discretion was necessarily constrained by the limited mission.

With the advent of the New Deal, the scope of the government mission became far broader, because for the first time it was possible to conceive of public ways to reorganize large segments of the economy, which by circumstance can be done only by administrative agencies. A quick look at a handful of relevant cases shows how the advent of new government powers diminished individual rights and transformed the legal and political landscape.

As an initial example, consider the allocation of the electromagnetic spectrum, the medium for radio communication, which had by that time become big business. In 1943, in National Broadcasting Co., Inc. v. United States, Justice Felix Frankfurter had to construe the deceptively simple words that defined the authority of the Federal Communications Commission under the Communications Act of 1934—“public interest, convenience and necessity.” Frankfurter read it broadly: “The Act does not restrict the Commission merely to supervision of the traffic. It puts upon the Commission the burden of determining the composition of that traffic.” The traffic officer sets the rules of the road, such that cars don’t crash and frequencies don’t interfere. But how does anyone determine the composition of the traffic? Frankfurter did not have a clue, so he blessed a broad delegation of that job to an equally clueless administrative agency.

With this maneuver, gone were the last vestiges of the non-delegation doctrine—the principle that Congress could not delegate its responsibility to write the country’s laws to administrative agencies—because no narrower grant of power could let the FCC look at some undefined range of supposedly relevant factors, as required by the statute. So when the FCC was forced to use “comparative broadcast hearings” to determine which of many applicants should receive a coveted government license, its ad hoc determinations made it difficult to articulate a sensible standard of judicial review. By 1965, the FCC had developed a decision process that rested on a list of seven malleable factors: diversification of control by media companies; full-time participation by station owners; proposed program service; past broadcast record; efficient use of frequency; character; and “other factors”—a meta-factor, as it were.

Until it was scrapped by the Telecommunications Act of 1996, this system of weak property rights cost the public in two ways. First, the wasteful and politicized hearings often assigned frequencies to parties for whom they had low value. The new licensees were then allowed to sell their licenses for cash, so that the system not only required a wasteful hearing, but also allowed the gains from the sale of the frequency to be captured by a private party, not government. Second, rights to use a frequency band were subject to all sorts of government restrictions that reduced its value, only for the lucky winner to assign its license to a third party. As a consequence, when technology improved, the licensee was prohibited by law from packing more channels into the allocated spectrum. It was as if a homeowner had to make sure that every room was exactly 20 feet by 20 feet.

And so the overconfident Frankfurter got everything wrong after all. The correct system to use here is a variant of the traffic-cop model: an auction of defined property rights, subject only to boundary conditions dealing with interference. But Frankfurter decided as he did because he understood that the statutory licensing system was necessitated by the New Deal’s rejection of market institutions. The elimination of property rights puts burdens on the administrative-law system that it simply cannot handle. Bad constitutional law begets bad administrative law.

NBC is a tricky case because government discretion is greater for resources it owns than for those it does not. But the same untrammeled discretion arises in cases in which the government interferes with private relationships. One of the worst statutes in that regard is the Fair Labor Standards Act of 1938, which sets minimum-wage and overtime regulations. There is no coherent reason this task should fall to a government in a competitive labor market. The key statutory move is to give a new definition of the term fair. A labor standard’s purpose ought to be to rule out force and fraud, whose use distorts wages in competitive markets. But the FLSA extends the notion of unfairness to the ever present forms of inequality of bargaining power. It then seeks to delegate to the Department of Labor the power to issue regulations to implement that flawed definition.

A 1944 case, Skidmore v. Swift & Co., involved a ruling by the DOL on overtime regulation, which held that firefighters on call after regular hours were entitled to overtime if they played games, but not if they slept, thus giving these workers a perverse incentive to stay awake. But the deeper problem was that the FLSA, in awarding time-and-a-half for overtime work, makes some sense only when the overtime hours are more arduous than regular work. Here, the prior employment contracts between the firefighters and the municipality got the compensation schedules right: low wages (and some in-kind benefits) while on call, and substantial payments when called out. The rigidity of the FLSA blocks that previous package. Since the correct answer was out of bounds, Justice Robert H. Jackson held that these principles “are entitled to respect,” no matter how senseless the resulting administrative rule. Once again, a bad substantive scheme led to an adoption of unwarranted discretion by an administrative agency.

A third pillar of administrative law is formed by the companion cases of Securities and Exchange Commission v. Chenery (I), in 1943, and Securities and Exchange Commission v. Chenery (II), in 1947, which involved a complex corporate reorganization carried out under the Public Utility Holding Company Act of 1935. The SEC sought to prevent some corporate insiders from sharing pro rata in the gains of a successful reorganization with other preferred shareholders, even though the agency conceded that the parties met their fiduciary duties of loyalty and fair dealing. In Chenery I, Justice Frankfurter held that it was permissible to impose liability on the insiders only if the new regulation were specifically authorized by either judicial decision or general rule. Chenery II upheld that imposition four years later. The New Deal mind-set treated the rules of fiduciary liability as infinitely malleable by routine judicial and legislative activities, not subject to any discernible principled substantive restrictions.

But that has to be wrong. Chenery II limited the insiders’ share of gains. Could the next case just wipe them out altogether on a principle to be announced later? No government actor should be able to strip shareholders of their interests by announcing that the retention of their fractional interest is now a legal wrong. The creation of fake liability rules is an open invitation for government to take its first steps toward a confiscatory scheme. The risks are equally great whether the new liability is imposed by common-law decision or administrative rule. The SEC’s proper role is combatting fraud and forcing full disclosure. Take it away from its moorings, and it can literally make just about any kind of conduct illegal and impose any sanction for that newfound illegality.

Each of these three statutory schemes—NBC, Skidmore, and the Chenery cases—involves conscious deviation from common-law schemes, without explaining why such novel liabilities should be imposed. At that point, the deadly cycle begins. The standardless rules lead to a massive delegation to administrative bodies, which are totally unconstrained in their statutory objectives. When the permissible ends become broad, the delegated authority cannot be effectively constrained on matters of means. The all-too-common pattern of judicial deference is far easier to maintain in a world that contains no fixed substantive commitment than a world in which these commitments bind. So if property rights become inherently fragile, if wage-and-hour regulations can be imposed at will, if ad hoc duties can be imposed on corporate actors, the world of voluntary markets becomes a more dangerous place.

The post–New Deal approach offers no constraints on ends. It does not treat force or fraud as the necessary predicate of government intervention in the market. In trade regulation, this approach is as comfortable setting up the milk cartels in 1934’s Nebbia v. New York as it is in striking them down, even though from a social point of view, free competition always outperforms cartelization.

The great flaw of New Deal legislation is that it unleashed a new system of regulation in which the ends were, and are, illegitimate. Strike down these ends, and there is no reason to worry about the administrative enforcement of invalid statutes. Let those statutes be sustained, and a vast and untethered administrative state will be erected to ensure their enforcement. And so it is that the lax constitutional protection of individual rights nourishes the uncontrollable administrative state America has today.

This story is part of the project “The Battle for the Constitution,” in partnership with the National Constitution Center.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.