In 1935, Justice Benjamin Cardozo complained of “delegation running riot.” Since then, it has only expanded in scope and scale as Congress has, time and again, delegated to administrative agencies the authority to draft, develop, and deploy regulatory regimes governing nearly all aspects of economic life in America. Might this be about to change?

For the first time since Cardozo’s complaint, it appears a majority of justices on the U.S. Supreme Court are prepared to reconsider the limits on congressional delegation of lawmaking authority to administrative agencies. Although the Court has not invalidated a federal statute on nondelegation grounds in over 80 years, it has opened the door to new nondelegation challenges.

In June 2019, in Gundy v. United States, the Court considered a nondelegation challenge to the Sex Offender Registration and Notification Act (SORNA), which authorized the U.S. attorney general to determine whether the act’s requirements applied to individuals who committed qualifying offenses prior to SORNA’s enactment. In the end, SORNA survived, as the Court majority was unwilling to declare that the grant of authority to the attorney general constituted an unlawful delegation of power, but Gundy revealed substantial discomfort with the extent to which the Court has allowed delegation to proliferate.

Chief Justice John Roberts and Justices Neil Gorsuch and Clarence Thomas dissented in Gundy and called for aggressive judicial enforcement of limits on Congress’s ability to delegate legislative power. Justice Samuel Alito voted to uphold Gundy largely because he did not want nondelegation concerns to apply disproportionately in criminal cases. He wrote a separate concurrence expressing his willingness to reconsider constitutional limits on delegation in an appropriate case.

Justice Brett Kavanaugh did not participate in Gundy, as he had not joined the Court until after the case was argued. But he has since affirmed that he, too, believes nondelegation concerns could “warrant further consideration in future cases.” If each justice is to be taken at his word, that would indicate a majority of the Court is open to taming legislative delegation.

Yet, reconsideration of the nondelegation doctrine may not mean much. If the Court’s practice with other revived constitutional doctrines is any guide, it may take more to curb delegation’s reach. In 1995, a Supreme Court majority announced its intent to police the limits of Congress’s Commerce Power, and yet few federal laws have seen their reach constrained. New and unprecedented assertions of federal power have been turned away, but the rest of the U.S. Code has been left intact. In much the same way, a nondelegation revival may target newly enacted outliers without doing much at all to curb those delegations that are already on the books. If delegation is to be curbed, it may have to be stopped at the source: Congress itself.

Delegation and Time

Delegation lies at the foundation of the modern administrative state. Federal administrative agencies have no inherent power to issue regulations, administer programs, or enforce federal law. Rather, Congress grants agencies those powers through legislation. In various statutes, Congress has granted agencies the authority to implement — and oftentimes direct — federal policy across a wide range of areas, and this practice of delegation has increased over time.

Congress may well have good reasons to delegate substantial policymaking and implementation to administrative agencies. Among other things, legislators may feel they lack the technical knowledge or expertise possessed by career staff at administrative agencies. It may also be easier to develop coherent policies on complex or controversial matters within a hierarchical structure than in a legislative committee. Agencies may also be able to act with greater speed and dispatch than a bicameral legislature, making them more suited to address urgent problems. At the same time, legislators may use delegation as a way of evading accountability for their actions.

However necessary the practice of delegation, it is not without its costs. Allowing extensive delegation may undermine accountability, entrench policy agendas, and compromise the democratic legitimacy of agency action. As the late legal scholar John Hart Ely observed, the concern with delegation is not necessarily that “ ‘faceless bureaucrats’ necessarily do a bad job as our effective legislators.” Rather, it is that “they are neither elected nor reelected, and are controlled only spasmodically by officials who are.” In this way, broad delegation can be viewed as a threat to deliberative democracy.

Much criticism of unbridled delegation focuses on the volume, range, and expansiveness of the legislature’s delegation of authority. Some statutes grant federal agencies the authority to make broad policy decisions with tremendous economic consequences, such as what to set as the acceptable level of air pollution in urban areas or how to regulate emerging telecommunications technologies. Other statutes give agencies minimal constraints on whether to adopt regulatory measures and what policy objectives such measures should pursue.

While most critics have focused on the breadth and scope of delegation, less attention has been paid to the time element. Agencies using their delegated power regularly draw on statutory authority granted many years (or decades) earlier. Agencies quite often rely on long-standing — and even long-dormant — authority when creating new regulations. Rules promulgated by federal agencies in 2020 may be authorized by statutes enacted 30, 40, or 50 years ago, drafted and voted on by legislators who may no longer be alive, let alone still serving their constituents.

This risks undermining the democratic legitimacy of much agency action. Consider that when the Federal Communications Commission first sought to adopt an “open internet” order, it relied on a 1934 statute that Congress had not substantially revisited in 14 years. Even with those revisions, the statute was out of date within a decade. The 1996 amendments to the Communications Act preceded Wi-Fi networks, let alone Facebook, Wikipedia, Netflix, and even Google. These amendments responded to time-specific technologies and market pressures, and presumed a desire for a “stovepipe” regulatory framework separating telecommunications and information services. However appropriate such ideas were in 1996, they are obsolete today.

The Communications Act is hardly an exception. Environmental law is likewise replete with statutes based on outdated or mistaken assumptions that limit their effectiveness. In some cases, these statutes relied on then-current scientific understandings of environmental problems and their causes. Yet, as scientific understanding and technical expertise concerning pollution and other environmental concerns have advanced, the statutory regimes have not kept pace. Much of the Clean Water Act focuses on pollution from point sources; relatively little of the act concerns nonpoint sources. However well-justified this emphasis may have been in 1972, it makes no sense today, as nonpoint source pollution now presents the far greater threat to water quality. Yet, the Environmental Protection Agency has been delegated relatively little authority to address that.

The Clean Air Act (CAA) is arguably the most expansive federal environmental law. It is also the source of authority for recent regulations adopted to limit greenhouse gas emissions in an effort to reduce the threat posed by global warming. Congress erected the CAA’s basic architecture in 1970 and made significant modifications in 1977 and 1990. As originally constructed, it focused most acutely on localized air pollution. The act’s core provisions define acceptable ambient concentrations of regulated air pollutants and direct states to adopt plans to ensure compliance with the designated National Ambient Air Quality Standards (NAAQS). Relatively little of the CAA’s core architecture concerned interstate air pollutants. Global climate change in particular was not yet a serious concern within Congress.

When Congress last modified the CAA in 1990, it tightened and revised the NAAQS provisions. It also expanded the statute’s scope to address non-localized air pollutants, such as those that contribute to acid rain and the depletion of stratospheric ozone. Separate provisions addressed each of those concerns. However, no provisions expressly addressed greenhouse gas emissions. Nor have any such measures been adopted since. Nonetheless, 17 years later, in Massachusetts v. EPA, the Supreme Court concluded that the plain language of the CAA was broad enough to cover greenhouse gases as air pollutants. Whether the Court was correct to interpret the CAA in this fashion, it is fair to say that Massachusetts v. EPA set in motion a series of regulatory initiatives that Congress never contemplated, let alone endorsed, and forced the EPA to retrofit a 20th century statutory regime to address a 21st century problem.

This temporal problem is not limited to regulatory programs. Older extant statutes often enable the executive branch to take actions Congress did not anticipate. For instance, Congress enacted the International Emergency Economic Powers Act in 1977 (IEEPA) to empower the president to take concrete actions in response to any “unusual and extraordinary threat … to the national security, foreign policy, or economy of the United States” arising from outside the country. This statute grants broad authority that has been invoked to address a wide range of foreign policy concerns. While Congress did not seek to delineate the precise circumstances under which the IEEPA could be used, it is quite unlikely the 1977 Congress intended to delegate authority to impose tariffs on Mexico in response to an alleged illegal migration crisis. Yet, that is how it was used in 2019.

While most critiques of excessive legislative delegation focus on the scope and range of discretion entrusted to administrative agencies, the temporal lag between legislative delegation and utilization of delegated authority raises distinct concerns about whether delegation is consistent with democratic governance. Agencies only have that power delegated to them by Congress. Yet, when decades pass between the enactment of statutes delegating authority to agencies and the exercise of that authority, there is a risk that the delegated authority will be used for purposes or concerns that the enacting Congress never considered. This may lead to situations where Congress has not provided the proper tool for the problem the agency is addressing. More broadly, agencies may be exercising power granted for one purpose to pursue another aim that Congress had never contemplated. This was arguably true with both the FCC’s initial effort to impose “net neutrality” and the EPA’s use of the CAA to address climate change.

When agencies rely on regulatory authority delegated to them in the past, there is also a risk that the power exercised is no longer in line with contemporary legislative majorities. The inertia inherent in the legislative process makes it difficult to revise delegations of authority and can entrench the dead hand of a past Congress. Consequently, agencies may often have the power (or even the obligation) to act based upon a prior Congress’s preferences that no longer command popular support. In this respect, the lag between delegation and regulation may create a particularly worrisome democratic deficit. The values ascendant at the time of enactment may no longer command widespread support. Particular policy concerns, much like given statutory language, may be obsolete.

Closing the Democratic Deficit of Time

While the nondelegation doctrine has not led to the invalidation of federal statutes, nondelegation concerns appear to have influenced various administrative law doctrines. Most notably, nondelegation concerns influence how the Court interprets statutes that may be understood to delegate authority to regulatory agencies. In particular, courts are not to lightly presume that Congress has delegated authority to agencies that might implicate constitutional concerns such as by intruding on state prerogatives or infringing upon constitutional rights. Yet, however much such doctrines may compensate for the potential democratic deficit caused by delegation, they do little to address the time concerns.

If delegation is to be curbed — and, in particular, if the democratic deficit of utilizing aged statutes to adopt new regulatory initiatives is to be constrained — it is possible courts are not the place for reform advocates to focus their efforts. If Congress is the source of delegations, perhaps Congress should also be the entity to exercise control.

The distinct temporal problem of broad delegation and related concerns over statutory obsolescence would be addressed if Congress were to return to the practice of enacting substantive legislation on a regular basis. This is easier said than done. Presumably, legislators would legislate if that was their preference. That is, if members of Congress believed the benefits of regularly legislating outweighed the costs, then that is how they would behave. For a variety of reasons, including competing demands on legislators’ time and alternative ways to invest their political capital, legislators choose not to legislate with any regularity.

The surest way to change legislative behavior is to change the incentives legislators face, and this is something self-conscious legislators may seek to do. In a wide range of contexts, Congress already enacts laws and adopts procedures with an eye to altering or ameliorating the incentives future legislators may face. If, as we argue, lawmakers do not revisit and reevaluate existing statutory frameworks as often as they should, Congress may be able to help solve this problem by seeing to it that inaction has consequences.

Temporary Legislation

One way Congress can encourage future legislators to revisit existing statutory frameworks on a more regular basis is through the use of “temporary” legislation. Legislation that “sunsets,” expires, or otherwise requires regular reauthorization could induce Congress to revisit, reassess, and recalibrate existing programs so as to ensure that such programs reflect current knowledge, focus on the most salient concerns, and are more in line with contemporary voter preferences.

This idea is not new. “Temporary legislation,” Harvard law professor Jacob Gersen has observed, “is a staple of legislatures, both old and modern.” Well before the birth of the modern regulatory agency, prominent voices extolled the virtue of legislation that needs to be renewed or revisited. Thomas Jefferson, for instance, argued that vices such as corruption make statutory expiration preferable to relying on the possibility of repeal. In Federalist No. 26, Alexander Hamilton argued two-year limits on military appropriations would require periodic deliberation and thereby check potentially unwise policy decisions. Temporary legislation was embraced by colonial legislatures and the early Congress. The Sedition Act of 1798, as enacted, expired in 1801, and the first two national banks were created with time-limited charters and allowed to expire as well.

During the New Deal, when Congress set about creating a range of new federal agencies, William Douglas — then on President Franklin D. Roosevelt’s Securities and Exchange Commission — urged consideration of limiting how long Congress’s new creations could operate without renewed legislative authorization. Prior to his appointment to the Supreme Court, Douglas advised Roosevelt to include sunset provisions because of the risk that a new agency would have exhausted its “great creative work” within a decade and risked falling prey to “inertia” and becoming “a prisoner of bureaucracy.” Sunset provisions, in Douglas’s view, limit rent-seeking within the administrative state. The late Cornell political scientist Theodore Lowi echoed this view in his book The End of Liberalism, in which he urged adoption of a “tenure of statutes” act that would require statutes authorizing administrative agencies to be periodically renewed. The idea was to require periodic reevaluation and review of administrative agencies so as to provide opportunities to eliminate wasteful or unneeded programs and bring wayward bureaucracies to heel.

Interest in sunset provisions for administrative agencies peaked in the 1970s, largely in reaction to widespread mistrust of government institutions. Inspired by Lowi, the watchdog group Common Cause pushed for the adoption of “sunset” clauses at the state level. Beginning in Colorado in 1976, this movement quickly spread across the United States. Within five years, sunset statutes of one sort or another had been adopted in 36 states. The details of those statutes varied from state to state, as did their success. As a general matter, the various state sunset laws required periodic review and reauthorization of state agencies. Some required extensive (and costly) review and evaluation prior to the sunset.

There are a host of arguments in favor of sunset provisions. The most obvious is that they increase the likelihood of culling outdated laws, programs, and agencies. Over time, things change and what was once necessary may no longer be. In the alternative, an agency may remain necessary but in dire need of reform. Sunset provisions can serve as an effective oversight tool when properly employed.

Regular reauthorization ensures that existing delegations of authority retain a degree of democratic legitimacy that may be absent when agencies continue to operate on authority granted to them years or decades earlier.

Limiting the duration of legislative authorization can have broad effects on the incentives faced by legislatures and the actions taken by administrative agencies. Most obviously, limiting the duration of legislation reduces the ability of legislative majorities to entrench their policy preferences and benefits contemporary majorities relative to their predecessors. In the context of regulatory programs, limiting legislative duration tends to strengthen the hand of the legislature relative to the executive. But perhaps most importantly, regular reauthorization ensures that existing delegations of authority retain a degree of democratic legitimacy that may be absent when agencies continue to operate on authority granted to them years, if not decades, before.

Reauthorization Today

Although Congress rarely takes steps to incentivize (let alone require) reauthorization of major regulatory statutes, it does utilize temporary legislation to induce periodic reengagement or reauthorization of potentially controversial or problematic programs across a range of policy areas. The use of temporary legislation in these contexts does not always produce more efficient or effective governance, but it does appear to produce policy outcomes that are consistent with contemporary legislative priorities and responsive to democratic demands.

Perhaps the best-known reauthorization legislation is the farm bill. A product of the Great Depression, this omnibus legislation delegates a wide range of authority to the U.S. Department of Agriculture. Typically, the farm bill requires reauthorization every five years. When Congress fails to do this, expiration has various consequences, including the reversion of key program parameters back to those dictated by the first farm bill, the Agricultural Adjustment Act of 1933. This broadly undesirable default baseline appears to provide ample incentive for regular reauthorization.

The Federal Aviation Administration is another example, as the federal agency must be routinely reauthorized. Although this often involves little more than updating various deadlines, the requirement that Congress do this causes legislators to focus on the agency, even if only for a moment, providing an opportunity to update and revise relevant requirements.

The Food and Drug Administration’s user-fee programs likewise require periodic reauthorization. Beginning in the 1990s, Congress passed several acts authorizing the FDA to implement the programs, which provide funds to improve the efficiency of relevant operations. The programs must be and have been reauthorized every five years. Most recently, with the FDA Reauthorization Act of 2017, four of these user-fee programs were reauthorized into 2022 as a collective. The act also updated the enabling statutes for these and other programs through clarifying revisions and some substantive modifications. While none of the revisions appear to be comprehensive, the routine reauthorizations keep these user-fee programs fine-tuned and in good working order. Reauthorization occurs regularly because the failure to do this would revert the FDA drug approval process to being based solely on congressional appropriations — wholly inadequate to timely process drug approval requests — and may require the FDA to lay off agency officials whose salaries the user fees fund.

Reauthorization as a Response to Nondelegation

The temporal delegation problem has taken on added significance with the decline in lawmaking by legislation and the rise of lawmaking by regulation. Although counting words, pages, and laws is by no means a flawless method for capturing the extent of this trend in federal lawmaking, it provides at least an imperfect snapshot. For instance, by the end of 2016, the Code of Federal Regulations exceeded 175,000 pages, 100 million words, and tens of thousands of agency rules. In 2016, federal agencies reached a new regulatory record by filling over 95,000 pages of the Federal Register with adopted rules, proposed rules, and notices — nearly 20% more than the 80,000 or so pages published in 2015. Roughly two-fifths of those pages in 2016 were devoted to 3,853 final rules, an increase from the 3,410 final rules federal agencies promulgated in 2015. By contrast, the 114th Congress, over that same two-year period, enacted just 329 public laws for a total of 3,036 pages in the Statutes at Large.

In other words, we live in an era when the vast majority of federal lawmaking does not take place in Congress, but within the hundreds of federal agencies spread across the modern regulatory state. And such lawmaking is often undertaken using authority Congress delegated decades before, based on legislative compromises to address different problems.

The time problem of delegation would be cured were Congress to legislate more regularly — especially were it to more jealously guard the power it now delegates to the president and the regulatory state. And one way to induce Congress to legislate more regularly — and ensure that the legislative instructions agencies follow are responsive to contemporary democratic concerns — is to utilize temporary legislation and mandatory reauthorizations.

At one extreme, Congress could consider enacting a universal sunset statute that would require the reauthorization of any federal agency or program within a certain number of years. Many state sunset laws, for instance, applied across the board. The failure to reauthorize would lead the sun to set on an entire agency or program, thus barring any subsequent appropriation.

This one-size-fits-all approach would be bold, yet foolish. Statutes vary and action-forcing reforms may not be appropriate for all regulatory contexts. For some federal programs and perhaps some entire federal agencies, it might make sense to incorporate express sunset provisions. Such a blanket sun-setting threat would force Congress to take a fresh look at the agency’s regulatory activities and whether the program or agency continues to effectively fulfill the purpose for which Congress created it. Yet Congress should seek to use mandatory reauthorization provisions, sunsets, and other forms of temporary legislation to induce more regular legislative action and engage greater political accountability.

Congress does not face a binary choice between a complete sunset of an agency or program and permanent legislation. It may choose to incorporate statutory “sunset defaults” to which the agency or program resets if not reauthorized within a set time period, as is done with the farm bill. Likewise, in some regulatory contexts, it might be advantageous to set the sunset default as something that would force Congress to revisit and reauthorize the agency or program. In the case of regulatory agencies, the lack of authorization could mean that an agency lacks the ability to act with the force of law. In effect, without a valid authorization, it could not be said that the agency has been delegated such authority.

Authorization for the Clean Air Act, to take one example, expired in 1998. Under this hypothetical proposal, the EPA would now lack the ability to promulgate new regulations, issue new permits to regulated facilities, and perhaps even initiate new enforcement actions unless and until the act was reauthorized. The expired authorization would not affect the validity of regulations already promulgated, however, nor would it prevent state-level enforcement under previously approved state implementation plans or the filing of citizen suits against facilities for violating existing permits, regulations, or statutory provisions. Nonetheless, the consequences of allowing the authorization to expire would provide ample incentive for environmental groups and regulated firms to come to the table, as each would find the default baseline undesirable. That, in turn, would provide Congress with the opportunity and need to revisit and reconsider particularly obsolete or ineffective provisions in the law.

Conclusion

Temporary legislation and mandatory reauthorizations are not a silver bullet to address the problems of excessive delegation and the lack of democratic accountability within the administrative state. Yet, were Congress to utilize this strategy more often, it would create more opportunities to discipline existing delegations and hold legislators accountable for the policies and measures their legislative actions endorse.

Although five Supreme Court justices have expressed interest in revitalizing the nondelegation doctrine, there are reasons to suspect the Court is unlikely to do much to curb delegation in the coming years. Consequently, Congress will continue to face myriad incentives to delegate broad statutory authority to federal agencies and few incentives to revisit those broad delegations. And the president and federal agencies will continue to leverage such delegated authority. It will be difficult to change the legislative process (or constitutional doctrine) to decrease the breadth of statutory delegation to federal agencies.

Insofar as delegation is a problem, perhaps it is time to focus more on its causes. Accordingly, Congress needs to return to a regular practice of legislating and, in so doing, revisit prior delegations of authority to federal agencies. To encourage such legislative action, Congress should engage in regular reauthorization of federal agencies and programs and should take seriously its foundational rule against appropriation without authorization. Such action would not cure all that ails the modern administrative state, but it would ensure administrative action has greater democratic legitimacy and could induce the legislature to engage more fully in the question of what authority federal agencies should have.

Readings

  • “Administrative Collusion: How Delegation Diminishes the Collective Congress,” by Neomi Rao. New York University Law Review 90(5): 1463–1526 (2015).
  • “Dynamic Legislation,” by Rebecca M. Kysar. University of Pennsylvania Law Review 167(4): 809–868 (2019).
  • “Hiding Nondelegation in Mouseholes,” by Jacob Loshin and Aaron Nielson. Administrative Law Review 62(1): 19–68 (2010).
  • “ ‘I’m Leavin’ It (All) Up to You’: Gundy and the (Sort-of) Resurrection of the Subdelegation Doctrine,” by Gary Lawson. Cato Supreme Court Review 2018–2019: 31–68.
  • Power Without Responsibility: How Congress Abuses People through Delegation, by David Schoenbrod. Yale University Press, 1993.
  • Sunset Legislation in the States: Balancing the Legislature and the Executive, by Brian Baugus and Feler Bose. Mercatus Center at George Mason University, 2015.
  • “Temporary Legislation,” by Jacob E. Gersen. University of Chicago Law Review 74(1): 247–298 (2007).