The Role of Congress in Monitoring Administrative Rulemaking

September 12, 1996 • Testimony


(This testimony draws heavily from material prepared by David Schoenbrod and Gene Healy for a forthcoming
Cato Institute Policy Analysis. )

In 1995 the 104th Congress attempted to deal with our $500 billion regulatory burden by regulating the regulators. Senate Bill 343, the Dole‐​Johnston Regulatory Reform Bill, would have grafted a new web of rules–mandating cost‐​benefit analysis and scientific risk‐ assessment–onto the impenetrably dense administrative structure that already exists.

Giving regulatory agencies a dose of their own medicine is a laudable idea, but it will not solve our current dilemma. Indeed, by focusing chiefly on the monetary costs imposed by the current regulatory regime, congressional reformers have misconstrued the nature of the threat that regime poses. Instead of reinventing the regulatory state, Congress should take back its power to make the law.

Since the New Deal, Congress has ceded more and more of its legislative authority to executive branch agencies. This delegation of lawmaking power is ill advised and illegitimate, for several reasons:

  • Delegation violates the Constitution, subverting the central structural principle of that document: the separation of powers.
  • Delegation severs the people from the law, undermining democracy by allowing vitally important decisions of governance to be made by unelected, unaccountable officials.
  • Delegation is a political shell game, allowing legislators to simultaneously support the benefits and oppose the costs of regulation.
  • Most importantly, by allowing those who enforce the law to make the law as well, delegation subjects the lives, liberty and property of Americans to arbitrary rule.

Reservations about delegation are not limited to one side of the political spectrum; recently, concerns about the extent to which Congress has relinquished its lawmaking authority have been expressed by civil libertarians such as the ACLU’s Nadine Strossen, good‐​government reformers like former Sen. Bill Bradley and Debra Knopman of the Progressive Policy Institute, committed Progressives such as the New Republic’s Jacob Weisberg, and constitutional originalists such as former Judge Robert Bork and Judge Douglas Ginsburg. Despite their disparate perspectives, these thinkers have in common a concern for the vitality of our republican system of government–a vitality that has been sapped by Congress’s refusal to take responsibility for the law. That vitality can only be reclaimed by forcing the peoples’ representatives to reclaim the law. Reclaiming the law will require a restoration of the scheme of separation of powers outlined by the Framers–a return, in other words, to the original design.

The Original Design

The separation of legislative, executive, and judicial powers is the central principle of our Constitution’s architecture. This structural principle, according to legal scholar Rebecca Brown, is “a vital part of a constitutional organism whose final cause is the protection of individual rights.” Indeed, it was because the powers of the federal government were both enumerated and separated that most of the delegates to the Constitutional Convention thought that individual liberty could be preserved without a Bill of Rights. Alexander Hamilton held that the Constitution’s system of separated and enumerated powers was “itself, in every rational sense, and to every useful purpose, A BILL OF RIGHTS.”

The doctrine of separation of powers attained its axiomatic status for the founding generation in part through the historical experience of the colonies in their struggle with Britain, and in part through the writings of a number of influential political theorists. The Declaration of Independence’s bill of particulars against George III indicted the British king for several violations of the principle, among them, subverting the independence of the colonial legislatures, and making “judges dependent on his will alone.” The doctrine had also been articulated by, among others, Locke, Blackstone, and, especially, Montesquieu, whom Madison called “the oracle.” As constitutional historian Forrest McDonald notes, “American republican ideologues could recite the central points of Montesquieu’s doctrine of separation of powers as if it had been a catechism.”

Like Montesquieu, the Framers viewed political liberty as a condition in which citizens are free from arbitrary power and can expect to be secure in their persons and property. As Montesquieu put it in The Spirit of the Laws, “The political liberty of the subject is a tranquillity of mind, arising from the opinion each person has of his safety. Concentration of two or more of the three classes of power– legislative, executive, judicial–in a single organ of government would destroy that tranquillity, for reasons that John Adams expressed succinctly in a pamphlet published in 1776: “Because a single assembly, possessed of all the powers of government, would make arbitrary laws for their own interest, execute all laws arbitrarily for their own interest, and adjudge all controversies in their own favor.” According the late Malcolm P. Sharp, “Solicitude for liberty and property, and not unreasonable fear of what majority rule might do to them” were the primary impetus behind the enshrinement of separation of powers in the various state constitutions and its role in shaping the federal constitution.

To the end of preserving individual liberty and the rule of law, therefore, the first three articles of the Constitution neatly apportion the legislative, executive, and judicial powers respectively, to three separate bodies. Article I states, “All legislative powers herein granted shall be vested in a Congress of the United States”; Article II vest‐​s the executive power in the president; and Article III provides that the judicial power shall be vested in the Supreme Court and any inferior courts Congress decides to create. Neither the Framers nor Montesquieu adhered to a doctrine of pure separation of powers–a theory that would hermetically seal each department from the others. But the deviations from that principle are few, and explicitly prescribed.’ Indeed, Madison devoted Federalist 47 to defending these minor deviations from a theory of pure separation, readily granting that, were the proposed constitution guilty of a tendency toward mixing the legislative, executive, and judicial powers, “no further arguments would be necessary to inspire a universal reprobation of the system.’,

The precise limits of each respective function are not defined within the text of the Constitution, but that does not mean that the differences between them are incapable of being discerned. In an elegant reduction Gary Lawson of the Northwestern University School of Law writes, “Consider, for example, a statute creating the Goodness and Niceness Commission and giving it power ‘to promulgate rules for the promotion of goodness and niceness in all areas within the power of Congress under the Constitution.’ ” Clearly, such a commission would both make and enforce the law. 2 The President participates in the legislative process via the presentment clause and his veto power. The Vice President is given the tie‐​breaking vote in the Senate. The Senate confirms treaties and important executive branch appointees. It also has the judicial power to try impeachments

Statutes that express goals, even specific ones, but leave it to the executive branch to generate, the rules binding on private conduct, delegate the power to make law, and are thus illegitimate. John Locke, whose authority among the founding generation was rivaled only by Montesquieu’s, held that the legislature “cannot transfer the power of making laws to any other hands, for it being but a delegated power from the people, they who have it cannot pass it over to others.”

A statute meeting the test of nondelegation should clearly resolve most cases that arise under it. A person interested in whether certain conduct is prohibited should, under such a statute, be able to discern the answer from reading it. All statutes require interpretation, but the job of a law interpreter in the executive or judicial branch is to look backward to what the lawmakers intended, rather than forward, to determine what would be wise public policy. Cornell University Law School professor Cynthia R. Farina states the relevant question thus, “Are decisions of public policy being made by someone other than those who the people have chosen as their representatives?” if so, then the statute in question fails the test of nondelegation contemplated by the Constitution. Under a revitalized nondelegation doctrine, there will indeed be hard cases‐ instances in which there is no “bright line” between interpreting the law and actually making it; however, the vast majority of regulatory rulemakings issued under the current system do not constitute hard cases.

The Constitution in Exile

Before the New Deal, wholesale delegation of legislative authority to the executive was largely unknown in the United States, at least during peacetime. with the coming of the Great Depression, President Franklin Delano Roosevelt sought sweeping authority to manage the U.S. economy. With the passage of the National Industrial Recovery Act of 1933, he got it. The NIRA authorized industrial and trade associations to draw up codes designed to raise prices and restrict production; if the president found the codes unacceptable, he was empowered to immediately issue and enforce them. Upon hearing of the NIRA, Benito Mussolini exclaimed, “Ecco un ditatore!” (“Behold a dictator!”)

In 1935 the Supreme Court emphatically rejected the industrial code provisions of the NIRA in A.L.A. Schecter Poultry Corp. v. United States. The Court, led by Chief Justice Hughes, argued that “Congress is not permitted. to abdicate or to transfer to others the essential legislative functions with which it is thus vested.” In his concurring opinion, Justice Cardozo famously characterized the industrial code provisions as “delegation running riot.” But after Roosevelt’s 1937 attempt to subvert the judiciary’s independence t,y enlarging the Court, the Court never again struck down a New Deal statute on delegation grounds. Fear of Court‐​packing concentrated the mind wonderfully, and the judiciary chose not to stand in the path of the administrative state.

The nondelegation doctrine joined the doctrine of enumerated powers in jurisprudential limbo, as an integral part of what Judge Douglas Ginsburg has called “the Constitution‐​in‐​Exile.” Along with their “textual cousins,” the Necessary and Proper, Contracts, Takings, and Commerce clauses, these doctrines have been, according to Ginsburg, “banished for standing in opposition to unlimited government.

By 1944 the Court recognized few if any limits on Congress’s ability to delegate. In Yakus v. United States it held that Congress could delegate to an executive agent the power to set maximum prices for virtually all goods throughout the economy. What has followed since the New Deal and the Second World War has been a line of cases in which “the judiciary typically waxes eloquent about the serious breach were Congress ever to transfer its legislative power to other parties;, after which it finds a way to uphold the delegation.”

That line of cases culminates in Chevron, U.S.A., Inc. v. Natural Resources Defense Council (1983), in which the Court showed extraordinary deference to administrative agencies, interpretations of their own authority. The Chevron case arose out of a dispute over the meaning of the term “source” in the 1977 amendments to the Clean Air Act. Initially, the Environmental Protection Agency under President Carter defined the term so that it applied to each source of emissions within any given factory. But under the Reagan administration, the EPA issued a more flexible rule that considered the plant as a whole to be the “source.” Though the Court found it impossible to discern a legislative intent with regard to this issue, it upheld the EPA’s decision, holding that when a statute is silent on a particular issue, Congress can be understood to have delegated the power to make the law to the agency. And, according to Justice Stevens’s majority opinion, “Such legislative regulations are given controlling weight unless they are arbitrary, capricious, or manifestly contrary to the statute.” Professor Cass Sunstein of the University of Chicago School of Law suggests that the Chevron precedent, which allows agencies to determine the extent and nature of their own authority, ignores the wisdom embodied in the old adage about trusting foxes to guard henhouses.

With the judiciary’s abdication of its constitutional role, we are left with a legal status quo that effectively centralizes all governing functions in the executive branch agency: Congress passes a statute endorsing a high‐​minded goal–accommodation of the handicapped, safe drinking water, protection of wildlife–the executive branch agency then issues and enforces the rules governing individual behavior; the judicial branch, for its part, grants “controlling weight” to the agency’s interpretations of its own authority. In this way, the modern administrative state comes perilously close to realizing the Framers’ definition of despotic government, articulated by James Madison in the Federalist 47: “The accumulation of all powers, legislative, executive, and judiciary, in the same hands, whether of one, a few, or many, and whether hereditary, self‐ appointed, or elective, may justly be pronounced the very definition of tyranny.”

Delegation Running Riot

The administrative state erected since the New Deal is a massive amalgamation of Professor Lawson’s “Goodness and Niceness Commissions.” In the service of broadly popular societal goals, Congress has delegated ever‐​increasing amounts of legislative authority to the executive branch. What follows are just a few illustrative examples of delegation’s role in modern government and the concomitant threats to civil liberty and government. This list makes no pretense of being exhaustive; huge swaths of our statutory law, and most of the Federal Register would fail the test of a revitalized nondelegation doctrine.

The FDA and Tobacco

On August 10, 1995 the Food and Drug Administration unveiled a proposed package of new regulations ostensibly designed to reduce teenage smoking. Among the proposals: (1) the FDA would ban all outdoor advertising within 1,000 feet of any playground, or elementary or secondary school; (2) in magazines that could conceivably be read by children, the agency would limit advertising to black text on white background; (3) all cigarette advertising would have to include the phrase “Cigarettes–A Nicotine Delivery Device”; (4) the agency would ban vending machines, self‐​service displays, sale and distribution by mail, individual cigarette sales, and cigarette packs of less than 20.

There are any number of constitutional objections to be made to the FDA’s proposal, among them that it: relies on an absurdly broad conception of the Commerce power; violates First Amendment protections of commercial speech; encroaches on state prerogatives under the Tenth Amendment–but most salient for the purposes of this paper is that the FDA’s proposal is based on a sweeping delegation of legislative authority.

The FDA’s proposal constituted a rather dramatic turnaround, since the agency had long held that it did not have the authority under the Food Drug and Cosmetic Act to regulate cigarettes unless the manufacturer made health claims. In fact, in 1977 the FDA commissioner rejected a petition filed by the anti‐​smoking activist group Action on Smoking and Health requesting that the FDA restrict the sale of cigarettes to pharmacies. ASH challenged the commissioner’s decision in the federal courts. In ASH v. Harris, the U.S. Court of Appeals upheld the commissioner’s decision, but made it clear that the FDA was free to take a more expansive view of its authority should the agency choose to do so: “Nothing in this opinion should suggest that the Administration is irrevocably bound to any long‐​standing interpretation and representations thereof to the legislative branch. An administrative agency is Clearly free to revise its interpretations. 24

Dr. David Kessler’s appointment as FDA commissioner in 1990 heralded the arrival of a new, more aggressive agency–one that was fully prepared to exploit such judicial deference. Kessler’s FDA crafted a creative interpretation of the Food Drug and Cosmetic Act– one which would allow the agency to significantly restrict tobacco products without endorsing outright prohibition. Rather than regulate tobacco itself as a drug–because cigarettes could not possibly be approved as safe and effective–the FDA bases its claim for jurisdiction over cigarettes on the 1976 Medical Device Amendments to the Food Drug and Cosmetic Act. The agency intends to regulate cigarettes using the restricted device provisions of that act. Specifically, the agency argues that cigarettes are nicotine delivery devices, and thus subject to regulation as a combination drug/​device product. “Chew” and “snuff,” tobacco leaves that are used orally, are also considered “devices” for the purposes of the regulations. The agency claims that its authority over restricted medical devices allows it to regulate cigarettes and other tobacco products without taking them off the market completely. This approach, according to the FDA, “affords the most … flexible mechanism for regulating the sale, distribution, and use of these products.”

The FDA’s proposed regulations were initially submitted to President Clinton, who approved them. The Washington Post’s front‐ page article on this development‐ led with this curious sentence: “President Clinton has given the Food and Drug Administration for the first time the authority to regulate cigarettes.” Though clearly inaccurate with regard to constitutional law–authority to enforce a statute through regulation derives from Congress, not the president– the Post writers’ phrasing accurately described the current legal environment of unrestrained executive authority. The same article reported that President Clinton promised to halt implementation of the FDA rulemaking if Congress would pass the proposed regulations into law.

Such regulatory blackmail demonstrates how far we’ve departed from our constitutional framework, in which Congress legislates and the executive branch enforces the law. Instead, pace Clinton’s suggestion, the executive branch can use its illegitimate and unconstitutional ability to make law as a bargaining chip to force Congress to legislate. Indeed since 32 senators voiced their opposition to the proposals in a December 28, 1995 letter to the FDA, it is safe to conclude that, absent executive blackmail, such restrictions could not be passed through normal constitutional channels.

The FDA’s ability to make the law stems from a combination of statutory vagueness and judicial deference. This dangerous combination has resulted in metastasizing authority for the agency, as its oversight of medical devices illustrates. The FDA’s definitional agility with regard to cigarettes is hardly its most expansive attempt to assert -jurisdiction using its medical device authority. In the last several years the agency has invoked that authority to claim oversight of such common consumer items as weight lifting equipment, mouthwash, sunglasses, shoe deodorizers, electric toothbrushes, and television remote controls.

Three months after releasing its draft proposal on cigarettes, the FDA demonstrated just how broadly it views its ill‐​defined powers to regulate medical‐ devices. On December 11, 1995 the agency employed that authority to head off a threat to‐​of all things–airline safety. Pilots for carriers serving Las Vegas had complained that the outdoor laser light shows put on by area casinos were occasionally blinding them and putting their passengers in jeopardy. The FDA, invoking its authority to regulate lasers as medical devices, imposed a moratorium on all laser light shows anywhere within 20 miles of the three airports serving Las Vegas. A spokesman for the FDA said that the agency would not hesitate “to extend its coverage to other locales or nationwide” if it becomes necessary.

Clearly, laser light shows are not intended to, and cannot be, used to diagnose, cure, treat, or mitigate disease, nor do they affect the structure or function of the human body. But FDA regulators do not view themselves as executive agents with defined and limited public authority. Instead, they see themselves as public guardians with an indeterminate and open‐​ended mandate to do good. Wetlands Regulation

With the Federal Water Pollution Control Act amendments of 1972, Congress delegated to the Army Corps of Engineers the authority to require permits for the dumping of dredged or fill materials into the “navigable waters” of the United States. By 1977 the Corps had defined its own mandate broadly enough to allow it to regulate wetlands, including “swamps, marshes, bogs and similar areas”– private property That was not “navigable” in the traditional sense of the word.

In U.S. v. Riverside Bayview Homes (1985) the Supreme Court upheld the Army Corps of Engineers, broad interpretation of its own authority. Citing Chevron, the Court, led by Justice Byron White, held that “an agency’s construction of a statute is entitled to deference if it is reasonable and not in conflict with the expressed intent of Congress.” To bolster the claim that the Corps’ definition of “navigable waters” was a reasonable interpretation of Congress’s intent, Justice White invoked the legislative history of the 1977 amendments to the Clean Water Act. Congressional critics of the Corps, power grab had attempted to insert a more restrictive definition of “waters” into the 1977 amendments. That definition, which would have limited the Corps to regulating waters that were actually navigable, passed the House but stalled in the Senate. Justice White argued that “A refusal by Congress to overrule an agency’s construction of legislation is at least some evidence of the reasonableness of that construction.” But this is to turn the Constitution on its head: the Framers erected significant barriers to the passage of legislation in an attempt to ensure that each new rule binding on private conduct would be duly considered. Under the Constitution, a law must meet with the approval, or at least the acquiescence, of the representatives of three different constituencies: the House, the Senate, and the President. But when Congress is allowed to delegate its legislative authority, the executive branch agency makes the law, and all the constitutional hurdles that are supposed to stand in the way of frivolous lawmaking then obstruct those seeking to repeal frivolous executive‐​branch lawmaking.

Emboldened by the Court’s approval, the Army Corps of Engineers issued an even more expansive definition of navigable waters. By 1987 “navigable waters” had come to mean land that contained certain kinds of vegetation, soil hydrology, or was saturated with water for at least seven days a year.

In 1989 Ocie and Carey Mills, a father and son from Florida, ran afoul of the Corps’ metastasizing authority over land use. The Millses were found guilty of “discharging pollutants into the navigable waters of the United States,” The “waters” in question consisted of a wooded waterfront lot t‐​hat had no standing water on it. The Millses were sentenced to 21 months in jail each, and one year of parole. Though sympathetic to the Millses, plight, Judge Vinson of the U.S. District Court (N.D. Florida) found himself bound by precedent to uphold their conviction. He wrote: “A delegation doctrine which essentially allows Congress to abdicate its power to define the elements of a criminal offense, in favor of an unelected administrative agency such as the Corps of Engineers, does violence to this time‐​honored principle … Deferent and minimal judicial review of Congress’ transfer of its criminal lawmaking function to other bodies, in other branches, calls into question the vitality of the tripartite established by our Constitution. It also calls into question the nexus that must exist between the law so applied and simple logic and common sense. Yet that seems to be the state of the law.”

The Abortion Gag Rule

Supporters of abortion rights had reason to lament sweeping delegation of lawmaking authority in 1988, when Secretary of Health and Human Services Louis Sullivan decided to change the rules governing federally funded family planning organizations. Title X of the Public Health Service Act, enacted in 1970, authorized the Secretary of H.H.S. to make grants to and enter into contracts with public or nonprofit private clinics offering “a broad range of acceptable and effective family planning methods and services.” Though the legislation prohibited the use of Title X funds to pay for abortions, it was silent as to whether advice about abortion could be given at federally funded clinics. But almost 20 years after the passage of the initial legislation, H.H.S. Secretary Sullivan issued regulations that summarily forbade clinics receiving Title X funding to provide information about abortion.

In doing so, Secretary Sullivan implemented by fiat a policy that two years earlier had failed to garner a majority of votes in Congress. In 1986 Sen. Orrin Hatch (R-UT) and then‐​Rep. Jack F. Kemp (R-NY) introduced legislation that would have prohibited Title X clinics from discussing abortion; that legislation was rejected by Congress. Sullivan’s 1988 regulations accomplished what Hatch and Kemp could not.

Pro‐​choice advocates were outraged by Sullivan’s implementation of the gag rule. They argued that it had never been Congress’s intention to prevent clinics from advising their clients, often indigent women, about all safe, legal, and available medical options.

Sullivan’s action was challenged in federal court and eventually upheld by the Supreme Court in Rust v. Sullivan (1991). In his dissenting opinion, Justice Blackmun argued that the H.H.S. rules violated constitutional rights; he condemned the rules as “content‐ based regulation of speech” and an assault on abortion rights. Rehnquist’s majority opinion makes a convincing case that the regulations did not impinge on constitutional freedoms, since ” A legislature’s decision not to subsidize the exercise of a fundamental right does riot infringe the right.“ ‘ However, it was not a legislature that made this far‐​reaching decision, but an executive branch appointee, insulated from democratic control.

Nonetheless, the Court once again held that executive appointees have broad interpretive authority. Citing Chevron once again, the Court, led by Chief Justice Rehnquist, reasoned that it was not necessary to “dwell on the plain language of the Title XI statute because we agree with every court to have addressed the issue that the language is ambiguous … When we find that the legislative history is ambiguous and unenlightening on the matters with respect to which the regulations deal, we customarily defer to the expertise of the agency.” But, as was the case with wetlands regulation, the Court’s deference essentially placed lawmaking power in the hands of the executive agency and forced opponents of the rule to leap the procedural hurdles the Framers erected to protect liberty.

As the Framers intended, those hurdles proved difficult to surmount. Though popular opinion was against the gag rule–a 1991 Harris poll found that 78 percent of Americans thought Congress should overturn it–Congress was unable to pass vetoproof legislation repealing Sullivan’s regulations. President Bush twice vetoed legislation revoking the gag rule, and the rule survived until President Clinton overturned it by executive order on January 22, 1993. It thus took five years and two intervening presidential elections to revoke Louis Sullivan’s decree.

It could be argued that some of the examples above more clearly represent usurpations of statutory authority, rather than over broad delegations. For example, when Congress tasked the FDA with reviewing and approving new medical technology, it could not possibly have intended that the agency involve itself in airline safety. Nonetheless, rule by bureaucratic fiat is the inevitable product of a political culture conditioned by wholesale delegation of legislative authority. That political culture, and its effects on the behavior of executive branch regulators, was noted by James Landis, one of the leading legal theorists of the New Deal and one‐​time chairman of the Securities and Exchange Commission. During his tenure at the SEC, Landis observed that: “One of the ablest administrators … never read at least more than casually, the statutes he translated into reality. He assumed that they gave him power to deal with the broad problems of an industry, and upon that understanding he sought his own solutions.”

Having vested unelected officials with the power to make the law, legislators should not be surprised if their delegates interpret that power broadly. Indeed, given the current legal environment of promiscuous delegation on the part of the legislative branch, coupled with blithe deference on the part of the judiciary, it is little wonder that regulators conceive of themselves as having virtually unchecked power to do good.

The Dubious Benefits of Delegation

Clearly, wholesale delegation of lawmaking power comes with significant costs. Does it bring corresponding benefits? Defenders of the current regulatory regime argue that modern government cannot operate without delegation of legislative authority. Indeed, the Supreme Court said as much in a 1989 case involving a statute authorizing a commission to make rules governing criminal sentencing: “Our jurisprudence has been driven by a practical understanding that in our increasingly complex society, replete with ever changing and more technical problems, Congress simply cannot do its job absent an ability to delegate power under broad general directives.” This argument reveals misplaced priorities–it puts the alleged needs of the modern administrative state ahead of the question of constitutional legitimacy. Congress’s “job” after all, is to safeguard the framework of ordered liberty envisioned by the Constitution. Even so, the claims that are often made for the efficacy of delegation are vastly overblown.

Rule by Experts?

According to defenders of delegation, agency officials are experts who make technical decisions, and legislators are generalists who make broad policy decisions. But, as discussed above, Congress usually cannot delegate the technical issues in lawmaking without also delegating the broad issues of policy. Thus, lawmaking inevitably reflects moral judgments about how to balance and attain competing goals. According to political scientist Robert Dahl:

No intellectually defensible claim can be made that policy elites … possess superior moral knowledge or more specifically superior knowledge of what constitutes the public good. Indeed, we have some reason for thinking that specialization, which is the very ground for the influence of policy elites, may itself impair their capacity for moral judgment. Likewise precisely because the knowledge of the policy elites is specialized, their expert knowledge ordinarily provides too narrow a base for the instrumental judgments that an intelligent policy would require.

Perhaps for this reason, as well as because of the politics of the appointment process, most agency heads are not scientists, engineers, economists, or other kinds of technical experts. From the EPA’s inception in 1970, seven of its eight administrators and seven of its nine assistant administrators for air pollution have been lawyers. Moreover, as one observer has noted, “the New Deal concept of the ‘expert agency, breaks down in the modern context of health and environmental regulation. An agency addressing complex scientific, economic, and technological issues must draw upon so many different kinds of expertise that no individual employee can know very much about all of the issues involved in a typical rulemaking.”

Meanwhile, generalist legislators often vote on laws–such as those setting the emission limits for new cars–the merits of which depend upon the resolution of hotly contested technical disputes. Although both agency heads and legislators often lack the expertise to evaluate technical arguments by themselves, they can get help from agency staff, government institutes (for example, the Centers for Disease Control), and private sources (for example, medical associations, private think tanks, and university scientists). In addition, legislators request advice from their own staffs, committee staffs, and various congressional offices. By paying attention to the source, amount, and tenor of competing advice, both agency heads and legislators can make judgments involving technical issues without fully understanding them.

Another problem with the theory of agency expertise is the assumption that agencies are sufficiently insulated from politics to make their decisions scientifically, rather than politically. But, agencies are, of course, not really insulated from politics at all, but rather are subject to all kinds of subtle and not so subtle pressures from members of Congress and the White House staff. Agencies are vulnerable to such pressure because they and their staffs have interests of their own, such as getting wider powers, a larger budget, and access to higher appointed positions. Perhaps agency lawmaking is somewhat more removed from legislative politics than is congressional lawmaking, but, in acting behind closed doors to pressure agencies, members of Congress are largely free from electoral accountability.

Is Congress Too Busy?

CEOs of large private organizations usually delegate details to underlings in order to leave enough time to decide the broad issues of policy. New Dealers argued for delegation on similar grounds; “time spent on details by Congress must be at the sacrifice of time spent on matters of the broad public policy.’, Yet Congress does not act like an institution too short of time to get involved in details, especially as it has turned from broad to narrow delegation. For example, the Clean Air Act and many other statutes give agencies copious instructions on the handling of many complex questions. The 2,823-page-long Internal Revenue Code legislates in great detail, often creating rules so specialized that they apply to only one taxpayer. Congress legislates about details on an even more massive scale in the annual federal budget, which in 1991 grew to 1,527 printed pages on five and a half pounds of paper. That budget, like others, not only decides broad policy–such as the allocation of funds among major program categories–but also dictates tiny particulars of program administration. For example, Congress decided that $2.5 million of the $55.3 billion gross Department of Agriculture budget should go for the planning, design, and construction of a Poultry Disease Laboratory, and that it should be located in Athens, Georgia.

But in delegating to agencies, Congress often leaves open broad policy issues. Delegating major policy choices to a coordinate branch of government is altogether different than delegating details to underlings. Congress cannot do all of the agencies’ work, but it can make the laws–that is, the rules binding on private conduct, which, after all, is the job the Constitution assigns to Congress. Under the Constitution, Congress can appropriately leave to the executive and judicial branches other tasks, such as deciding how to enforce those rules (for example, interpreting the laws and exercising prosecutorial discretion), organizing and running agency operations (for example, assigning tasks to the staff, hiring employees, buying equipment), managing public enterprises (for example, the post office or other government operations or property), and making recommendations to Congress (for example, proposing changes in laws).

Legislated laws can be quite general. For example, one section of the 1990 Clean Air Act Amendments mandates that the EPA base emissions limitations for many categories of sources on the levels achieved by the cleanest 12 percent of the plants in each category. Through this general formula, Congress established a rule of conduct applicable to many pollutants from many kinds of sources by stating the criterion separating permissible and impermissible conduct.

Enacting laws forces legislators to take political responsibility for imposing regulatory costs and benefits. In contrast, delegation allows Congress to stay silent about what the agency will prohibit, thus severing the link between the legislator’s vote and the law, upon which democratic accountability depends.

Congress could achieve the public purposes that it now pursues through delegation in far less time than agencies take to make laws and in less time than delegation takes Congress in the long run. Acting by itself, Congress would not have to go through the same laborious processes that it requires of agencies. Congress currently accompanies delegation with detailed instructions on substance and procedure that constrain agency discretion. Writing such instructions would be unnecessary if Congress made the rules.

Congress could, however, ask for an agency’s help in drafting law. For instance, it could require the agency to propose statutory language, prepare supporting analyses, and hold hearings on proposals. The agency’s analysis undoubtedly would make use of the kind of information that now is considered in administrative rulemaking. The New Deals leading theoretician of the administrative process, James Landis, advocated exactly this approach. He wanted agencies to propose laws, but not promulgate them. Landis wrote that agencies would have a better chance of breaking the stalemates that often prevent them from protecting the public if they could act as “the technical agent s in the initiation of rules of’ conduct, yet at the same time … have the elected lawmakers share in the responsibility for their adoption.” As Landis recognized, since controversy often paralyses the administrative process, “it is an act of political wisdom to put back upon the shoulders of the Congress” responsibility for controversial choices.

Delegation saves Congress from political accountability, but it does not save time. Delegation is time‐​consuming because instructing agencies on how to make the law is a complex task, as the length of the various Clean Air Acts suggests. Moreover, the issues that one Congress ducks by delegating often reemerge to consume the time of succeeding Congresses. Although the 1970 Clean Air Act sailed through with hardly a dissenting vote, half of the sessions of Congress from 1970 to 1990 undertook major efforts to rewrite the act, in addition to the large amount of time spent overseeing implementation of the act and doing casework on EPA lawmaking.

Is Legislation Quick Enough? Some political leaders fear that the separation of powers mandated by the Constitution is unworkable because it leads to gridlock when the president and majorities in the House and the Senate do not all come from the same party. Those who use the negative term “gridlock,” however, ignore the fact that the political inertia it describes is an integral part of the American Constitutional design. The Framers believed that laws should not be made unless they have the broad support that usually is necessary to get them through the Article I process. As Madison put it in Federalist 62, the Constitution is designed to curb the “facility and excess of law‐​making.”

Some see delegation as a cure for divided government. Broad discretion allows agencies to make law without the permission of the House, Senate, or the president. However, because the president, the legislators, and their staffs influence the agency, the stalemate often continues after the delegation, but in a new context. Yet, because delegation has ostensibly given the agency the job of making the law the elected lawmakers can shift to the agency much of the blame for failing to resolve the dispute. Delegation thus short‐ circuits the nation’s only authoritative method of resolving disputes about what the law should be, and so puts protection of the public into an administrative limbo. The EPA’s delays in producing the rules required by the Clean Air Act are typical of what happens under many other statutes.

The purported ability of agencies to protect the public quickly is more apparent than real for other reasons. The Administrative Procedures Act theoretically allows agencies to make law in two months, and even less in an emergency. It is tempting to compare such potential speed with the years that can pass while bills languish in Congress. Yet Congress can react quickly when it senses public support for quick action, while agencies ordinarily need years to make law.

The Real Reason Congress Delegates

As the discussion above indicates, the typical rationales offered to support delegation are flimsy. Congress does not need to delegate in order to seek expert help; nor does Congress need delegation to ease its workload; still less does delegation help Congress avoid delays in addressing issues of broad public concern. Why then does Congress delegate?

One of the main reasons Congress delegates is to manipulate voter perceptions. Delegation allows legislators to represent themselves to some constituents as supporting an action and to others as opposing it. Legislators, I‐​or example, can write different letters about the same issue to different groups of constituents, with each letter crafted to make the legislator appear to sympathize with that group, position. Such letters are, of course, far more private than publicly recorded yea or nea votes. As former EPA administrator Lee Thomas described delegation under the Clean Air Act, “Everybody” is accountable and nobody is accountable under the way Congress is setting it up, but the legislators have got a designated whipping boy.”

Congress’s penchant for covering up its tracks via delegation is nowhere more starkly illustrated than in the congressional pay‐​raise controversy of 1988–89. In 1988 Congress used delegation to try to give its members a 50‐​percent pay raise without losing votes in the following election. It passed a statute that delegated to the Commission on Executive, Legislative, and Judicial Salaries the power to set pay for themselves and other top officials whose pay they linked to their own. Under the statute, if the commission were to grant a pay increase, another statute passed before–but not after– the increase went into effect could cancer it. When the commission recommended the 50‐​percent increase, some legislators introduced bills to cancel it. But this was part of a plan in which the congressional leadership would prevent a vote on the bills until it was too late to stop the increase. Legislators could then tell their constituents that they would have voted against the increase if given the chance‐ thus getting both the pay raise and the credit for opposing it.

However, the size of the increase, in an atmosphere of antipathy to Congress, provoked such a storm of protest and publicity that the public came to see through the charade. Embarrassed, the House leadership conducted a secret ballot among members to determine whether to hold a roll‐​call vote on the pay increase. Fifty‐​seven percent of the members who responded opposed a roll call vote, although 95 percent of the House members surveyed by Public Citizen claimed they had supported it. After public opposition to the pay raise approached 90 percent, Congress passed a bill to cancel it.

The pay raise controversy illustrates Congress’s willingness to use delegation to manipulate voters’ perception of its activities. In that particular case, manipulation failed–indeed backfired–because the public, aided by perceptive journalists, saw through the ruse. But manipulation through delegation is usually successful, because routine government action is neither so readily understood nor so pregnant with symbolic value as the pay raise was, and so eludes the sustained attention of the press and the public.

Not only does delegation let legislators avoid publicly recorded votes on hard choices, it also allows them to actively please conflicting interests by doing casework on their behalf. Casework, unlike roll‐​call voting, is not a matter of public record. Delegation thus allows members of Congress to function as ministers rather than legislators; they express popular aspirations and tend to their flocks rather than make hard choices.

Congress’s huge reelection rates are in part testimony to the fact that the delegation ruse generally works. During the 1980s only 88 of 2,175 congressional seats were turned over because of an incumbent’s defeat. With delegation members can usually escape being ejected from office except upon grounds that would oust a minister from the pulpit–scandal. In those exceptional cases in which incumbent legislators do lose elections, their defeat is far more likely to be caused by some escapade or by voting for a real law, such as a tax increase, than by how they shaped the law through delegation.

Getting There from Here

Despite the palpable political benefits of delegation–which allows congressmen simultaneously to support the benefits and oppose the costs of regulation–by the mid‐​1970s Congress began growing increasingly uneasy about the amount of power it had ceded to the executive branch. Popular complaints against arbitrary, capricious, and indecipherable regulatory laws began to have political effect, and Congress began to make noises about “reigning in the regulators.” That noise has grown louder by the Congress, until today it is one of the clearest themes of the celebrated congressional “Contract With America”.

The Legislative Veto

Rather than refrain from delegating legislative authority, Congress originally attempted to retain some control over executive branch agencies through the use of legislative veto provisions. The legislative veto usually takes the form of a clause in a statute that stated that executive branch action pursuant to the power delegated in the statute would take effect only if Congress did not veto it by resolution within a given period of time. Use of such clauses increased significantly with increased regulation during the ‘70s.

This legislative tool was declared unconstitutional by the Supreme Court in the 1983 case INS v. Chadha. Congress had authorized the Attorney General to use his discretion in suspending deportation proceedings for‐ selected “hardship cases” among illegal aliens. That discretion, however, was subject to disapproval via resolution by either house of Congress. Jagdish Rai Chadha was one of 340 illegal aliens whose deportation was suspended by the Attorney General in 1974. In 1975 the House of Representatives passed a resolution reinstituting deportation proceedings for Chadha and five others on the Attorney General’s list. A majority of the Court, led by Chief Justice Burger, held that this one‐​House veto provision violated the separation of powers embodied in the Constitution. According to Burger’s majority opinion, the legislative veto contained in the Immigration and Nationality Act allowed one House to make law without the participation of the other House and the president. According to the Court, the veto provision violated the Constitution’s Presentment Clause, article I, section 7, clause 3, which requires that “every Order, Resolution, or Vote to which the Concurrence of the Senate and House of Representatives may be necessary… shall be presented to thin President” for his signature or veto.

The Court’s reasoning was somewhat perplexing; as Martin Shapiro of U.C. Berkeley School of Law has pointed out, if the veto provisions are legislative in nature, and thus violate the Presentment clause, what of the regulations that are vetoed? Writes Shapiro, “If the congressional veto was unconstitutional because it failed to allow for a presidential veto, then the delegation of its rulemaking powers by Congress to the agencies was even more unconstitutional.” Nonetheless, the Court’s decision invalidated scores of legislative veto provisions contained in other statutes.

The Breyer Proposal

In a lecture given at the Georgetown University Law Center later that year, Judge Stephen Breyer, now associate justice on the Supreme Court, presented a plan for a “veto substitute” that would allow Congress to retain control of the law while following the requirements of Chadha. Breyer’s proposal would replace the legislative vetoes with statutory language stating that “the agency’s exercise of the authority to which the veto is attached is ineffective unless Congress enacts a confirmatory law within, say, sixty days.” Thus, under Breyer’s scheme, the executive branch would largely be stripped of lawmaking power; agencies would recommend particular courses of action, but they would not have the effect of law until they passed through the normal constitutional channels.

But how could Congress possibly handle the volume of rulemaking that modern administrative government is said to require? In his Georgetown lecture, Breyer suggested changing the House and Senate rules to allow a special “fast track” for proposed regulations subject to the confirmatory law requirement. Thus, under the new Senate rules Breyer envisioned, when an executive branch agency proposed rules subject to such a requirement, a bill containing the text of that regulation would be introduced automatically under the name of the Majority Leader. That bill would not be referred to committee, nor would it be amendable, debatable, nor subject to filibuster; instead, the Senate would vote yea or nea on the bill within 60 days of its introduction. The House would adopt similar rule changes.

Breyer’s proposal would allow Congress to follow the formal requirements of Chadha while preserving the substance of the legislative veto. Under Breyer’s plan, if one House disapproves of a regulation subject to congressional oversight, it can essentially “veto” it. But the confirmatory law requirement Breyer proposed would change the political dynamic considerably: “The veto substitute imposes on Congress a degree of visible responsibility for the actions it confirms, a burden that the veto system allowed it to avoid.”

The Nickles Amendment

Nineteen ninety four’s Republican takeover of Congress gave new impetus to regulatory reform, and‐ generated renewed interest in Breyer’s proposal. Sen. Don Nickles (R‐​Okla.) offered an amendment to the 1996 debt ceiling legislation that embodied a weak form of the Breyer proposal. Passed into law as P.L. 104121, it delays implementation of major regulatory rules, giving Congress 60 days to pass a joint resolution invalidating a proposed rule. That resolution would then have to be signed by the president. But as Rep. Nick Smith. (R‐​Mich.) has pointed out, the president is unlikely to sign a bill overriding a rule promulgated by his own administration. Thus, in many cases, the Nickles Amendment would require a two‐​thirds, supermajority vote by Congress to overturn a regulation. As such, it is little better than the status quo, since it requires the opponents of bad law to leap all the constitutional hurdles originally set in place to check overzealous lawmaking.

The Significant Regulation Oversight Act

Representative Smith has introduced a better bill, one that comes closer to the confirmatory law requirement envisioned by Breyer. H.R. 2990, the “Significant Regulation Oversight Act of 1996,” introduced on February 28, 1996, would require significant new rules to be affirmatively approved by both houses of Congress before going into effect. Which rules would be considered “significant” would be defined in the initial statutes providing for regulation. Thus, Congress would decide initially which types of rules could be passed by departments and agencies through the process outlined in the Administrative Procedures Act, and which would have to be legislatively enacted by Congress. For “significant” regulations, the agency would have to send its draft proposal to Congress.

Following Breyer’s recommendations, the Smith bill provides for rules changes in the House and Senate allowing for “fasttrack” consideration of regulations. The agency’s submission of a proposed regulation automatically creates a resolution to be introduced by the Majority Leader of each house. But, in contrast to Breyer’s scheme, that resolution then goes to the relevant committee. Within 45 days, the committee must decide whether to report the resolution or vote affirmatively not to report it. If it does neither within the allotted time, the resolution goes to the floor automatically for an up or down vote, no amendments permitted.

The Smith bill also includes a provision for revising or revoking regulations passed prior to the bill’s enactment. (Section 5) A petition to change or repeal such a regulation would be accepted when signed by 30 senators or 120 members of the House of Representatives. Such a petition would require the Majority Leader to introduce a joint resolution revising or repealing the regulation in question This provision would make it easier for regulatory reformers to force floor votes on controversial regulations. A minority of reformers in either house could force their colleagues to take publicly recorded stands on issues they might prefer to duck. Milton Friedman referred to the weight of existing regulatory legislation–hotly debated, but once passed, untouchable–as “the tyranny of the status quo.” Smith’s bill provides a legislative weapon that can be used to fight that tyranny. As Smith notes, “By placing regulatory power once more into the hands of officials that ordinary citizens could speak with, influence, and vote for, those citizens would retain more control over their lives.”

The Congressional Accountability Act

Freshman Rep. J.D. Hayworth (R‐​Ariz.), chairman of the House Constitutional Caucus, has introduced legislation that is more sweeping than either Rep. Smith’s bill or the Breyer proposal. Unlike Smith’s bill, H.R. 2727, the Congressional Responsibility Act of 1995, is not limited to “significant” regulations. Both the Smith bill and the Breyer proposal require Congress to affirmatively identify areas of authority that it wishes to subject to a confirmatory law requirement in contrast, the Hayworth bill leaves almost nothing to the agencies’ discretion: “This Act ends the practice whereby Congress delegates its responsibility for making regulations to unelected, unaccountable officials of the executive branch and requires that regulations proposed by agencies of the executive branch be affirmatively enacted by Congress before they become effective.” The only regulations that the Congressional Accountability Act would exempt from congressional review are regulations pertaining to agency organization, personnel, and the like.

Like the Smith bill, the Hayworth bill operates along the lines originally suggested by Breyer. Agencies must submit their proposed regulations to Congress, whereupon the Majority Leader of each house is to introduce a bill Enacting the regulation. Instead of the bill being referred to a committee. Under Hayworth’s framework, any member of the respective house can move to proceed to consideration of the proposed regulation. The bill is unamendable, and debate is limited to one hour. All such bills must be voted on within 60 calendar days of their introduction. However, if a majority of either house votes to suspend the “fast track” rules outlined above, the bill will be considered in the same manner as other bills.

Rep. Hayworth’s bill, if enacted, would represent an important first step towards ending the constitutional crisis caused by unrestrained delegation. one problem with the approach originally outlined by Breyer, and adopted by Rep. Smith, is that it allows Congress too much discretion over when to delegate. The Smith bill requires Congress to decide with each new statute, which decisions it would like to be held accountable for. Thus, the Smith approach requires Congress to strive continually not to delegate, despite the very real political benefits of doing so.

The Hayworth bill cuts the Gordian knot, defining regulation broadly at the outset, and holding Congress accountable for anything that can properly be construed as lawmaking.

The main defect of the Hayworth bill, however, is that it would not effect delegations of legislative authority that occurred before its enactment. The tyranny of the status quo would continue unabated even if the Congressional Responsibility Act were to pass. Some rules changes along the lines that Rep. Smith’s bill proposes–allowing petitions to expedite regulatory repeal–are therefore essential.

Regulation Without Delegation

Would the end of delegation spell the end of the regulatory state? Many of delegation’s defenders seem to think so. Theoretically, however, the entire code of federal regulations as it exists today could have been enacted under the rules changes proposed by Rep. Hayworth.

But of course it would not have been. The point here is not to show that under a revived nondelegation doctrine the current regulatory regime can survive unchanged. It is more likely that a return to nondelegation will mean a return to prescriptive laws, a new respect for federalism, and a renewed appreciation of the Framers’ view that the chief danger to republican government lies in legislative overzealousness, not legislative inaction. If Congress is to reclaim the law, it will be necessary for Congress to do less, do it properly, and be held accountable for the results.

Defenders of the administrative state view regulators, freedom from accountability as a virtue of the system. As FDA commissioner David Kessler puts it, “There’s a reason FDA commissioners aren’t elected.” Perhaps so, but it is not a reason that defenders of republican government are bound to respect. Indeed, the reason that Kessler hints at–regulators’ sweeping authority to act on what they perceive to be the public good, free from the meddling of the people’s representatives–is inimical to our free institutions as originally conceived by the Framers. In the original design, only judicial appointees were wholly insulated from public pressure. But, properly understood, the judiciary’s constitutional role makes it “the least dangerous branch.” Its power is essentially negative–it strikes down laws that violate the Constitution. The type of power Commissioner Kessler champions, and exercises, is of a different nature entirely. It is power over people, the power to make laws binding on private conduct. That power should not–must not–be exercised without responsibility.


Forty years ago 75 percent of Americans professed faith in the federal government to do the right‐ thing most of the time. Now three quarters tell pollsters that– they lack such faith. So strong is the public’s distrust of government that a third of respondents in a recent Gallup poll agreed that the federal government represents “an immediate threat to the rights and freedoms of ordinary citizens.” There has been much handwringing of late over this sea change in public opinion. Following E.J. Dionne, pundits and polls repeatedly ask why Americans hate politics. They lament the current political culture in which Americans lack faith in their government; feel that they have no influence in the political process; and curse the politicians who offer them nothing but stale platitudes and non‐​issues in every campaign.

Few analysts, however, have examined delegation’s contribution to this state of affairs. What plagues American political culture right now isn’t really politics, if by politics we mean open, daylight debate over the affairs of state. What plagues us is runaway administrative (government. The most important, far‐​reaching decisions in American government are no longer made by elected officials: they’re made by executive branch appointees. Americans are right to believe that they have no control over the engines of government. And they’re right to curse politicians who run on issues on which they can have very little impact: abortion, family values, support for “diversity,” and the like. over the past 60 years our elected representatives have abdicated their constitutional responsibility to make the law. A Fourth branch of government has effectively been created out of whole cloth. Our political culture will remain poisoned until we neutralize this branch and force Congress to accept direct accountability for its actions.

About the Author