As a practical matter, this norm can be enforced only partially—there is only so far into the regulatory weeds that even the most diligent Congress can go. But partial enforcement is far preferable to no enforcement, which is where we are now. "Underenforced constitutional norms," to use University of Texas at Austin law professor Lawrence Sager's term, include the Fourteenth Amendment's equal protection clause and have an honored place in the constitutional firmament. The question is how to navigate various controversies in order to restore congressional responsibility to that firmament.
Controversy churns about the "administrative state"—agencies that de facto make laws and have been criticized for breaking free of democratic control. Controversy also churns about the composition of the Supreme Court, in part because some justices have said the Court's job includes enforcing the constitutional norm that law be made not by agencies but by an elected Congress. Both controversies shape the extent to which we live in a republic: a state controlled by the public rather than a political boss or an elite.
A Personal Oddessy
A surprising light is thrown on all this by my experience long-ago in the 1970s, trying to enforce the Clean Air Act (CAA). I was an attorney for the Natural Resources Defense Council (NRDC), which was on the losing side in Chevron. The members of Congress who wrote the statute said the CAA would "protect health" from all pollutants with "an adequate margin of safety" by the end of the 1970s. But they left almost all the specifics of what that entails—which is to say, the real policymaking—to the Environmental Protection Agency and the states.
The legislation benefited plenty of lawyers on both the corporate and environmental side. Speaking personally, I won major lawsuits, got my name in the news, and received foundation grants. The corporate lawyers collected big fees. Yet, because Congress didn't do the hard work of lawmaking, tens of thousands of my clients died, as I explain below.
I left the NRDC and went into academia to understand why the CAA fell so short of the politicians' promise and to try to do something about it. Try I did. In articles and a book, Power Without Responsibility (Yale University Press, 1995), I argued that the courts should enforce the Constitution's requirement that lawmakers in Congress must vote for the laws and thereby shoulder personal responsibility for their consequences. Angry at what had happened to my clients, I argued that Congress could have—should have—made the law. However, at the time, I didn't sufficiently consider the practical and political difficulties of the Court holding unconstitutional statutes that failed to do so.
In other articles and a book, Saving Our Environment from Washington (Yale University Press, 2005), I pushed for Congress to adopt a statute that would put responsibility on the elected lawmakers by requiring them to vote on the major agency-made laws . In response to this and other arguments, Congress and President Bill Clinton approved the Congressional Review Act. The legislation was derived from my proposal, but with a perverse change. The requirement that lawmakers vote on agency-made laws was turned into an option to vote on them, which they usually don't take. (See "Cleaning Out the Statutory Junk," Summer 2018.)
Despite these failures, a promising new path to success has recently appeared with the likely modification of the Chevron doctrine. By basing the modification on the constitutional norm, the Court could reveal that members of Congress fall far short of their duty under the Constitution to take responsibility for the laws. This would be politically powerful because polls show that the public strongly prefers Congress—rather than the administration—to make policy. That is why members of Congress cover up their shirking by pretending to want responsibility.
To point the way, this article explains the essence of our republic, why it has declined, and in particular how the Court got trapped into participating in the cover-up, and finally the way for the Court to extricate itself and our republic.
Our Republic
With the Declaration of Independence having proclaimed that governments derive "their just powers from the consent of the governed," the Constitution sought to empower the people to sack key policymakers. Article I vests "all legislative powers herein granted" in a House of Representatives and a Senate legislating in tandem with a president and mandates roll call votes on controversial issues. That requires these officials to take personal responsibility for the hard choices. There could be "no taxation without representation" or, for that matter, regulation without representation. It is no wonder then that school civics courses once taught that it's Congress's job to make the laws and that its members are called "lawmakers."
Article I meant that Congress had to make the law itself rather than delegating that job to others. Debate at the Constitutional Convention proceeded on that premise. The power to regulate was understood at the beginning of the republic to be the power to make the rules of private conduct. In Federalist 75, Alexander Hamilton wrote, "The essence of the legislative authority is to enact laws, or, in other words, to prescribe rules for the regulation of the society." Thus, to regulate, Congress must itself state the rules that bind the public in understandable terms, such as a rule limiting the amount of pollution from designated factories. The enacted rule would thus allow voters to hold their representatives responsible for the consequences of lawmaking. That serves the bedrock purpose of Article I.
In contrast, a statute like the CAA that tells an agency to make rules to "protect health" states a goal rather than a rule. As such, the legislation conflicts with Article I. Stating goals is insufficient because Congress avoids responsibility for resolving major political controversies that ensue from pursuing the goal. For example, "protect health" is a pleasing goal, but when Congress passed the Clean Air Act of 1970 most pollutants were known to cause some harm to health at almost any ambient level conceivably attainable by end of the 1970s. Sound policymaking would thus require Congress to strike a balance between health protection and feasibility. Further, Congress would have to decide which pollution sources would bear the burden of cutting emissions sufficiently to achieve the desired ambient level. Such decisions entail heavy political cost for members, so they punt them to the agencies. That gives legislators plausible deniability for unpopular consequences of the rules that get announced on agency letterhead.
Of course, even a carefully drafted, legislated rule will require interpretation in some cases. Nonetheless, rule interpretation differs from rulemaking. Interpretation is based on evidence of how the legislature that enacted the statute would have clarified the law's ambiguities rather than on what makes good policy sense to the interpreter. One source of such evidence is the rule itself because, by making clear what's required in most cases, it reveals the relative weight the legislature gave to conflicting policy goals such as enhancing regulatory protection vs. minimizing regulatory burdens.
In Wayman v. Southard (1825), the Court defined Congress's regulatory job in a way that jives with my Congress-must-state-the-rule idea. The Court held that Congress must deal with "important subjects" that Article I assigns to it, but may allow others to "fill up the details." Other officials could "vary minor regulations which are within the broad outlines marked out by the legislation in directing the execution." "Fill up the details" can be understood as either part of a second definition of the constitutional norm or a way to only partially enforce the state-the-rule definition of the norm. Either way, members of Congress would still have to take personal responsibility for the politically salient choices and so serve the purpose of Article I. I will not pause to find the exact words to reconcile the two definitions because, as illustrated by the CAA provision quoted above, Congress now comes nowhere close to complying with either. The point of this article is to show how the Court could begin to bring Congress closer to its constitutional duty rather than to state the exact perimeters of that duty.
Our Republic in Decline
For its first 70-plus years, Congress tried to deliver what the Constitution promised: enacting the rules of private conduct. The young nation had administrative officials who did important work, but with rare exception they applied enacted rules rather than themselves imposing rules.
The Progressive Movement, which grew strong toward the end of the 1800s, began to unwittingly push the republic down a slippery slope to the administrative state. Many of the Progressives believed in separation of powers, including a Congress that makes the law. They saw their statutes as empowering experts in agencies insulated from politics to use scientific methods to find correct ways to apply statutory law rather than to make law. For example, they thought of the Interstate Commerce Act of 1887 as authorizing such experts to apply a rule on railroad rates that courts had previously applied.
Later, however, it began to emerge that the experts were not insulated from politics and science did not provide correct answers to questions of policy. By then, Washington had many such agencies. Unwilling to roll all that back, in Hampton v. United States (1928) the Court held that Congress does its job if it states "an intelligible principle" to guide the agency. That might have been meant to be only slightly looser than "fill up the details." Yet, the statute upheld in that case gave the president broad rulemaking power on a hot topic—tariffs—and the Court failed to say how closely Congress's instructions must control the agency's choices.
Whatever the Court meant in 1928, only five years later Congress passed a statute that provided little control but granted vast powers to President Franklin D. Roosevelt to regulate industry in response to the Great Depression: the National Industrial Recovery Act (NIRA). The Italian dictator Benito Mussolini stated admiringly of Roosevelt's sway under the statute, "Ecco un ditatore"—that is, "Behold a dictator." In 1935, a unanimous Court, finding that the statute failed to state an intelligible principle, declared it unconstitutional.
Roosevelt infamously struck back at the Court by proposing a statute authorizing him to appoint additional justices. Congress did not pass this so-called "court-packing plan," but the president nonetheless prevailed. With retiring justices replaced by Roosevelt appointees and the emergencies of the Great Depression and World War II, the scolded Court upheld every regulatory statute it saw from then down to the present, including some whose "intelligible principle" was "the public interest." Whatever the Court meant by "intelligible principle" in 1935, it now means next to nothing.
The modern Supreme Court vows to stop Congress from abdicating its legislative power. But the Court trivializes that vow by holding that Congress does not do so when it states an intelligible principle, even if that principle is a vague goal. The justices upheld the CAA's "protect health" provision. The Court, in effect, lets the members of Congress, despite their self-interest, judge whether they have made themselves sufficiently responsible to voters. Thus, "consent of the governed" has become farce.
The Court's Errors
The Court has erred twice. First, it adopted a meaningless definition of Congress's job. Second, it does not acknowledge that Congress shirks its responsibilities.
Supreme mush / The Court justifies its part in this farce by claiming that judges lack a judicially manageable test of whether Congress violates its duty. Yet, the Court created this problem itself by replacing its early manageable definitions of Congress's job, both of which served to make the legislators responsible for the politically salient lawmaking choices. It has replaced that with the unintelligible "intelligible principle" test.
The Court's progressive abandonment of its early formulations is, however, understandable because of growing concerns about whether Congress could fully do the job. It became increasingly hard to imagine that Congress could enact all the federal rules as the nation grew in area and population, more activity spanned state boundaries, and the economy grew more complex.
Nonetheless, the Court's abandoning its earlier formulations was still an error because there are many ways to accommodate concerns of practicality other than trashing the Constitution's norm of legislative responsibility. For example, courts may desist from issuing an order to right a wrong when the order would create an "undue hardship." Undoing a massive swathe of federal statutes and regulations would massively harm the public.
Moreover, many constitutional standards of liability themselves contain carve-outs to avoid impracticality. Governments can act in ways that otherwise would deny equal protection of the laws if, depending on the discrimination, there is a "compelling state interest" or a "rational basis," which is a notoriously permissive test. For another example, governments can constrain the content of speech if the speech would create danger—the proverbial shouting "Fire!" in a crowded theater. Similarly, the Court might decide to refrain from enforcing Article I where Congress has left issues that are not politically salient to the agency. That could be seen as using Wayman's "fill up the details" concept to limit enforcement of the state-the-rule norm.
If the Court adopted the state-the-rule norm but indicated that it was willing to limit enforcement, Congress could meet the Court halfway by adopting the proposal discussed in the next section.
Capital shirking / Lacking any intelligible concept of Congress's job, the Court fails to acknowledge that Congress could come much closer to discharging it.
Anyone who observes Congress knows it shuns responsibility. This section focuses on one example of how it could shoulder more. It's a proposal that came from James Landis, once the New Deal's leading expert on administrative law and later dean of Harvard Law School. Major new regulations, he urged, should not take effect until Congress votes to approve them. Later, Judge (now Justice) Stephen Breyer published an essay showing that Congress could do such work. A statute could require both houses to vote on whether to approve agency actions by a deadline with limited debate and no filibuster, and then present the approvals to the president for signature.
Congress could find the time for this work. Members and their staffs would have the benefit of the agency's rulemaking record. There are about as many "major regulations," as defined by the Office of Information and Regulatory Affairs (OIRA) as votes on symbolic public laws such as the naming of post offices. Voting on major regulations would require legislators to shoulder more responsibility, but exercising legislative powers is in their job description while naming post offices isn't.
One might object that this would transfer rulemaking power from expert agencies to politicians. But agencies and their personnel have interests, so both Democratic and Republican presidents have subjected them to control by OIRA within the Office of the President. Meanwhile, regulatory priorities have swung wildly from president to president. Whatever one thinks of presidential control of rulemaking, we don't necessarily have far-sighted, broad-minded wisdom at the center of it.
Although the legislators themselves would spend little time on each regulation and voters would not read them, the legislators would still feel responsibility for their long-term consequences. After all, a challenger in some future election could point out that the incumbent had voted for a rule that now has unpopular consequences. It is recorded votes on rules, not debate and sound bites on popular goals, that make members of Congress responsible for regulations in future elections.
Democratic legislators could hardly say the proposal is anathema to good government when it came from a leading New Dealer (Landis) and a Supreme Court justice who is an expert in regulation and was appointed by a Democratic president (Breyer). Nor can Republican legislators say this when their contingent in the House has repeatedly passed a bill that incorporates it—as I discuss below.
The errors' consequences / By letting the legislators, in effect, judge whether they do their duty and failing to acknowledge that they could do much more, the Court covers up for an irresponsible Congress. The cover-up matters because polling shows voters want Congress rather than the president to "make policies," as Yale political scientist David Mayhew reports in The Imprint of Congress (Yale University Press, 2017). In 1958, 1977, and 2004–2005, the margin of support for this position was overwhelming: about three to one. Knowing that, members of Congress pretend to want responsibility but in practice shun it—at a terrible cost to their constituents.
Consider these instances. The year 1970 brought the first Earth Day and sharp criticism, led by Ralph Nader, that people died from air pollution because a string of clean air laws Congress enacted in the 1950s and 1960s left the difficult policy decisions to federal agencies.
In response, Sen. Edmund Muskie, who had a hand in several of those earlier laws, authored the 1970 CAA, which he boasted made the "hard choices." But this statute also ducked the tough decisions, with disastrous consequences. Take, for example, the pollution that came from refiners using lead additives in gasoline. (See "The EPA's Faustian Bargain," Fall 2006.) The statute promised that health would be protected from lead completely by 1976. Yet, politicians on the left and right effectively blocked those protections until the mid-1980s—and relented only after the big refiners found that they could save money if the EPA banned lead additives to gasoline. In the mid-1970s, one-half of the children in New York City had blood lead levels of 19 micrograms per deciliter (μg/dL) or above. In comparison, President Barack Obama declared a state of emergency in Flint, MI in 2016 because 5% of the children had blood lead levels of 5 μg/dL or higher. Back in the 1970s, my medical experts told me that, although lead in paint caused tragically high lead levels in many children, the population-wide contamination came primarily from lead in gasoline. The unqualified promise that the CAA would "protect health" was a pious fraud.
I began to wonder what would have happened if Congress couldn't pass the buck on lead in gasoline. Doing nothing on lead wasn't an option because in 1970 "Getting the Lead Out" was a chief demand. Congress itself, in a singular exception to the statute's general flight from responsibility, decided that new cars had to emit 90% less of a list of pollutants by 1975, but left lead off the list. If Congress couldn't have passed the buck on lead, I suspect it would have reduced lead exposure by at least half by 1975.
This quicker start on lead would have averted about 50,000 deaths in the United States, about equal to American deaths in the Vietnam War. There should be a monument on the scale of the Vietnam Veterans Memorial to these victims of Congress's shirking. I suggest locating it on Capitol Hill. Nonetheless, the 1970 CAA was a political bonanza for members of Congress. They got credit for the popular promise of healthy air, yet the blame for the failures and the economic burdens fell on the EPA and the states.
Legislators continue to feign a desire for responsibility. For example, Republicans in the House have repeatedly passed the Regulations from the Executive in Need of Scrutiny (REINS) Act. REINS does include the Landis–Breyer proposal, but swaddled with anti-regulatory provisions that ensure it will never pass and so will never inflict responsibility upon them. (See "Cleaning Out the Statutory Junk," Summer 2018.) One such provision would require agencies to cut the cost of existing regulations to offset the cost of any new regulations. That allows Republicans who support REINS to take credit with their party's base for wanting to control regulatory costs, but they shift to agencies the blame for limiting regulatory protection. Meanwhile, Democrats who support existing regulatory statutes can take credit with their party's base for wanting regulatory protection but shift to agencies blame for the regulatory burdens. This is a perfect recipe for polarization.
Democrats, for their part, have introduced no alternative versions of Landis–Breyer stripped of the anti-regulatory sections. They prefer instead to blast Republicans for being against regulatory protection. The upshot is a legislative stalemate in which lawmakers from both parties escape responsibility for the laws. This illustrates in a microcosm how the polarization, brought on by both parties ducking responsibility, produces gridlock.
Congress should adopt the Landis–Breyer proposal. If legislators had to vote on major regulations, their responsibility would motivate them to update statutes to allow agencies to promulgate regulations that produce more regulatory protection bang for the buck. That would inflict responsibility on the "Honorables," but would be win–win for the constituents they claim to serve.
The Trap
The Court's errors might suggest that it cares not for the consent of the governed. But it has repeatedly said otherwise and has decided cases insisting that legislatures exercise legislative powers. For example, in Clinton v. New York (1998), justices from the left and right joined to strike down a 1996 statute in which Congress had given the president line-item veto power over budgets Congress had approved.
The Court does, however, seem indifferent to Congress doing its job when it comes to delegating to agencies the power to make rules of private conduct. On that, the Court got trapped by history. The Court fell for the Progressives' conceit that experts in agencies were applying science rather than making law. When that conceit lost credibility, the Court tried to maintain the status quo with the "intelligible principle" test. Then, faced with the exigencies of the Great Depression and World War II, the Court let administrative lawmaking mushroom, except for the NIRA.
The Court let this mushrooming continue after the war. To see why, consider what would happen if it strikes down a statute based upon some meaningful definition of Congress's job. The decision would come as a bolt from the blue. Chaos would ensue because many of the regulatory statutes in the U.S. Code would probably flunk this test. A huge swath of the U.S. Code and the Code of Federal Regulations would be at death's door. That would leave people in pervasive doubt as to their regulatory duties and protections.
In response, Congress could respond by simply passing a brief statute enacting much of the Code of Federal Regulations. But that would make a mockery of congressional responsibility. Besides, it would take time for Congress to decide exactly what to include and what to leave out of the statute. For example, there would be strong arguments to exclude regulations presently undergoing judicial review. Even after the enacting statute passed, Congress and agencies would struggle to meet the ongoing need for changes in statutes and regulations. These prospects mean the Court may face an assault on its independence more successful than Roosevelt's court-packing plan.
The Court might make itself a smaller target by stating that the constitutional norm would be only partially enforced, but deciding the scope of enforcement would require a judgment of policy. This would be far more contentious than those in other constitutional decisions because so much of our current regulatory infrastructure is based on the Court's consent to Congress shirking its duties and many members of Congress would be enraged. Private discussions with former justices from the left, right, and center left me with the impression that they would have liked to push Congress to do its job but were unsure how to escape the trap.
The Escape
In many institutional reform cases dealing with violations whose remedy depends upon policy judgments within the province of a reluctant legislature, lower federal courts and state courts have found that they can succeed by starting a conversation with the legislature about the remedy. My New York Law School colleague Ross Sandler and I discuss this at length in our book Democracy by Decree (Yale University Press, 2004). The legislatures in those cases were at the state or local level, rather than sitting high on Capitol Hill. But the Court has a potential ally that is even more powerful than Congress: public opinion.
A conversation is a better starting place than the guillotine of holding statutes unconstitutional. The Court can start the conversation through statutory interpretation. It has long held that courts should interpret statutes to reduce serious constitutional concerns where the statute would bear such an interpretation. So, courts have sometimes interpreted statutes to narrow delegations of lawmaking power, but the extent to which they do so is limited. The Court's leave-it-to-Congress approach makes it hard to imagine a serious constitutional concern with delegation.
Courts would have far more occasion to invoke the nondelegation canon of statutory construction if the Court recognized congressional responsibility for lawmaking as a partially enforced constitutional norm. An appropriate occasion seems near. In a dissenting opinion in City of Arlington v. Federal Communications Commission (2013), Chief Justice John Roberts expressed discomfort with Chevron. He wrote, "Before a court can defer to the agency's interpretation of the ambiguous terms …, it must determine for itself that Congress has delegated authority to the agency to issue those interpretations with the force of law." As the American Enterprise Institute's Peter Wallison shows in his excellent new book Judicial Fortitude (Encounter Books, 2018), five current Supreme Court justices (Roberts, Clarence Thomas, Samuel Alito, Neil Gorsuch, and Brett Kavanaugh) have expressed support for changing Chevron. The Court should. It is a pro-delegation canon of statutory construction.
The Court should not only modify Chevron, but do so based on the constitutional norm of congressional responsibility. The norm would raise serious constitutional issues about many if not most regulatory statutes and therefore occasion many decisions narrowing the scope of delegations. In so doing, the Court need not and should not decide the extent to which this norm needs to be only partially enforced. It need not because, although it would be impractical to strike down every statutory provision that empowers agencies to make law, the courts could read statutes to narrow delegations of legislative power where the words of the statute would bear the interpretation. It should not because the Court should use such decisions to call upon Congress to decide how to address these concerns of practicality.
Switching statutory interpretation from pro-delegation to anti-delegation would not in itself result in Congress taking much more responsibility, but it would result in teachable moments. With many cases in the courts at all levels dealing with statutes in which Congress arguably didn't do its job, the courts would have plenty of opportunities to show the public that Congress falls short of public demands as well as the constitutional principle that lawmakers set policy. In contrast, the Court now lends its prestige to covering up for Congress. The Court's refusal to remain a party to the coverup would tend to prod Congress to engage in the policymaking conversation.
So, too, would pressure falling on Congress from its bearing more blame for regulatory outcomes. Chevron often allows legislators to plausibly deny responsibility for regulatory protection not granted or regulatory burdens imposed by saying that the agency misused its discretion under the statute. After replacing Chevron, however, the courts would affirm that many interpretations are based upon what Congress previously wrote and, if the legislators don't like it, they should change it. So many of the hard choices that Congress avoided would return to haunt it.
Of course, Congress could continue to avoid taking responsibility—or, worse, it could cede legislative powers to agencies more openly. In response, the Court would need to consider invoking the Constitution to strike down grants to agencies of the power to make rules of private conduct, but do so in a way that is practical. Rolling out the guillotine would be easier after having tried a conversation first. The public having been educated, the Court's constitutional intervention would not come as a bolt from the blue.
Moreover, the Court would have made clear that it would have preferred that Congress had made the policy judgments about how to stop its evasion of the "consent of the governed." The Court might take practicality into account by, for example, limiting itself to new "major" regulations as now defined by OIRA. This would be an altogether judicially manageable standard.
Conclusion
The Court has no more supreme duty than judging compliance with the Constitution. None of the Constitution's goals is more supreme than "consent of the governed." Yet, in response to claims that Congress frustrates consent of the governed by outsourcing lawmaking to agencies, the Court outsources judgment to Congress. That's poetic injustice. And it's time for that to end.
