After Courts: Democratizing Statutory Law

Ryan D. Doerfler* & Samuel Moyn** March, 2025

In Federalist No. 78, Alexander Hamilton argued for locating interpretive authority over law separately from those institutions tasked with formulating it. Hamilton’s vision, never accurate as a description of American practice, has not been credible for a long time. To the extent enormous power is still allocated to judges, our legal institutions have been out of step with our legal theory, which has long regarded them as political actors and policymakers. More practically, every term it is clearer and clearer that the role of the Supreme Court in statutory cases (including checking administrative rulemaking and other processes) is, if anything, more menacing than its role in the rare instances when it deploys its heaviest weaponry of constitutional invalidation. Against progressive calls to reclaim the judiciary, this Article extends our proposal to disempower courts exercising lawmaking authority—to include when they are interpreting statutes alone. Indeed, the same considerations that counsel the constitutional disempowerment of courts counsel their statutory disempowerment, and the allocation or reallocation of their authority over law to politically accountable agents. The heart of our Article offers a survey of court disempowerment strategies and tools, which are comparable to, though not identical with, the disempowerment mechanisms that have been proposed in the arena of constitutional reform. Such strategies and tools are appealing in the short term; but the long term requires a fuller rethinking of our institutions of legal interpretation. Available and existing disempowerment strategies for courts are best conceived as early and partial versions of full-scale allocation of interpretive authority over law to “political” branches and openly political control.

[Judges] have battered their way to supremacy with their double axes; one edge is the control over legislation by its unconstitutionality, the other is such free interpretation of statutes as suits their purposes.

—Learned Hand¹

Introduction
After the Hamiltonian Republic

Over just a few days at the end of October Term 2023, the Supreme Court did more damage to the powers of the administrative state than it had in almost a century. It was one thing for the court, in the more prominent case of Loper Bright Enterprises v. Raimondo, to put the final nail in the coffin of its deferential regime articulated in Chevron USA Inc. v. Natural Resources Defense Council, Inc., which had allowed agencies to resolve ambiguities in statutes unless their attempts to do so were clearly erroneous.² In two more 6–3 decisions (both with conservatives on one side and liberals another), Corner Post, Inc. v. Board of Governors of the Federal Reserve System and SEC v. Jarkesy, the Court sharply limited the powers of agencies in other ways—respectively, by effectively depriving agencies of statutes of limitations protections from suit against facial challenges to their rules, and by making it nearly impossible for them to enforce their rules on their own when certain kinds of interests are in play.³ In doing so, the Court reopened basic jurisprudential and political questions about institutional interpretive authority over statutory law. Initial responses to the cases ranged from proposals for damage control to rage that a past inter-branch, power-sharing settlement over statutory interpretation had been upset.⁴ There is an opportunity, however, to follow the Court in returning to first principles about control of statutory meaning—even if doing so will show that the Court is leaning further than ever into a juristocratic syndrome that democratic authority should displace.

In Federalist No. 78, Alexander Hamilton defended the life tenure of judges by locating the judiciary’s interpretive authority over law separately from those institutions tasked with formulating it.⁵ Such separation, Hamilton wrote, counted as “the best expedient which can be devised in any government, to secure a steady, upright, and impartial administration of the laws.”⁶ Hamilton’s vision, never accurate as a description of American law, has not been credible for a long time. It is not the case that “interpretation” of law in adjudication can be separated out conceptually, as Hamilton did in contrasting judicial “reason” with political “will.”⁷ Nor, more generally, is there a self-evident line between lawmaking and law-application that could ever track the formal identity of different government institutions. Therefore, there is no way to claim, as Hamilton did, that “[t]he interpretation of the laws is the proper and peculiar province of the courts.”⁸ All parts of government are involved in lawmaking and law application.⁹ Judges do not enjoy a monopoly on legal interpretation. Nor do they stay out of lawmaking through the interpretations they offer. Especially in a system with an apex court that resolves interpretive controversy without formal institutional constraints, there is a great deal of policymaking in adjudicative settings.

Our legal institutions have been out of step with our legal theory: To the extent massive power is still allocated to Article III judges, legal theory has long regarded them as political actors and policymakers. Since Legal Realism in the 1920s and ’30s, not to mention successor movements, no one save assorted formalists or neoformalists have doubted that courts’ “interpretive” power is political and of major significance to political life.¹⁰ But the reverse is also true: Our legal theory has long been out of step with our legal institutions. To the extent we have already seen the transfer of a great deal of interpretive authority over law outside the Article III judiciary, our institutional legal theory has not caught up either descriptively or prescriptively with what it would mean to acknowledge that legal interpretation, especially in the adjudicative setting, is a political act. Hamilton’s republic has long been dead, if it ever lived; this Article analyzes this reality and calls for disposing of the remnants of that republic for good. We can do without Article III courts and should mostly dispense with them.

Memorably calling it the “least dangerous [branch],”¹¹ Hamilton also stressed that the judiciary is institutionally weak.¹² But today it looks like its residual strength remains to be fully countered. Some successors of the Legal Realists, associated with the “Legal Process” school after World War II, discarded Hamilton’s premises but maintained his institutional assumptions that judges should play a pivotal role in our legal order.¹³ Once we abandon Hamilton’s pre-Realist premises about legal interpretation, the question is simply which institutions are best situated to decide on the content and future of the law and most accountable in doing so.

Federalist No. 78 is remembered for its defense—later partly reprised in Marbury v. Madison—of the specific role of judges in marking constitutional limits for other branches.¹⁴ But Hamilton was clear that his proposals amounted to a broad institutional theory of law, applicable to statutory interpretation too: “[I]t is not with a view to infractions of the Constitution only, that the independence of the judges may be an essential safeguard against the effects of occasional ill humors in the society.”¹⁵ He further wrote,

[J]udicial magistracy . . . . serves to moderate the immediate mischiefs of those [statutes] which may have been passed[. It also] operates as a check upon the legislative body in passing them; who, perceiving that obstacles to the success of iniquitous intention are to be expected from the scruples of the courts, are in a manner compelled, by the very motives of the injustice they meditate, to qualify their attempts.¹⁶

Though remembered for its theory of constitutional invalidation of statute, Federalist No. 78 offered a theory of judicial control over the legal order quite apart from that power. According to Hamilton, life-tenured and unelected judges played a defensible role in interpreting statutes for the substantive ends of containing the excesses of legislatures.¹⁷ Hamilton didn’t explain how these substantive ends—countering the “iniquity” or “mischief” that he associated with legislatures—fit with his theory that judges were mere interpreters of legislative enactment, rather than policymakers in their own right.¹⁸ Nor did he explain how judiciaries could be seen as reassuringly weak; their ability to block and moderate legislative power presupposed strength. Apart from any supposed distinction between “reason” and “will,” then, Hamilton’s substantive defense of judicial authority was candidly aristocratic and conservative. Our time shows, in turn, that this whole theory, not merely the noxious penchant of the Supreme Court to censor legislation, needs to be junked. Like the rest of Federalist No. 78, our prior arguments for democratizing the American legal order focused on judicial review.¹⁹ This Article pursues the far broader and probably more consequential power that judges retain over statutory interpretation and investigates mechanisms to curtail or eliminate that power. Every term it is clearer and clearer that the Supreme Court in statutory cases (including checking administrative rulemaking, enforcement power, and other processes) is, if anything, more menacing than its role in the rare instances when it deploys its heaviest weaponry of constitutional invalidation. Indeed, traditions of critique that obsessed about such invalidation on its own,²⁰ as well as mainstream expert processes that privileged reflection on the body’s constitutional authority,²¹ have failed to register both the scope of the problem with our judiciary and how far into remediation the republic already is. Recognizing these facts, this Article appraises what it would take to more fully complete the job.

Unlike popular calls to reclaim the judiciary, this Article completes our proposal to disempower Article III courts exercising lawmaking authority—including when they are interpreting statutes alone. For many progressives, statutory interpretation in courts is inevitable, and so political control through judicial appointment or even “court packing” is most important.²² Indeed, some have claimed that the residual interpretive powers of courts over statute, make arguments about constitutional disempowerment moot—on the theory that courts can easily circumvent any constraints on invalidating laws simply by reinterpreting them, as judges do every day. Other progressives regard courts as the most promising site of political agitation—especially if reclaimed through the appointment of friendly judges—in order to protect the vulnerable and weak from majority discrimination and popular oppression.²³

This Article argues, against all these perspectives, that the same considerations that counsel the disempowerment of courts to invalidate statutes counsel their disempowerment to interpret statutes. The Article then advocates for the allocation or reallocation of authority over law to politically accountable agents. Just as in constitutional law, statutory interpretation is frequently enough policy choice as to make it anything but obvious that such a task should be diverted from candidly political actors. It is precisely because policy-laden interpretation in the course of applying law is inevitable that empowered courts are neither exclusive nor ideal as a forum for that interpretation.

Part I starts outlining this vision by exploring why legal interpretation cannot be allocated (let alone exclusively so) to judiciaries as if they were apolitical in the sense Hamilton declared and that many lawyers—in spite of what they know—adopt the pretense of believing. The heart of the Article, Part II, then turns to what institutional transformation has already occurred in limiting the authority of courts over statutory interpretation and what further steps are possible, sooner or later. It offers a survey of disempowerment strategies and tools, which are comparable to, though not identical with, the disempowerment mechanisms that have been proposed to curb the judiciary’s power to invalidate statute.²⁴ Such strategies and tools are appealing in the short term; but in the long term, Part II argues, rethinking our institutions of legal interpretation is required. Indeed, as we show, available and existing disempowerment strategies for courts are best conceived as early and partial versions of full-scale allocation of interpretive authority to “political” branches and openly political control. Part III anticipates several objections, such as neo-Hamiltonian arguments that the “rule of law” requires a formally independent judiciary for statutory interpretation, as well as understandable worries that the political branches of government are not very democratic themselves.

Statutory Interpretation as Policymaking

Much of the motivation for democratizing courts comes from the recognition that statutory interpretation is often policymaking; judges frequently are, in other words, “politicians in robes.”²⁵ Throughout the 1920s and ’30s, Legal Realists observed the work of appellate courts in particular consists mostly of answering questions that lack “correct” or “unique” legal answers.²⁶ In statutory and common law cases alike, the availability of well-formed legal arguments in opposing directions meant that, consciously or not, judges were exercising discretion—appealing to their “sense of the situation,” as Karl Llewellyn put it—to resolve the disputes before them.²⁷ Developing and refining the positivist theory Realists presupposed,²⁸ H.L.A. Hart told a similar story, identifying what he called a “penumbra of uncertainty” around legal rules, a space in which a “judge must choose between alternatives.”²⁹ Into the 1970s and ’80s, various members of the Critical Legal Studies movement emerged as the intellectual successors to the Legal Realists from a half century ago. They situated jurisprudential observations about legal underdeterminacy within a more radical social theory, and, in some instances, advanced a more radical—and more philosophically dubious—account of underdeterminacy itself.³⁰ And even amidst the Reagan Revolution, conservative scholars and jurists like Frank Easterbrook and, to a lesser extent, Antonin Scalia,³¹ continued to emphasize the limits of law in appellate decisionmaking, with Easterbrook remarking, for example, that the “interesting questions in litigation involve statutes that are ambiguous when applied to a particular set of facts,” and that the “construction of an ambiguous document” is unavoidably a “work of judicial creation.”³²

Within the Realist tradition especially, the claim that judges enjoy substantial discretion rests upon observations concerning legal practice familiar to most attorneys.³³ To say that judging is political in the United States is largely uncontroversial when talking about constitutional cases adjudicated by the Supreme Court.³⁴ By attending to abstract guarantees like “equal protection of the laws”³⁵ or “freedom of speech,”³⁶ as well as murky constitutional histories and prior judicial decisions that are binding unless they are not, the justices purport to derive legal answers to assorted politically significant questions. In recent years, these questions have included whether or under what conditions institutions of higher education may consider race or ethnicity in admissions,³⁷ or the extent to which state legislatures may regulate the disclosure of physician prescription histories.³⁸

Recognizing the mismatch between the uncertain materials cited and the highly specific (and often conflicting) conclusions reached, few today insist that constitutional disputes addressed by the Court admit unique legal answers (though this may be changing among legal conservatives³⁹). Instead, the prevailing attitude among lawyers, at least, is that the justices in such cases enjoy limited but meaningful discretion (or, to use more euphemistic language, that deciding such cases involves the exercise of “judgment”).⁴⁰ But even if many or even most constitutional disputes are, in this sense, political, one might think that statutory cases are different.

Chief Justice John Marshall, for example, defended a more pragmatic approach to constitutional interpretation by observing the underlying document lacked “the prolixity of a legal code.”⁴¹ More recently, defenders of an apolitical Supreme Court have emphasized the Court’s ordinary statutory rulings, contrasted with its “blockbuster” constitutional decisions.⁴²

It is true that statutory law often is more specific than corresponding constitutional provisions. Statutes like the Civil Rights Act⁴³ or the Americans with Disabilities Act (ADA),⁴⁴ or legislative proposals like the Equality Act,⁴⁵ offer much more detailed protections against invidious discrimination than the Constitution’s promise of “equal protection.”⁴⁶ The same is true of the Voting Rights Act (VRA)⁴⁷ or the proposed For the People Act,⁴⁸ which articulate specific guarantees of electoral participation not readily discerned in the Fourteenth Amendment, let alone the “conception of political equality” that supposedly pervades the larger document.⁴⁹ And this is to say nothing of complex social legislation like the Medicare and Medicaid Act⁵⁰ or the Social Security Act,⁵¹ or of proposals like the Medicare for All Act⁵² or the College for All Act,⁵³ or legislation establishing organizing rights like the National Labor Relations Act (NLRA)⁵⁴ or the proposed Protecting the Right to Organize (PRO) Act,⁵⁵ notwithstanding efforts of some scholars⁵⁶ to tie them to the Constitution’s amorphous “guarantee” of a “Republican Form of Government.”⁵⁷ The reasons for these and similar disparities in detail are mostly practical. As Chief Justice Marshall observed, a constitution works less well as a unifying political object if it is unduly complex.⁵⁸ Statutes, meanwhile, are frequently addressed to judges and bureaucrats and, in that sense, are written by lawyers, for lawyers. Apart from accessibility, constitutions—including, notoriously, the U.S. Constitution—are often made more difficult to amend than ordinary law. Although everyday legislators are limited by foresight, constitutional drafters are even more constrained; they are less able to commit to specifics since constitutional mistakes, whether substantive or strategic, are more difficult to correct later on. But to say that legislators are generally less constrained than those who draft constitutional provisions is not to say that members of Congress legislate without constraint. They are also subject to material and other epistemic limitations, whether in the form of inadequate staffing, an excess of topics to be legislated upon, or an inability to predict social or technological change. Add to these limits the need to forge compromises across ideological divisions and, relatedly, the unpredictability of being able to reach such compromises in the future. Further, legislation is often relevantly nonspecific, leaving to later specification whether or how a statute should apply to circumstances unforeseen or unaddressed at the time. The ADA, for example, states with much greater clarity than the Equal Protection Clause that no persons with a “disability” shall be subject to certain forms of discrimination.⁵⁹ And yet, that same statute also leaves relatively underspecified which conditions are qualifying disabilities. ⁶⁰ Similarly, the VRA prohibits “denial or abridgement” of the otherwise inferred “right . . . to vote” based on “race or color.” Unaddressed, though, is whether, for instance, policies with a disparate racial impact, as opposed to invidious intent, constitute such denial or abridgment.⁶¹

Statutes, then, tend to be more specific than constitutions but also nonspecific in predictable and sometimes unpredictable ways. As the Legal Realists observed a century ago, it is where statutes are nonspecific that courts, and appellate courts in particular, are most often asked to weigh in.⁶² Given the cost of litigation, it would make little sense, for example, for some individual to sue the federal government claiming Medicare eligibility at sixty-four years old.⁶³ On the other hand, it might make a great deal of sense for a person or class of persons to appeal the decision not to designate some novel treatment as “medically necessary.”⁶⁴ Similarly, a recalcitrant university would be wasting its resources insisting it is not an “employer” for purposes of the NLRA.⁶⁵ And yet, it might be a wise, if morally dubious, investment to contest the status of certain graduate workers as “employees.”⁶⁶ These and myriad other examples suggest that the work of appellate courts, specifically in statutory cases, is specification. Where Congress has “spoken directly,” so to speak, economic rationality typically disfavors litigation that contests the apparent legal standard. Where statutes are “silent or ambiguous,” by contrast, interested parties often have reason to ask courts to (favorably) fill in the “gap.”⁶⁷

That this sort of interstitial lawmaking is so common is, ironically, sometimes offered as a reason why calls to disempower courts are misguided. Because it would be neither feasible nor desirable for Congress to try to specify everything in advance, the argument goes, we must rely upon courts to adjust the meaning of statutes to circumstances as they arise.⁶⁸ This line of reasoning mirrors familiar arguments for the use of legal “standards” rather than “rules.” As Louis Kaplow put it, because specifying the law’s content is too “expensive” ex ante, it will sometimes make sense for Congress to rely on “adjudicator[s]” to give the law content ex post.⁶⁹

To say the law should sometimes be developed case-by-case, though, is not the same as saying it should be developed by courts. Assuming Congress lacks the capacity to adjudicate such cases itself, this leaves open upon what sort of adjudicators it ought to rely. By insulating themselves from “political” pressures, courts distinguish themselves from other would-be adjudicators both in the popular consciousness and—in the federal system—as a matter of constitutional design.⁷⁰ Across the ideological spectrum, judges emphasize values like “neutrality” and “impartiality,” contrasting themselves with “politicians” and those whose decisions are “ideologically” driven.⁷¹ This self-presentation corresponds to constitutional structure, with judges insulated from democratic pressures by means of life-tenure and salary protection.⁷² At both the level of ideology and design, courts thus contrast with other adjudicators within the federal system—most immediately, administrative agencies.

Given the discretionary nature of most statutory cases, at least at the appellate stage, one might even ask, why should courts be involved at all?

Consider two recent examples. In Sackett v. EPA, the Supreme Court considered whether the Clean Water Act (CWA) permits the EPA to regulate backfilling of dirt and rocks into wetlands.⁷³ The CWA gives the EPA jurisdiction over the “discharge” of such materials into “navigable waters,” and, in turn, defines “navigable waters” as “the waters of the United States, including the territorial seas.”⁷⁴ Justice Samuel Alito observed, writing for the Court, that the meaning of “the waters of the United States” as used in the statute had been a site of contestation “[f]or more than a half century,” with both agencies and courts adopting varying interpretations over time.⁷⁵ In 2023, the EPA under President Joe Biden had construed the phrase to include all waters that could “affect interstate or foreign commerce,” including “neighboring” wetlands with a “significant nexus to a traditional navigable water,” a broader reading than the EPA had adopted five years earlier under President Donald Trump.⁷⁶ The Court, for its part, had substantially narrowed the meaning of “the waters of the United States” through a series of 5–4 decisions that rejected an earlier interpretation, according to which the phrase encompassed even “isolated waters” if those waters could be used by migratory birds and endangered species.⁷⁷ The conservative majority based its conclusion, in that case and others, largely on concerns about “federalism.”⁷⁸

In Sackett, all of the justices rejected the EPA’s interpretation.⁷⁹ Writing for five conservatives, Justice Alito explained the EPA’s reading put “property owners in a precarious position” because it was “often difficult to determine whether a particular piece of property” was covered by the CWA.⁸⁰ This, in turn, meant those looking to “build on their property” often needed to engage in a permitting process that “can take years and cost an exorbitant amount.”⁸¹ Determining that Congress had not articulated such burdens on ownership with adequate clarity, Justice Alito concluded that “the waters of the United States” encompasses only wetlands with a “continuous surface connection” to traditional navigable waters, such that it is “difficult to determine where the ‘water’ ends and the ‘wetland’ begins.”⁸² (In so concluding, Justice Alito swept aside the “ecological consequences” marshaled by the EPA in support of its reading, insisting “the CWA does not define the EPA’s jurisdiction based on ecological importance.”)⁸³ Justice Brett Kavanaugh and the liberals, meanwhile, agreed the Act did not cover the wetlands at issue but would have included within the CWA’s scope those “separated from a covered water only by a man-made dike or barrier, natural river berm, beach dune, or the like.”⁸⁴ As Justice Kavanaugh explained, excluding such wetlands did linguistic violence to the CWA’s specific inclusion of wetlands “adjacent” to covered waters; this effectively limited the Act’s coverage to “adjoining” wetlands.⁸⁵ Further, in a separate concurrence, the liberals criticized Justice Alito and those joining his opinion for placing a “thumb on the scale for property owners” despite Congress having enacted a statute that is “all about stopping property owners from polluting.”⁸⁶

Turning from civil to criminal: Dubin v. United States addressed whether a medical professional’s inclusion of a patient’s Medicaid reimbursement number in a falsified reimbursement claim constituted “[a]ggravated identify theft.”⁸⁷ Under the relevant statute, anyone who “uses” “a means of identification” “during [or] in relation to” healthcare fraud or other covered offenses is subject to an additional mandatory minimum sentence of two years in prison.⁸⁸ Federal prosecutors argued this was a “clear aggravated identity theft case” because the defendant’s “fraudulent billing,” which, among other things, exaggerated the services provided to the patient, included the patient’s reimbursement number—in that sense, they “use[d] . . . a means of identification” to “facilitate[] or further[]” the fraud.⁸⁹ The Court, again, unanimously rejected the Government’s interpretation. Justice Sonia Sotomayor, writing for eight Justices, observed that on this capacious reading “a waiter who serves flank steak but charges [the patron’s card] for filet mignon” would be guilty of “aggravated identify theft”;⁹⁰ so too a “lawyer who rounds up her hours from 2.9 to 3 and bills her client electronically.”⁹¹ Recognizing the absurdity of these outcomes, Justice Sotomayor rejected the Government’s suggestion that such cases could be avoided through prosecutorial discretion.⁹² Instead, she reasoned, the language at issue must be understood to include only circumstances in which the misuse of someone else’s “means of identification” is “at the crux” of the underlying offense.⁹³ That narrower reading separated “intuitive” cases of identity theft—for example, a “bank employee who passes along customer information to an accomplice”—from those like the earlier hypotheticals, where the use of identification was merely an “ancillary feature” of some billing method.⁹⁴

Justice Neil Gorsuch wrote a separate concurrence.⁹⁵ He praised the other Justices for rejecting the Government’s “unserious position” but lamented that the statute remained “not much better than a Rorschach test.”⁹⁶ The Court’s “crux” test presumably required some sort of “causation.”⁹⁷ But “how much ‘causation’[?] ” Justice Gorsuch asked.⁹⁸ The Court’s opinion offered little more than intuitions. With modest “tweaks,” though, “to the Court’s hypothetical[s],” intuitions begin to fade.⁹⁹ What if the restaurant charged the customer for “an appetizer he never ordered” or a “$5,000 bottle of Moët”?¹⁰⁰ What if the restaurant “waited an hour” to run these fraudulent charges?¹⁰¹ “A day? A year?”¹⁰² Although “debating” such scenarios “may seem good fun,” Justice Gorsuch worried that the Court’s test would “simply become a fig leaf for judges’ and jurors’ own subjective moral judgments.”¹⁰³

In both cases, the legal questions were highly specific. The underlying statutes, meanwhile, were relevantly not. The different Justices, in turn, resorted to open-ended moral reasoning to reach their varying conclusions, balancing property interests against ecological concerns, worrying about excessive punishment versus theft, and so on. And what else was there to do?¹⁰⁴ Where Congress leaves a gap in the law, filling it necessarily involves the exercise of discretion. Whatever one makes of the Justices’ individual moral assessments, there is still a threshold question: Why are these people tasked with making them? That is, why are such inherently political decisions reserved to persons specifically removed from the democratic process?

As we discuss in Part II, there are different ways one might try to limit the role of courts in statutory cases, ranging from less to more comprehensive. Here, though, it helps to ask simply, who else might decide? With Sackett, the answer is reasonably straightforward. The CWA empowers the EPA to enforce the statute directly,¹⁰⁵ which means that the agency, led by a political appointee directly accountable to the President and benefitting from the input of interested parties, articulates binding answers to interpretive questions about “the waters of the United States” and similar terms and phrases both through rulemaking and adjudication. One way to make interpretation of the CWA more democratic, then, would be to insulate (again, to varying degrees) EPA rules and orders from judicial review.

The story with Dubin is more complicated. Partly for constitutional reasons, authority to establish ex post criminal liability here and elsewhere rests with courts and, hypothetically, at least juries. Unlike the EPA, then, the Department of Justice (DOJ) cannot interpret the “aggravated identity theft” statute authoritatively through adjudication, and thankfully so, perhaps, given the carceral orientation of that institution. Under Title 18, the DOJ also lacks rulemaking authority and so cannot (again, maybe happily) create ex ante liability either.¹⁰⁶ Congress could, however, change that, or empower some other agency to implement clarifying rules. Beyond the executive, one could try to revive the American jury, once seen as a source of democratic participation in American criminal justice.¹⁰⁷ Or one could, as Justice Gorsuch suggested in his separate concurrence, try to send the question back to Congress.¹⁰⁸ Conceding that any specification of the vague language at issue would necessarily be legislative, Justice Gorsuch urged his colleagues not to enforce the statute until Congress had “fix[ed]” the “problem.”¹⁰⁹ But while Justice Gorsuch’s proposal relied on judicial self-restraint, Congress might restrain courts externally to a similar end, imposing, for example, some kind of voting rule that makes it more difficult for judges to create “federal common law crimes.”¹¹⁰

Implicit in Justice Gorsuch’s reasoning is that judgements of political morality are best left to openly political actors. That sentiment, shared here, has pervaded legal conservatism going back at least to the late 1970s. Embracing both originalism and textualism as signature methods of interpretation, “modern” and even “late” period conservatism has portrayed and continues to portray judging as an exercise in identification as opposed to decision.¹¹¹ (Chief Justice John Roberts’s analogy of judges and “umpires” in his confirmation testimony is perhaps the most famous example.)¹¹² Liberals, in turn, are criticized for engaging in “activism,” or, more crudely, for “making things up.”¹¹³

This “rhetoric of restraint” has done little to limit judges in practice.¹¹⁴ Some prominent conservatives like Judge Easterbrook and John Manning urged judges to stay their hand in statutory and constitutional cases, respectively, in the absence of clearly identifiable law.¹¹⁵ In the 1980s and ’90s, and even into the early 2000s, those urgings manifested in doctrine, albeit sporadically, with conservatives championing, for example, judicial deference to interpretations of statutes by administrating agencies. Even during this period, though, other conservatives, most notably Justice Scalia, limited the practical significance of such doctrinal commitments by (implausibly) declaring legal texts “clear” most of the time.¹¹⁶

In the years leading up to Loper Bright Enterprises v. Raimondo, conservative judges mostly ignored Chevron and similar doctrines, though continuing, in many cases, to declare legal texts “clear”—or at least “clear enough.”¹¹⁷ And in Loper Bright itself, Chief Justice Roberts made the genuinely stunning assertion that “statutes, no matter how impenetrable, do—in fact, must—have a single, best meaning.”¹¹⁸ A more recent spate of conservative scholarship concedes that statutory and other legal texts very often “run out,” but also insists (again, implausibly) that other “background” sources of law let judges decide cases without resorting to non-legal, that is, political reasons.¹¹⁹ This scholarship relates to more recent trends in the Court’s jurisprudence, including, most relevant here, the conservative Justices’ increasing reliance on “substantive” canons¹²⁰ like the “major questions doctrine” or various pro-federalism rules, as well as the Court’s broader turn to “history” and “tradition” in both constitutional and statutory disputes.¹²¹

If conservatives accept that discretionary judgments are “political” rather than “legal,” at least some liberals are more resistant. Justice Stephen Breyer, for example, insists that most disagreements among Supreme Court justices are “jurisprudential,” as opposed to “political.”¹²² “Politics,” for Justice Breyer, involves crude motivations like favoritism towards one’s political party or those to whom one owes one’s position—dispositions, he asserts, that are “not present at the Court.”¹²³ Instead, Justice Breyer attributes differences in how to understand phrases like “the waters of the United States” to differences in “judicial philosoph[y],” which, although to be contrasted with differences in “personal ideology,” may or may not be distinguishable from differences in “political philosophy.”¹²⁴ Setting aside his naivete about the occasionally transactional nature of judging, Justice Breyer concedes further the “difficult” cases that “come before [the] Court” are ones with “considerable merit on both sides,” such that a “judge’s background, experience, and personal views” inevitably “make a difference” in how they vote.¹²⁵

Theorizing within the liberal academy often lands in a similar place. Take legal philosopher Mitchell Berman, who, in an effort to vindicate the idea that “there is law in many hard cases,” says judicial reasoning in those cases is best understood as involving the application of multiple, potentially competing “principles.”¹²⁶ In disputes that implicate federalism, for example, Berman offers that a judge or justice may need to weigh a lawmaker’s apparent “legal intentions” against “historical practice” and “judicial precedent,” as well as the general sentiment that “states matter” within our federal system.¹²⁷ Each principle, Berman maintains, is appropriate and, indeed, necessary to consider as a matter of law.¹²⁸ Unlike his intellectual predecessor Ronald Dworkin, however, Berman does not insist there is only one permissible way to assign those principles relative weights. Indeed, what makes a case “hard” is that judges disagree about whether to weigh one (say, apparent legal intentions) or the other (say, that “states matter”) more heavily—with no way to adjudicate such disagreements beyond seeing which side gains more adherents.¹²⁹

For both Breyer and Berman, then, resolving hard cases involves attention to distinctly legal materials. And yet, what to make of those materials depends upon one’s “personal,” which is to say, ideological, disposition.

Other liberals worry less that the Court is political so long as it is favorable terrain. Buoyed by collective memories of the Warren Court and traumatized by the Reagan Revolution of the 1980s, many older liberals continue to think of judges as comparatively enlightened elites who can and should wield power in promotion of justice.¹³⁰ Against this juristocratic picture, these liberals contrast a system of “crude” majoritarianism in which the comparatively bigoted masses suppress, through their elected representatives, women, people of color, religious minorities, and so on.¹³¹

Although flattering to legal elites, this narrative about the relative progressive credentials of judges fares poorly in view of the Court’s actual history. As Nikolas Bowie and Daphna Renan have documented,¹³² the judiciary has played a malignant role, especially with respect to racial justice—most egregiously through antebellum and Reconstruction-era decisions like Dred Scott v. Sanford¹³³ and the Civil Rights Cases,¹³⁴ but also well into the late twentieth and early twenty-first centuries, with cases like Adarand Constructors, Inc. v. Peña¹³⁵ and Shelby County v. Holder.¹³⁶ And this is to say nothing of the material harm visited upon vulnerable populations through the negation of things like workplace protections,¹³⁷ welfare protections,¹³⁸ and organizing rights.¹³⁹

Given this track record, the preference for juristocracy over democracy among some liberals is both objectionable on its face and failed as a political strategy.¹⁴⁰ For progressives, at least, courts have proven remarkably hostile terrain, even compared to the decidedly uneven alternatives.

Mechanisms of Disempowerment

A. Disempowerment Between Reform and Reallocation

Once one distinguishes between interpretation and courts, the question becomes which institutions should apply, and therefore interpret, the law. Interpretation could—perhaps should—occur wholly outside of courts.

Allocation of interpretive authority beyond Article III courts has been a feature, not a bug, of American governance from its beginning.¹⁴¹ “Nineteenth-century federal courts subjected federal agencies to minimal oversight,” leading historians recently observed.¹⁴² This was true of lawmaking both by early bureaucratic federal entities and in early adjudicative settings, including legislatures at the state level.¹⁴³

In short, notwithstanding Alexander Hamilton, there was no commitment at the Founding to exclusively allocate lawmaking to legislatures, or forms of it involving fact-based application to formally independent judges. Classic examples were land allocation courts, military courts, and territorial courts—none Article III institutions and all extremely well attested in American history.¹⁴⁴ As Jerry Mashaw, Thomas Merrill, and others have shown, the early-to-mid twentieth century saw the crystallization of what Merrill dubs an “appellate review” model, which made federal courts more indispensable than before¹⁴⁵—even as the growth of the federal government across the same period also massively expanded non-Article III legal interpretation.¹⁴⁶

Regardless of that history, however, this Article makes a jurisprudential case for a post-Hamiltonian perspective. After all, the supremacy of courts in lawmaking became commonsensical, despite the vagaries of institutional practice, and Article III juristocracy remains something of an imaginative default even today.¹⁴⁷ Furthermore, the political right ever more exclusively asserts the need for judicial power in its quest to eliminate or minimize the administrative state.¹⁴⁸ More disturbingly, liberals treat that same power as desirable or even necessary.¹⁴⁹ Courts today operate within a culture of “juristocracy,”¹⁵⁰ in which judicial authority to “say what the law is” goes largely unquestioned across the political spectrum.¹⁵¹ This juristocratic culture persists notwithstanding how much law application has already been transferred to other institutions. Indeed, the ideological prominence of courts in some ways outstrips their real power, which has been already eroded by increased administration for more than a century.¹⁵²

The progressive goal should be the demystification of courts as default or desirable sites of legal interpretation in order to defang them. In the name of short-term reform, democratizing interpretation could mean democratizing courts. And to the extent that those courts have been allocated authority over law in the first place, it is open to Congress to reallocate it later, if not sooner. For some time, courts will remain a default for many observers, if only culturally and imaginatively. In part for this reason, however, progressives will need to adopt a dual strategy: disempowering courts in the short term without entirely (though perhaps in pockets) reallocating their functions and powers elsewhere, while also contemplating wholesale reallocation of those functions and powers to other institutions in the long term.

If this is true, the debate is really about how to control current bearers of adjudicative and interpretive authority, and when and how to allocate and reallocate their functions to their best imaginable bearers. Short-term, partial disempowerment of current bearers can and will coexist for a long time with long-term reallocation.

Some might suggest that there is no need to disempower courts and reallocate (quickly or slowly) their residual functions to openly political institutions; it would suffice to make explicit the current political functions courts perform.¹⁵³ This is a mistake. As we have seen above, myths of apolitical neutrality permeate the history and theory of courts. By design and by necessity, these legalistic fictions—that judges find rather than make the law—operate to mask courts’ legitimacy deficit in making policy choices. Such fictions limit, for this reason, an alternative strategy: making courts’ political functions explicit. By employing that alternative strategy, one asks why they have been assigned those functions to begin with.

At stake, ultimately, is a legal culture that dispenses with sharp institutional lines between law-application and lawmaking. Any beneficent features of courts, such as the utility of private rights of action to encourage collective monitoring of governance, do not depend on Article III institutions.¹⁵⁴ And if, as part of the culture of juristocracy, government agencies actually rely on courts when it comes to enforcement—especially of penalties involving “private rights”—that culture is contingent, rather than inevitable, and therefore subject to change.¹⁵⁵

We organize this Part by examining strategies to constrain courts and strategies for what we call “reallocation.” Both are forms of institutional disempowerment. Constraint is partial disempowerment; reallocation is total disempowerment. We note in advance that the first could as easily delay as incite the second. For this reason, part of our discussion of disempowerment strategies is oriented to the first-order question of what each reform would accomplish; but it must also attend to how each short-term approach might entrench adjudication in courts and therefore hamper long-term reallocation. Every reform strategy, after all, also provides a potentially safer refuge for residual power.

B. Reforming Courts

1. Personnel Versus Disempowerment Reforms

The more popular approach to reforming courts among liberals today is what we have referred to elsewhere as “personnel” reform.¹⁵⁶ It takes for granted that courts wield their current power and seeks to change their ideological makeup. The continuing hegemonic fiction—that legal interpretation is apolitical—thus coexists with a frank and openly political recognition that everything depends on who serves in the judiciary. The moral and political outlook of those tasked with applying the law bears massively on what the law means and says, in effect.¹⁵⁷ No wonder, then, that Democrats have worked uncommonly hard to restaff the judiciary after the renowned focus and success of the first Trump administration’s campaign to remake the judiciary in its image.¹⁵⁸ (Over two hundred new federal judges were appointed between 2016–21, and in the three and a half years after, over two hundred nominated by Biden have been confirmed.)¹⁵⁹

Choosing personnel in successive waves to remake the federal judiciary will remain an appealing strategy of legal innovation (or stasis) so long as judicial power remains unchecked in other ways. Disempowerment remedies that attempt to alter the judiciary’s authority over law, rather than haggle over its staffing, have attracted less interest thus far, requiring a more wholesale rethinking of our institutional arrangements.¹⁶⁰ Our basic project in this Part is to explore reversing this priority. Our previous work spotlighted the limitations of personnel reform in constitutional adjudication, in favor of the promise of disempowering reform.¹⁶¹ Many of the same arguments apply to adjudication generally, though the details of the disempowerment remedies may differ.

2. Doctrinal Versus Institutional Disempowerment

There are a host of tools for constraining the interpretive power of courts. Most prior analysis concerns doctrinal tools. Doctrinal tools are judge-made self-restraint mechanisms supposed to ensure deference to other, more democratically accountable actors. Most famously, Professor James Bradley Thayer attempted to elevate the judicial deference owed to jury fact determinations into constitutional law, so as to doctrinally limit judicial interference with legislative output on constitutional grounds.¹⁶² But there are many other examples.¹⁶³ Take, for instance, old case-or-controversy¹⁶⁴ and new standing rules, one rationale for which was and is to limit judicial involvement in the nation’s political affairs.¹⁶⁵ Or stare decisis in statutory decisions, which has been defended, and its strengthening proposed, in hopes of containing regular judicial rewriting of interpreted laws, sometimes by affording Congress a chance to react to initial judicial interpretation.¹⁶⁶

Compared to such doctrinal tools, institutional tools are very different. We privilege them in what follows, on the grounds that doctrinal tools—while far from meaningless—are likely to constrain judges far less, especially over time. The basic reason for this is straightforward: Doctrinal tools work through self-constraint; institutional tools do not.

Self-restraint is possible; but it is hard, and it can break down, especially when countervailing imperatives—such as ideology—override.¹⁶⁷ In constitutional adjudication, the Court adopted a principle of judicial restraint following its conflict with President Franklin Delano Roosevelt, declaring in the landmark 1937 case of West Coast Hotel Co. v. Parrish that judges should leave in place any “regulation which is reasonable in relation to its subject and is adopted in the interests of the community.”¹⁶⁸ Within less than a decade, though, and then especially under Chief Justice Earl Warren, the newly liberal Court mostly abandoned that principle in an effort to promote civil liberties and related ends.¹⁶⁹ Conversely, a conservative Court throughout the 1980s and ’90s embraced Chevron deference in statutory cases, willing to defer to interpretations by agencies that tended toward deregulation.¹⁷⁰ As Republican control of the White House became less certain, though, and particularly as conservatives gained a lock on the judiciary in the late 2010s, deference to agencies was replaced by assertions of “judicial power” to “resolve [statutory] ambiguities over time.”¹⁷¹

We have previously argued for a shift from doctrinal to institutional tools in the special case of constitutional adjudication; but run-of-the-mill statutory adjudication is an equally if not more important setting for such experiments. In turn, partial institutional disempowerment can be a step towards wholesale institutional transfer: from the reform of an existing adjudicative institution to the choice of a new, more democratic one.

3. Voting Rules

Supermajority thresholds for decisions have already attracted attention in constitutional adjudication; they warrant (and have received some) consideration in statutory adjudication as well.

Consider one prominent example. Chevron, before it was overruled, called for courts to defer to an agency’s interpretation of law, unless the law clearly precluded the agency’s reading.¹⁷² Though failed as a doctrinal solution, this sort of deference could be re-implemented in at least two different ways: by telling judges to defer to “reasonable” interpretations¹⁷³ or, more mechanically, by requiring them to reach a supermajority voting threshold.¹⁷⁴ (This statutory constraint parallels proposals to institute a supermajority rule to achieve Thayer’s self-restraint regime through a mechanical institutional device.)¹⁷⁵ Professors Jacob Gersen and Adrian Vermeule advocate for a supermajority voting threshold on the grounds that it is likely to achieve the same benefits as Chevron, with fewer costs (including costs incurred in the often merely rhetorical deference characteristic of doctrinal regimes).¹⁷⁶

A supermajority rule for statutory cases would work differently in different situations. When an agency interprets a statute in the context of an enforcement proceeding, for example, that interpretation binds the parties involved unless and until some court intervenes. A supermajority rule in that situation would work to promote deference, partially insulating the agency’s interpretation from later judicial displacement. By contrast, when the Government offers its interpretation in a criminal proceeding, to take one example, that interpretation lacks legal force unless and until some court (and, potentially, some jury) accepts it.¹⁷⁷ In that setting, a supermajority rule operates as a substitute, not for Chevron, but for the rule of lenity, insulating defendant parties from legal liability in the absence of a “clear” violation.¹⁷⁸

Unlike supermajority rules for constitutional adjudication,¹⁷⁹ analogous rules for statutory decisions have mostly been limited to theory rather than practice.¹⁸⁰ There are good reasons to think, though, that a supermajority decision threshold for statutory interpretation generally would be a good thing on balance. In administrative law cases involving a prior, legally binding interpretation, a higher threshold would function to leave agency decisions in place more of the time, promoting democratic responsiveness, as well as technical competence, in turn. And in cases with no prior binding interpretation, the higher threshold would protect, albeit imperfectly, against the retroactive imposition of legal penalties.

More generally, the basic goal in statutory interpretation for judicial decisionmakers (to the extent their role is retained) ought to be to facilitate the specification and development of the law within permissible bounds. Theories vary as to how to draw that line, but there is common agreement around the goal of constraining interpretation within some acceptable and inevitable zone of variation.¹⁸¹ A supermajority rule seems tailor-made for this purpose. After all, interpretations that have succeeded not by a bare majority but by more general consensus are far less likely to exceed the permissible zone, however it is defined.

Consider the examples of hot button cases lately, starting with an example of overturning agency action (in particular, an agency’s understanding of its own powers under statute). In West Virginia v. EPA, the Supreme Court found that certain environmental regulations exceeded the agency’s rulemaking authority.¹⁸² In a 6–3 ruling, the Court appealed to the emerging “major questions doctrine” to conclude that putatively vast agency assumptions of power under the color of a statutory grant of authority could not be assumed to have been delegated.¹⁸³ The regulations were therefore void. In reaching this conclusion, the Court extended a line of resistance to Chevron deference that imposed limits on just how important and new agency rules were, effectively requiring more Congressional guidance. “[I]n certain extraordinary cases,” Justice Roberts wrote for the court, “something more than a merely plausible textual basis for agency action is necessary.”¹⁸⁴

West Virginia v. EPA illustrates both the promise and limits of supermajority rules as a limit on judicial authority in statutory cases. Applying a traditional two-thirds rule, the outcome would have been the same despite the conservative majority relying on what Justice Elena Kagan called a “made-up” interpretive doctrine.¹⁸⁵ Under a rule requiring seven (or eight, or nine) votes, however, the Court would have been unable to reach a decision, leaving the EPA’s regulatory authority in place.¹⁸⁶ As this case illustrates, then, a great deal would potentially turn on which proxy for “law” Congress were to adopt. A relatively non-demanding proxy (like a two-thirds rule) could be overcome by a significant ideological bloc, as West Virginia v. EPA shows. A more demanding proxy (ranging from a 7–2 rule to a unanimity requirement) would, by contrast, be more difficult to overcome but would also risk enabling holdouts in what otherwise seem like “clear” legal cases.

Now consider an example of the second kind of case, without a prior, binding interpretation of statute in the political branches. In Ohio v. American Express Co. (Amex), a 5–4 Supreme Court held that “anti-steering” provisions in contracts between Amex and merchants were not “contract[s] . . . in restraint of trade”¹⁸⁷ for purposes of Section 1 of the Sherman Act.¹⁸⁸ The DOJ, along with several states, had argued those provisions, which prohibit merchants from encouraging (“steering”) customers to use competitor cards, resulted in higher merchant fees.¹⁸⁹ A conservative majority rejected that framing, explaining the relevant “market” included not just merchants (as the DOJ had suggested) but also cardholders, and that the Government needed to consider the effects of anti-steering provisions on the latter (which it had not).¹⁹⁰

A supermajority rule would have made no difference to the outcome of Amex.¹⁹¹ Under the current statutory scheme, the DOJ can only establish Section 1 liability in the context of a judicial proceeding,¹⁹² which means the Government would have needed two additional votes (assuming a 6–3 rule) to prevail had a supermajority rule been in place.¹⁹³ Suppose, however, Congress were to amend the Sherman Act to let the DOJ (or some other agency, like the Federal Trade Commission (FTC)), enforce Section 1 directly and allow courts to consider such cases only on appeal. Under that scheme, a supermajority rule would change the outcome in an analogous case, with the conservatives needing to garner an additional vote (or two, or three, or four, depending on the rule) to displace the agency’s interpretation. As Amex illustrates, then, adopting a supermajority rule requires consideration not only of voting threshold, but also of whether different statutory schemes would, when paired with such a rule, yield more deference (analogizing to Chevron) or less enforcement (analogizing to the rule of lenity).

Amex highlights, in turn, that, for progressives, adopting a supermajority rule alone might do more harm than good in some situations. Federal courts have proven themselves hostile to antitrust enforcement, and requiring DOJ or FTC officials to secure six (or seven, or eight, or nine) votes would result in even less enforcement than at present.¹⁹⁴ The same is true of various pieces of civil rights legislation, including, for instance, Section 1983.¹⁹⁵ To the extent that more aggressive enforcement of such laws promotes democracy, adopting a supermajority rule while continuing to rely upon courts for initial enforcement would involve a tradeoff, however modest.¹⁹⁶ Rather than presenting progressives with some sort of dilemma—namely, whether to promote democracy via undemocratic means—such examples merely remind them that the benefits of certain democratizing reforms sometimes depend upon the adoption of other reforms. Thus, while progressives may prefer the combination of judicial enforcement and a supermajority requirement (up to and including a unanimity requirement) in some contexts, like immigration, contexts like antitrust or civil rights might be ones in which the (re)assignment of enforcement authority is even more important than attending to the applicable voting rules.

It is important to note that a supermajority rule could have significant effects even beyond such closely divided decisions. Consider Glacier Northwest, Inc. v. Teamsters, which involved a dispute about whether the NLRA preempts tort suits under state law by employers against unions for property destruction allegedly caused by strikes.¹⁹⁷ By a vote of 8–1, the Court held that the NLRA might not preempt such suits, interpreting the statute and the judiciary’s own earlier precedents about its meaning.¹⁹⁸ However, the Court held that the possibility of proceeding with the tort suits was itself conditional, since it was made to turn on legal findings that (1) the property destruction was intentional, imminent, and foreseeable, and (2) the unions took reasonable precautions to avoid it.¹⁹⁹

Commentators immediately surmised that some liberal justices helped construct the five person majority opinion to avoid a decision even more prejudicial to labor rights.²⁰⁰ It was self-evident that a majority decision rule shaped the bargaining positions and ultimate choices of the justices. Equally self-evident, a heightened voting standard would have altered the liberals’ bargaining position, potentially leading the five who signed the controlling opinion to strengthen the preemption regime, or to a nondecision.

The costs of a generalized supermajority rule regime in statutory interpretation are partly apparent and partly real. In a system organized like the American one, a supermajority consensus would be unavailable in a number of cases. To the extent that courts are supposed to perform a “dispute-resolution” function, this sort of judicial indecision might seem disturbing. As with any deference rule, though, limiting one decisionmaker works merely to empower other—in this case, more democratic—decisionmakers. More generally, a court’s refusal to decide a case due to the absence of “clearly” identifiable law (evinced here by the absence of a supermajority) is, as Easterbrook observed, just the same as (and, for constitutional purposes, an instance of²⁰¹) dismissal on jurisdictional grounds, which courts do all the time.²⁰²

A related worry has to do with disuniformity. Should the Supreme Court fail to reach the necessary supermajority to decide some statutory case, it could leave in place regional variation in interpretation in the form of federal circuit splits or varying state court decisions. Setting aside that such variation is the default in a system of discretionary Supreme Court review, as Jed Shugerman has shown in the constitutional setting, this problem can be solved through implementing rules that condition lower court relief upon a higher court’s failure to weigh in, even if weighing in takes the form of nondecision.²⁰³ Building on the idea that nondecision is, in effect, dismissal on jurisdictional grounds, a rule concerning relief provided by lower courts would make such a jurisdictional determination apply to courts generally rather than the Supreme Court or some other appellate court, specifically.

This is to say nothing of the traditional argument that deference to federal agencies promotes geographic uniformity in the law by prioritizing a single national agency over regional courts of appeals. As Justice Gorsuch observed in his concurring opinion in Loper Bright (the case that overruled Chevron), such an approach does conduce to temporal disuniformity insofar as agencies are more disposed than courts to change their minds about what statutes mean, owing in part to agencies’ quicker ideological turnover.²⁰⁴ We think that sort of disuniformity is a virtue, though, since it merely reflects the ideal that elections have consequences.²⁰⁵

The effects of a supermajority rule in interpretation for statutes are also likely to have beneficial political effects beyond constraining courts’ interpretive ambit. Under the supermajority regime, the politics of taking over judiciaries for the sake of controlling legal interpretation correspondingly becomes less appealing, precisely because it is harder to achieve. If a bare majority rule has allowed a titanic contest for control of the Supreme Court to be a central feature of America’s national politics for a half century,²⁰⁶ a supermajority rule would marginalize the court and force ideological strife to more openly political settings. Indeed, one objection to bare majority control over powerful institutions is that they incentivize bets on control that hinge on accidents like death and retirement, rather than contests resolved by at least potentially more deliberate events such as elections. A supermajority rule would have the immediate effect of reducing the impact of a random event.

Some progressives may object, however, that bare majorities in statutory interpretation can promote egalitarian ends as well. For example, the reinterpretation of the Title VII of the Civil Rights Act of 1964 in Bostock v. Clayton County to bar discrimination based on gender identity and sexual orientation won six of nine votes.²⁰⁷ In that case, the LGBTQ plaintiffs sought a judicial order²⁰⁸ and so, if there were a stringent enough supermajority rule, the case would have led to a different outcome based on bargaining with otherwise dissenting justices to achieve consensus (or to avoid nondecision and consequent dismissal of the case). Here again we are reminded of the interaction between voting rules and the assignment of enforcement authority in a statutory scheme.

Of course, it is true that supermajority rules could have similar costs; it is harder for all political agendas (not just that of the currently prevailing reactionary judiciary) to be achieved through litigation. But overall, their benefits are greater, and, in any event, such “good” results do not depend on the judiciary. If federal and even state courts were disempowered, that would merely require progressive legislation, or progressive reinterpretation of statute by other entities, such as agencies, rather than coaxing favorable interpretations from friends in democratically insulated high places. Defenders of courts as indispensable sites of progressive politics must, then, explain why the results they seek are not overridden by the aggregate output of adjudication in courts, and why those results are not available any other way. Our point is not that cases like Bostock failed to deliver good things. It is that the virtues of majority rule in judicial decisionmaking come along with massive vices. Silver linings in a storm are not much to celebrate.

This is not to say that there are no difficulties in using supermajority rules to disempower courts in statutory cases. One basic problem is determining the proper threshold, especially in our current situation in which conservatives already control a two-thirds supermajority. To the extent supermajority agreement is understood as a proxy for law, present conditions make the over-inclusiveness of a two-thirds rule obvious. And yet, a more demanding rule, up to and including a unanimity requirement, would allow lawless and otherwise incompetent judges to act as holdouts, risking under-inclusiveness in turn.

The most important drawback to this tool for statutory control is that it does not improve the decision’s transparency. No more than majority rules can supermajority rules do anything to make plain when decisionmakers are effectuating law and when they are making political choices. As described above, the prevalence of policy choice masquerading as legal interpretation is the general reason legislators should disempower courts and shift decisionmaking to popularly accountable and permissibly ideological actors. But supermajority rules neither work to identify which cases involve plausible legal constraint nor necessarily work to improve the quality of the policy choices that remain. In this respect, supermajority rules are inferior to their doctrinal analogues, which, notwithstanding their unreliability, have the virtue of requiring judges to acknowledge the absence of clearly identifiable law.

At best, supermajority rules can be envisioned as proxies to raise confidence that legal constraint is operating on the grounds that more agreement of ideologically diverse judges might be taken to force them to bracket political agendas. But a proxy is merely that, and there are many ways to account for why supermajority thresholds are attained by shifting coalitions of judges. A likely effect of supermajority rules, then, is merely to alter the terms of bargaining. Similarly, the attainment of the supermajority threshold does not indicate that the decisionmakers are making a better policy choice, by whatever standard; it just implies more consensus about the propriety of that choice than in closely divided circumstances. Like some of the best policy choices that judges have made in the Supreme Court’s history, some of the worst have been made with supermajority agreement, including Dred Scott v. Sandford ²⁰⁹ or the Civil Rights Cases.²¹⁰ (Though, as discussed above, supermajority and even unanimous decisions sometimes emerge as a result of bargaining, the terms of which might vary under a supermajority rule.)

As the attainment of the higher threshold in those and other cases indicates, the most basic limitation with supermajority rules is that it raises the question posed by any disempowerment remedy that stops short of fully transferring legal interpretation to politically accountable actors: Why not go further? To the extent judges are left any power to engage in policy choice with the aura of legality, especially in a world in which so much interpretation has been transferred from courts, reform will still invite the question of why it is not more interventionist.

Other limits that have been claimed for supermajority rules are less substantial. One general critique of them, in constitutional and statutory cases alike, is that they can be construed as violations of the rule of United States v. Klein, which forbade Congressional interference with the outcome of cases.²¹¹ But if Klein doesn’t forbid a supermajority threshold for statutory invalidation, it shouldn’t for statutory interpretation.²¹² Beyond this legal objection, there are various technical objections to instituting supermajority rules, especially in a federal legal system like America’s. Since district court decisions are made by one judge, such a rule could not apply at that level (but then, a majority rule does not either). In effect, three-judge panels in the courts of appeals already work by supermajority rule, since a majority of three is effectively a two-thirds threshold. One could, however, go further, requiring unanimity from such panels. Regardless, both the Courts of Appeals meeting en banc and the Supreme Court allow bare majorities to control the law. A supermajority rule there would have the disempowering effect outlined above, even as a two-thirds rule.

4. Shunting Mechanisms

Another mechanism shunts legal interpretation outside the courts when specific triggering conditions are met. This tool goes beyond requiring higher levels of consensus within the courts by shifting the authority to interpret the law elsewhere. The reasons to entertain this mechanism are broader than simply because courts make closely divided decisions, either in agency review or first impression statutory interpretation. After all, judicial lawmaking doesn’t take place only when judges disagree whether and how to do so.

Many interpretive decisions are close because the legal materials present a parting of ways: mutually exclusive options for how to resolve ambiguity or fill gaps. Some of these decisions track existing partisan splits and political ideology, but many do not. Sometimes the difference among judicial policymakers occurs not within courts but between and among them. Functionally, supermajority voting rules distinguish interpretations based on bare majorities—including potentially major interpretive departures—from those that enjoy some higher level of consensus. In the American system, however, there are circuit splits between and among the federal courts that occur because a policy choice must be made, one way or the other.

Such splits themselves are evidence that there is no legally compelled answer—these splits fit poorly with the mythology that one branch of government engages in “will” and the other in “reason” or “judgment.”²¹³ However consensual the judicial outcome, such policymaking episodes demand politically accountable actors to make the call—even if it is uncontroversial. In this case, a supermajority rule designed to allow decisions where there is a high level of consensus but forbid decisions with low or none may not help much. And, of course, even in those latter cases the heightened threat of nondecision from a supermajority rule may persuade some observers to seek other tools.

One such tool is a mechanism that would force judicial uncertainty out of the judicial branch. This shunting mechanism would serve as a reverse advisory jurisdiction mechanism. Under traditional advisory jurisdiction, legislatures ask courts for non-binding interpretation assistance; what if there were a mechanism for courts to get help in interpretation when either they could not agree (under majority or supermajority rules of decision) or they confront a policy dilemma with only a policy resolution?

Consider a rule that shunted disagreements among regional circuits in a federal system, and/or cases with close enough disagreement, out of courts to legislatures—if not for adjudicative resolution and law application then at least for legal interpretation. This rule could also apply to judicial review of agency interpretation, providing a second institutional mechanism (beyond the voting rule fix prior scholars have suggested²¹⁴) for promoting deference to agencies. When circuits disagree about whether agency interpretation is valid, a shunting rule would simply hold that the agency automatically wins by shunting the final legal interpretation back to the agency from the courts.

As an example of how such a mechanism would function, consider a current circuit split involving the Federal Arbitration Act.²¹⁵ That law mandates enforcement of arbitration agreements, unless the dispute involves workers engaged in interstate commerce.²¹⁶ But courts have differed about who may claim this exemption. In Carmona v. Domino’s Pizza, LLC, the U.S. Court of Appeals for the Ninth Circuit held that drivers who deliver pizzas were covered by the exemption, even though the entirety of their work took place locally, on the grounds that delivery of a pizza is the last step in the sale of a product made with ingredients originating across the country.²¹⁷ But in Lopez. v. Cintas Corp., the Fifth Circuit held the opposite, forcing a uniform delivery driver into binding arbitration.²¹⁸

No legal considerations arguably compelled either result. Before reaching its general exemption for workers engaged in interstate commerce, the statute specifically exempts “seamen” and “railroad employees.”²¹⁹ But the statute gives no answer about who is covered by its closing general exemption—even if one appeals to a canon of statutory interpretation like ejusdem generis that restricts the meaning of a general item on a list in light of specific other items on it.²²⁰ After all, seamen could spend their whole careers in the same port and railroad workers fixing the same stretch of intrastate track. Resolving this circuit split, and so many others like it, is just a policy choice. Comparably, when the Supreme Court resolved last term in Bissonnette v. LePage Bakeries Park Street, LLC the related question of whether the exemption was reserved for transportation workers, a possibility its own precedent had raised,²²¹ it concluded in the negative. If workers were engaged in interstate commerce, the Supreme Court decided, it didn’t matter if their company wasn’t a transportation business. Once again, the underdeterminacy of the statute let the Court come out either way—even though it chose its result 9–0.²²²

What if circuit splits required Congress to resolve the interpretive dispute before courts applied the law to facts? If one’s response is that it is crucial for outlying circuits to be brought into conformity with national consensus, then the shunting mechanism could be triggered conditionally on a single circuit split, and fully only after a sufficient number of circuits reached opposite conclusions about the meaning of a statutory provision. There would be some cost in This eventuality in delay would cost litigants with pending Supreme Court appeals blocked as the scope of national disagreement clarified. But this kind of cost—however great—has to weigh the costs of judicially resolving policy choices, which can hurt a lot of people.

Another virtue of shunting mechanisms is even more compelling. They respond creatively to the risk of judicial non-decision that supermajority rules involve, to some greater or lesser extent. The shunting mechanism, after all, could be adopted in cases in which a court like the Supreme Court decided statutory cases by bare majorities, or even by higher consensus levels. And far from incurring the risk of non-decisions, the shunting mechanism could require political decisionmakers to act or, at least, take responsibility for the risk of interpretive disagreement themselves.

One response to shunting mechanisms could be that they obviate a principal institutional virtue of judicial interpretation: that political actors may not be able to reach consensus on policy themselves or want to escape responsibility for doing so by shifting interpretive authority to other agents. But it is an open question whether this “virtue” is in fact a vice: a shunting mechanism would require policy to be made by accountable policymakers, even if they did not want the responsibility to do so or found it hard to shoulder.

5. Facilitation of Legislative Override

Then there is legislative override of judicial interpretation of statute. There are emergent proposals to facilitate this in new circumstances. Our contribution is to portray this mechanism in the larger universe of disempowerment remedies, and to place it on a continuum toward the goal of reallocating adjudicative functions.

Compared to the previous two disempowerment mechanisms, legislative override has been a continuing practice and is therefore reasonably well-known.²²³ Examples can be traced at least as far back as the early twentieth century, when court-curbing agendas began to spike in the face of a conservative judiciary.²²⁴ They have been documented for recent American history,²²⁵ too, and scholars have explored how big an effect legislative override has on subsequent judicial behavior.²²⁶ Formally, it is already clear that Congress has the last word in statutory interpretation. But to the extent that Congress needs encouragement to use it, the existing literature on override should lead to creative new toolmaking.

Many of the emergent proposals urge congressional reengagement with lawmaking by extrapolating from an existing statute. In 1996, Congress passed the Congressional Review Act (CRA)²²⁷ in response to the Supreme Court invalidation²²⁸ of the legislative veto in INS v. Chadha.²²⁹ The CRA facilitated congressional “lookback” at agency regulations, and override when appropriate, which meant it has generally been used when the presidency switches parties, and the new administration can work with a friendly Congress to undo the rulemaking of a prior administration.²³⁰ Overriding agency regulation concerned legislative reclamation of power from the executive branch, not from the judiciary. But even today the CRA’s provision²³¹ for fast-tracking congressional consideration of another branch’s output treatment remains a model for court disempowerment. In both cases, there is disempowerment and functional transfer of policymaking back to the legislature.²³²

The basic idea of facilitating override in response to court decisions is that institutional triggers could encourage or mandate that some judicial acts of statutory interpretation undergo congressional review and, where desirable, replacement. Professor Ganesh Sitaraman called for such a mechanism in 2019²³³ and Senator Sheldon Whitehouse formally proposed a version of it in 2022.²³⁴ Whitehouse’s bill envisioned extending the existing CRA, requiring the Comptroller General to trigger congressional engagement with statutory interpretation, and to do so within short order of certain kinds of Supreme Court rulings.²³⁵ One can likewise imagine extensions of the Act that would prompt Congress to restore judicial deference to agency interpretation of laws, such as the one Professor Christopher Walker has proposed to counteract the major questions doctrine.²³⁶ Were this approach to gain traction, the override trigger would go far beyond policing agencies, the CRA’s original purpose, in order to police judicial interference with agencies.

As an example, West Virginia v. EPA in part provoked Senator Whitehouse’s bill .²³⁷ If his bill had been good law, the Comptroller General would have responded to the decision within two days, and otherwise applicable procedural roadblocks with committee and full chamber amendment of the law by majority vote could follow. Senator Whitehouse was at pains to emphasize that overriding is not “overrul[ing],” since it is not a judicial act, let alone a decision in a case or controversy.²³⁸ Rather, it is amendment to existing statutory law based on prior judicial interpretation of it, forcing the judiciary to steer a different course on the basis of a new law.²³⁹

As with supermajority rules, there are obvious limits to the facilitation of legislative override. The distinctive limitation of this tool is that it presupposes ongoing legislative control and enough will to supersede juristocracy. The endless number of statutes over which judiciaries have seized, and been allowed to keep, control contrasts starkly with these presuppositions. And nothing guarantees that legislatures act to override decisions that were political, as opposed to faithful interpretations of the legal constraint that political decisionmaking originally imposed. Finally, given the difficulty of legislative override practices where allowed, no amount of facilitation would abolish the default of the last word for the judiciary.

Legislative override does, however, make clear that institutional disempowerment can take escalating form. Each more intrusive tool can provide stronger functional versions of earlier ones. For example, the rules that potentially trigger legislative override of judicial statutory interpretation could privilege statute reconsideration that bare majorities interpreted or circuits split on. The same is going to be true as we turn to full-scale reallocation. After all, all disempowerment—if it is institutional rather than doctrinal—is partial reallocation. It hollows out zones of policymaking that courts have enjoyed. Escalation therefore could ultimately lead to their total closure and replacement within explicitly policymaking institutions.

C. Replacing Courts

In Federalist No. 78, Alexander Hamilton famously distinguished between the “will” of political actors and the “judgment” of allegedly nonpolitical ones like judges²⁴⁰: “The courts must declare the sense of the law; and if they should be disposed to exercise will instead of judgment, the consequence would equally be the substitution of their pleasure to that of the legislative body.”²⁴¹ It is often forgotten, however, that Hamilton added the radical suggestion that if the distinction between will and judgment turned out to be untenable, an interesting implication would follow: “The observation, if it prove any thing, would prove that there ought to be no judges distinct from that [legislative] body.”²⁴² Exactly! Well said.

Interpretation both in open rulemaking and in adjudication have been allocated to political agents from the beginning. Article III courts, to the extent they were allocated or seized power, are still open not just to disempowerment, but replacement. At the end of the escalating process of disempowerment—to which there is no natural stopping point except what feasibility allows and prudence requires—is complete allocation or reallocation of interpretive (including adjudicative) functions outside courts. There is, in fact, a deeper record of existentially disempowering courts than of stopping short of doing so through piecemeal reforms.²⁴³ Additionally, reallocation of extant power away from judiciaries has been dwarfed by the creation of new interpretive powers ex nihilo in the political branches, which skips the step of allocating them to the judiciary first.

This Section begins by reframing how, consistent with Article III, that allocation could have gotten so far without scholarship or the Supreme Court registering it. However shocking the allocation of legal interpretation to political control may seem, that is more a matter of ideological fixation than legitimate expectations: belief in judicial control even when it has already been undermined to a great extent. The Section then goes on to recapitulate just how much wholesale allocation or reallocation of power has occurred throughout United States history with the aim of ensuring statutory control. Against this backdrop, attention can now focus on possible extensions and expansions to allocating legal interpretation fully into the political domain. Under current doctrine, constitutional limits might forbid the complete realization of a system of openly political interpretation of statute—Hamilton’s nightmare but a credible progressive dream. But it ought to be uncontroversial that, otherwise, Congress can and should pursue a full reallocation project.

1. The Mystique of Article III

The conventional view is that Article III allocates all judicial power enjoyed by the federal government to the federal courts.²⁴⁴ Of course, functionally, no constitution could ever give any element of government a monopoly over legal interpretation; even formalists recognize exceptions to the institutional allocation of adjudication to Article III courts in particular.²⁴⁵ There are vast swathes of adjudication in both the executive and legislative branches of government.²⁴⁶ The Supreme Court has granted those “exceptions” broadly, and has even avoided declaring whether the federal judiciary must remain the ultimate resort for other constitutional claims (other than whatever constitutional norms underpin the endurance of Article III courts).²⁴⁷

Both conservatives and liberal scholars have responded with what one might call Article III theory-crafting. On the assumption that there must be some authority by which Congress may not deprive the federal courts, locating the formal line between permissible and impermissible allocation of interpretive power outside the judiciary becomes an intellectual puzzle to solve. Liberal solutions to the puzzle are typically ethical, theorizing how independent life-tenured judges are uniquely positioned to safeguard liberty. Conservative accounts are typically historical, acknowledging longstanding practices; William Baude, for instance, has offered a creative textualist reading that places Article III courts in a universe of other courts (such as territorial and tribal courts) as well as various kinds of putatively non-adjudicatory action. ²⁴⁸ On this theory, Congress may properly allocate to political decisionmakers cases involving government benefits or “public rights”—because they do not implicate the “private rights” of life, liberty, or property.²⁴⁹

All such solutions are rationalizations of a path-dependent direction Congress has headed in partial allocation or reallocation of interpretive power outside the judiciary. Of course, conservative solutions to the puzzle can take on momentous importance in an era in which critics urge the Supreme Court to transfer authority back to Article III adjudication²⁵⁰—for example, in October Term 2023 case SEC v. Jarkesy, which declared that an agency may not adjudicate securities fraud without a civil jury, appealing again to the concededly “arcane” and “confusing” category of “private,” as opposed to “public,” rights.²⁵¹

Our main point is that this scholarly debate characteristically fritters on the margins of a fait accompli. No matter what happens to the nondelegation doctrine and its surrogates, a better and more honest and realistic debate should begin with how undeniable it is that interpretive authority over law is already understood to be a political matter. On the formalists’ own account, in which agency rulemaking cannot be lawmaking (in spite of its name), but only legal interpretation,²⁵² then it has long been untrue that courts have a monopoly over legal interpretation. More important, most federal adjudication goes on outside Article III.²⁵³ Nor, as Professors Adam Cox and Emma Kaufman have pointed out, does the public rights doctrine account for what kinds of matters have been allocated beyond or transferred outside the ken of Article III courts²⁵⁴: “[T]o be a formalist about the separation of powers, one has to bend over backward to explain why there are so many courts and judges inside the administrative state.”²⁵⁵

But to the extent this is true, rationalizing the judicial retention of power after the fact, whether in conceding the authority of political actors or longing nostalgia for Article III adjudication, is the wrong move. Both to capture the state of play and to envision the protection of crucial values consistently with a vision of law as popular self-rule, foreseeing allocation or reallocation of interpretive authority in political fora is the right one.

2. Political (Re)allocation to Date

Congress has regularly allocated authority over both interpreting law and adjudicating it beyond the federal judiciary. Theoretically, a “presumption of judicial review” is taken as orthodoxy in administrative law,²⁵⁶ and Congress has affirmatively presumed an important enforcement role for Article III suits in effectuating its will. For example, the Fair Labor Standards Act and a host of others rely mostly on private suits against private actors, in which the federal judiciary provides the forum of statutory enforcement, and therefore interpretation.²⁵⁷ Antitrust law requires the government to prove its case for violations in federal court; notoriously, this choice led to major interpretive revolutions as judicial ideology drove doctrine in a deregulatory and libertarian direction.²⁵⁸ But Congress also routinely opts to leave the judiciary out, in the first instance or entirely. The NLRA created a board that played the predominant role in adjudication and interpretation of the statute—though it left a residual place for judicial control of board action,²⁵⁹ which, as Karl Klare’s classic study showed, moderated the radicalism of the law.²⁶⁰ Complete allocation of legal interpretation to political control without the prospect of judicial interloping is common in practice, especially in regimes like immigration in which “jurisdiction stripping” has already occurred.²⁶¹

We can distinguish between two different modes of disempowerment through allocation or reallocation elsewhere. First, Congress has set up alternative venues for statutory control. Second, Congress has protected those venues from court interference, whether by stripping jurisdiction over statute interpretation altogether or obviating any second-guessing of agency interpretation of its provisions. As before, these full-scale allocation mechanisms can be presented as escalating techniques. One tactic is to secure first-impression interpretation of statute within the political branches by creating alternative venues. Another is to protect first-impression interpretation of statute from judicial reinterpretation.

Congress has created various Article I and II venues of statutory interpretation policymaking. These have come in two forms: free-standing specification, through administrative rulemaking, and adjudicative processes located outside the Article III judiciary (or in special courts within it). Article I legislative courts have generally, though not unerringly, interpreted statute through case-by-case adjudication. Article II agencies, by contrast, famously regulate by initial rulemaking processes, which agencies then enforce, often settling disputes through their own adjudicative corps.²⁶²

The assignment of subject matters to Article I and II venues may seem random, as befits the path-dependent emergence of the administrative state; and, as John Ferejohn and Larry Kramer have observed, “[t]here is, in truth, little or no difference between a legislative court and an executive agency with respect to the power to implement and apply law.”²⁶³ The archetypical Article I legislative court has a subject matter focus that requires centralization in one venue or expertise unavailable in generalist courts.²⁶⁴ If Article I interpretative institutions have been sparingly used to date, for whatever reason, examples of Article II agencies are legion.

The earliest examples of legislative courts, after a prehistory in the early republic,²⁶⁵ were the Court of Claims and the Court of Customs and Patent Appeals (later merged in the Article III Court of Appeals for the Federal Circuit).²⁶⁶ The same Mann–Elkins Act that legislatively overrode a prior judicial interpretation of statute also created an exclusive new venue, the short-lived Commerce Court, to consider challenges to Interstate Commerce Commission decisions (over railroads and early telecommunications).²⁶⁷

Congress also has the power to create alternative Article III venues and bar other courts (other than, perhaps, the Supreme Court) from interfering with their work.²⁶⁸ During World War II, Congress accorded exclusive jurisdiction to a new Emergency Court of Appeals considering challenges to the Office of Price Administration (though allowing ultimate Supreme Court review).²⁶⁹ One possible reason for such venues—which the Foreign Intelligence Service Court exemplifies now—is that specialized topical bodies are likely to be more deferential, counteracting the risk in generalist and geographically decentralized Article III courts of excessive cultural or physical distance from the agencies they review. If that is true, the Article III requirements of life tenured judges with salaries immune from reduction still permit experimentation beyond the existing federal court system, and the grievous threat it poses to Congress’s statutory order; and if those models prove unworkable, there are always more openly political courts for Congress to create.²⁷⁰

Under whatever Article—I, II, or III—these alternative venues have been approved by the Supreme Court as constitutionally permissible entities, within unclear limits.²⁷¹ Although alternative interpretive venues under legislative auspices have been used intermittently and sparingly, interpretive venues under executive auspices make up an enormous administrative endeavor. Congressional output is paltry compared to that of the administrative agencies, which make thousands of rules specifying statute annually, and employ thousands of administrative judges, who apply policies to dispose of over a million cases each year.²⁷² Agencies are well-known for minting regulations under framework statutes. And like adjudication everywhere, administrative judges exercise a legislative function in small or large ways, since adjudication so often involves it. The Social Security Administration, “probably the largest adjudication agency in the western world,” decided over six hundred thousand claims in 2013 alone.²⁷³ Immigration judges decided another two hundred thousand, and Medicare and Veterans Administration judge tens of thousands each.²⁷⁴ By contrast, the federal courts deal with no more than half that number of cases in an average year.²⁷⁵

Cox and Kaufman have, therefore, called our system an “adjudicative state.”²⁷⁶ That case-by-case adjudication, alongside statutory interpretation through rulemaking, sometimes occurs under ultimate (much disputed) Article III supervision or under a presumption of reviewability does not change how massively allocation and reallocation outside Article III has already happened by Congress’s will. The extraordinary role of agencies in lawmaking embarrasses any formalist account of the separation of powers. And the sheer volume of adjudication in the “political” branches that itself inevitably has a lawmaking dimension does too. Most legal interpretation is not judicial in our world. It is already openly political.

As recent debates about the historical credentials of a “nondelegation doctrine” have proved, devolution of lawmaking authority to entities other than Article III courts goes almost as far back as the legislature’s own powers.²⁷⁷ But there was a lot of contingency in the belated decision to build agencies on a colossal scale within the executive branch rather than as arms of the legislature, especially given the examples of legislative courts.²⁷⁸ If the law was to be “the public’s law,”²⁷⁹ statutory control and, particularly, containment of interpretive revolutions instigated by actors other than the formal legislative assembly hardly warrants presidential control, whatever claims exist about the executive as an accountable and representative actor in their own right. Of course, this does not mean that Article I and II courts are constitutionally different entities under current doctrine, only that Congress could privilege legislative courts if it wanted to ensure adjudication outside Article III did not fall under executive auspices.

Congress has also legislated to protect interpretive power it constitutes outside the judiciary. Indeed, most acts of allocation or reallocation, creating new interpretive institutions under Article I or II or III, have been explicitly or implicitly constraining on extant Article III interpretation. For example, the World War II price control measure that brought about the Emergency Appeals Court deprived all lower courts of jurisdiction to consider either prices set under the Act or the validity of the Act itself.²⁸⁰ As the then-Chief Justice put it in Yakus v. United States, “it was thought desirable to . . . minimize the confusion that would result from inconsistent decisions of district and circuit courts rendered without the aid of an administrative interpretation.”²⁸¹ In other words, the channeling of jurisdiction outside traditional Article III venues was an expedient to maintain interpretive control.

The Supreme Court has generally acceded to massive allocation and reallocation of interpretive authority over law, especially when some privileged set of so-called private rights is not involved, sometimes requiring appearances to be maintained by the hypothetical ultimate availability of an Article III check.²⁸² Notwithstanding the general sense that the creation of the Administrative Procedure Act (APA) after World War II “settled” earlier disputes by conditioning the exercise of agency powers on judicial checks to promote rule of law imperatives like accountability, non-arbitrariness, and reason-giving, most federal interpretive authority over law is not judicial but political.²⁸³ With consent of the judiciary, most administrative law has already become “internal,” without functional reliance on Article III oversight, despite its supposedly haunting specter;²⁸⁴ in everyday practice, the “accountable bureaucrat” does not depend on judicial supervision.²⁸⁵

Furthermore, Congress has in many situations successfully eliminated even hypothetical judicial checks on political branch interpretation of statute. Bars on judicial review—of agency rules and of agency application of those rules in both enforcement and adjudication—are pervasive in administrative law. “Jurisdiction-stripping,” treated as futile or impossible under constitutional law, is routine and uncontroversial when allocating authority over statutory interpretation.²⁸⁶ Rulemaking tends to have more ultimate checks outside agencies than administrative adjudication does.²⁸⁷ But even interpretation of statutes is sometimes explicitly exempted from external review—most famously when the relevant statute precludes review or commits discretion to agencies as a matter of law.²⁸⁸ According to Professor Laura Dolbow’s data, no fewer than 190 provisions in the U.S. Code bar federal court jurisdiction over agencies in some way or other.²⁸⁹ In recent years the Supreme Court has been especially permissive of such bars in “high volume” adjudicative domains as Medicare and immigration.²⁹⁰

Far from bugs, such phenomena are features of administrative law, expressing how far political branches have allocated interpretive authority to their own institutions and protected that authority from judicial interference. Professor Vermeule has claimed that it is inevitable, no matter the intentions of administrative law framers, that legislators will create zones of either immunity to external checks, reviewability only up to a point, or outright unreviewability of agency rules and actions (and when asked, judges generally defer to such arrangements).²⁹¹ This isn’t, he argues, just something that crops up in ad hoc carveouts—such as the exclusion of the military²⁹²—or for the sake of exceptional political emergencies. “Ordinary” administrative law is riddled with a porous border between legality understood as requiring external checks and legality understood as requiring fewer or none.²⁹³

The implications for our analysis are obvious: Statutory interpretation has already been recognized as inherently political to an extraordinary extent, and doctrinal limits have been eroded to the point where it is possible to envision completing the process, when there is will to do so.

3. Completing Political (Re)allocation and (the Lack of) Constitutional Limits

Congress could create more legislative courts. It could expand agency adjudication. It could, for that matter, sidestep the federal courts within Article III, as in the emergency price controls example. It could bar review of more agency rulemaking than it already has.

“[T]o reject the view that all federal adjudicative tribunals must be article III courts is to state the problem, not solve it,” insisted Professor Richard Fallon some time ago.²⁹⁴ But from the perspective of ongoing allocation of legal authority, the nostalgic attempt to discover or reclaim the necessary and proper scope for judges in statutory interpretation presupposes that there has to be some such role, notwithstanding the extraordinary amount of allocation and reallocation elsewhere that has already occurred.

In fact, allocation and reallocation of political control over law is a fait accompli, and current doctrine allows it to progress even further. Sometimes how far it has gone is treated as a melancholic inevitability, for better or worse: “If judicial review provides no safeguard against potential abuses of power in most regulatory activities,” one commentator writes, “we must turn to other mechanisms.”²⁹⁵ Why not, however, treat it as a desirable outcome? Professor Nicholas Bagley has argued that the accreted default—that presumption that judicial review of agencies is available—ought to be discarded given the amount of historical allocation without it so far, especially since the benefits of retaining the presumption are outweighed by its costs.²⁹⁶ Instead, the opposite default makes more sense. Congress should be presumed to determine from statute to statute when and how—and if—external review is possible, with no automatic presumption favoring judicial scrutiny.²⁹⁷ This sort of analysis generalizes far beyond its context, since what holds for the presumption of reviewability may well hold for judicial review of a statute itself. It could be desirable to go beyond weakening the presumption of reviewability to presuming non-reviewability and the abandonment of courts altogether as sites of interpretation.

Further, Congress could shutter the Article III judiciary almost entirely, according to current doctrine, with the possible exception of leaving it as a venue for constitutional claims. No one is an “Article III literalis[t]”²⁹⁸ any more, given the vast legal interpretation that already occurs outside Article III courts—though some do develop fallback historical theories purporting to justify case classes they believe are necessarily decided by an independent judiciary.²⁹⁹ In truth, the Constitution only imposes Article III review for cases of minor original subject-matter jurisdiction of the Supreme Court (cases involving ambassadors and public ministers or consuls, and cases in which a state is a party).³⁰⁰ Otherwise, Congress’s power to channel jurisdiction is plenary. It can have or not have a lower court system.³⁰¹ Moreover, the Exceptions Clause allows for depriving the Supreme Court of appellate jurisdiction in most domains of law.³⁰²

These are conclusory statements, which preempt an extremely long-winded debate about limits. Ironically, we can abstain from summarizing it here, and stick with the conclusory statements, for two reasons: First, there is the irony that, at a higher level, the law is unclear about how much power the judiciary must constitutionally retain to deal with unclarity at a lower level. The very reasons that statutory interpretation is policy, to be openly treated as such, suggest there is no apolitical legal answer to the boundaries of and limits to policymaking, which are themselves policy dilemmas. Second, this area of constitutional law is famously overtheorized and underdecided.³⁰³ The very fact of academic controversy amidst judicial silence suggests there is no answer forthcoming about the limits of Congress’s jurisdictional control, any more than centuries of arcane theology could make up for God’s silence.

If there are hardwired limits to allocation and reallocation projects, they are likelier to surface when constitutional norms (other than the Article III limitations discussed above) are said to be at stake.³⁰⁴ Often, this is because Congress avoids negating constitutional claims even when engaged in transfer. Despite stripping jurisdiction from all other courts, the Emergency Price Controls Act, for example, did not purport to strip the Supreme Court of entertaining constitutional challenges to the Act or to regulation pursuant to it.³⁰⁵ Even in other situations, the proviso that current doctrine might restrict Congress is, of course, serious. How serious it is, however, is unclear, in part because the Supreme Court has always avoided the confrontation that reserving its authority would involve.

Congressional stripping of authority to entertain constitutional claims may itself be constitutional, or at least not clearly unconstitutional. Regardless, what is left is the widest possible scope for democratizing statutory interpretation. Most notably, when they have spoken on the matter at all, Supreme Court justices have always avoided defending residual limits to their own disempowerment even while ceding massive territory to prior transfer projects.³⁰⁶ SEC v. Jarkesy does not deprive the federal judiciary of agency action review; Jarkesy itself came about by invoking the Seventh Amendment jury trial guarantee.³⁰⁷ Hypothetically, were Congress to respond to the Court’s decision by stripping constitutional challenges to the agency, there is little reason to believe that the Seventh Amendment would require some Article III venue,³⁰⁸ especially if the one Congress chose operated with procedural constraints.

In any event, the conclusion that Congress’s authority to allocate statutory control is essentially unlimited, absent constitutional claims, is already radical enough for present purposes. No one can say how easily or quickly full allocation of interpretive authority to political actors could or will be finished. It is a political project. And we express no view on its most desirable forms, though there is an enormous debate to have over how to organize the politics of legal interpretation and specification, especially once one has given up the formalist picture of legislatures as exclusive sites for lawmaking, and judiciaries as exclusive sites for interpretation. For a long time, no doubt, the fig leaf of a hypothetical purpose for judges may be maintained, as in monarchies where lip-service is paid to men and women with crowns who no longer exercise power. But the drift of history is clear, and allocation beyond judiciaries has occurred for good reason; the imperative now is to imagine its furtherance.

Towards Democratic Law

The argument until now has been that, in statutory and constitutional law, we do best to allocate what is ultimately political work away from courts and toward other more democratically responsive, and more openly political actors. It would be, apparently, infeasible in many (though not all) cases to send that work back to the legislature, so much of the suggestion has been that we should insulate interpretations by agencies from judicial displacement.

This Part takes up three obvious and understandable objections to this approach to democratizing of statutory interpretation. One is that doing so does away with the rule of law; another is that the great value of the fiction of apolitical judges with insulated power is their exercise of it for the vulnerable and weak, over whom government would otherwise trample; the last is that our official politics is hardly democratic, leaving the unofficial politics of judges no more inferior in this respect (and, perhaps, superior).

At the outset, it is worth making two observations. First, these objections have not forestalled the fait accompli of disempowerment and (re)allocation of interpretive authority over law to political actors—indeed, on a massive scale.³⁰⁹ Alone or together, the objections either purport to explain why to resist the final stages of disempowerment and (re)allocation, not whether to begin them in the first place. Second, both prior and current struggles to protect this allocation from judicial clawback prove that there is a great deal at stake—not that ensuring the rule of law, protecting the vulnerable, or enhancing democratic self-government within openly political settings simply requires Article III institutions to migrate to a different institutional location.

A. The Rule of Law

What about the rule of law? Even for those sympathetic to the outright rejection of constitutional review of ordinary legislation, the prospect of reducing or even eliminating judicial oversight of the executive can be cause for unease or even alarm.

Professor Jeremy Waldron, for example, contends famously that judicial review of legislative outputs is undemocratic, and also appears to concede that limiting review of executive action is inconsistent with the “rule of law.”³¹⁰ For those, like Waldron, who wish to distinguish sharply between review of legislation and review of executive action, the puzzle becomes how to prevent courts from negating statutes through cramped interpretation rather than explicit invalidation—a puzzle for which democratizing statutory law provides a ready solution. All the same, worries about executive “tyranny” are not baseless; we turn to them here.

One way to understand this “rule of law” objection is as a concern about concentrated authority. Even conceding that judges have no monopoly on faithful interpretation of congressional enactments, one might reasonably fear that a charismatic (or even not so charismatic) individual vested with the power to decide whether to wield the authority of the state poses a greater risk of arbitrariness than does a system in which legal assessment and so, ultimately, power is divided. Indeed, as Professor Noah Rosenblum has written, the architects of the modern administrative state were highly attentive to such concerns.³¹¹

But to date our solution has been internal checks within politically accountable government. Hoping to balance the efficiency of a Progressive-era vision of largely unbounded administration against concerns about consolidation of power as observed in different European fascist regimes, New Deal reformers, in addition to providing for light-touch judicial review, embraced separation of powers within the executive branch.³¹² By doing so, Rosenblum explains, members of the President’s Committee on Administrative Management and others tried to build an effective bureaucracy that was, at the same time, something more than an “extension of the personality of the chief executive.”³¹³ Today, such internal strategies of dividing power are ubiquitous. As Dolbow observes, with some frequency jurisdictional bars on review of agency action also create alternative oversight mechanisms.³¹⁴ This is to say nothing of the various checks and obligations that agencies impose on themselves, in addition to what the APA or the relevant organic statute requires.³¹⁵

The worry with such internal separation of powers is that it merely recreates antidemocratic pathologies of the sort we have criticized here. A juristocracy of administrative law judges (ALJs) is not obviously better than our present situation. Nor, to note an analogous concern, is a financial regulatory system managed by “technocratic” economists something progressives should be eager to defend or promote.³¹⁶ To recognize these anxieties, however, is not to embrace the “unitary executive.”³¹⁷ Even if the goal of democratization is for political decisions to be made by overtly political actors, there are different strategies for bringing that reality about that do not require the consolidation of political power within a single individual.

One strategy is to substitute other forms of democratic control for presidential control. We have mentioned above the possibility of legislative courts as non-Article III venues.³¹⁸ As we discuss below, the administrative state has been and promises to be a site of participatory democracy in addition to its comparative representative democratic advantage. Notice-and-comment rulemaking is one familiar if limited example.³¹⁹ To the extent that decisions by agencies can be made with the input of affected communities, insulating such decisions from presidential control as well as judicial displacement could plausibly help “democratize” statutory law while also dispersing power in the way concerns about tyrannical accumulation of legal authority recommend. Further, the entrenchment of “countervailing” powers within the administrative process might help build our democracy generally by contributing to organizations and institutions that might successfully resist domination of our politics by monied interests.³²⁰

Another strategy is to benefit from, but also subordinate, the opinions of “independent” or “expert” officials. ALJs provide an instructive example. As Professor Emily Bremer documents, the APA creates a scheme under which administrative adjudication is subject to various procedural requirements and, even more relevant, are presided over by administrative law judges.³²¹ In turn, ALJs are subject to restrictions and requirements—prohibiting, for example, ex parte communications with interested parties outside the agency involved in the investigation or prosecution—meant to ensure their “impartiality.³²² ALJs also enjoy “impartiality” protections for promotion, like the assignment of salary determination to an entity outside the agency and the requirement that any adverse employment action be taken “only for good cause established and determined” by a centralized board.³²³

While the law governing ALJs mirrors the requirements and protections applicable to judges under Article III, the APA also “preserves,” as Bremer notes, “control over the various policymaking aspects of adjudication” to agency heads.³²⁴ In addition to power over appointments and the authority to issue rules and guidance ALJs are required to follow, agency heads may order ALJs to certify the corresponding “record” to agency heads for official “decision.”³²⁵ Alternatively, agencies may instruct ALJs to issue “initial” decisions that “may become final in the absence of review by the agency head.”³²⁶ In this way, the APA lets agency heads position themselves similarly to appellate courts in relation to trial courts—though, with the advantage that agency heads can reason explicitly as political actors. Comparably, the APA regime separates the prosecutorial and judicial functions and avails agencies of opinions of persons who are relevantly “neutral.”³²⁷ At the same time, the regime affords ultimate decisional authority—including, for example, deciding whether a statute requires or merely permits a particular reading, or whether a determination by the factfinder below is a finding of fact or an interpretation of law—to actors who understand their role as making policy within legal bounds.³²⁸

It follows that political allocation of interpretive authority is not incompatible with internal separation of powers, while also permitting flexibility that promotes accountable decisions and overtly political forms of policymaking. The same considerations explain why allocation projects have not been futile, thus reproducing juristocracy elsewhere, like an air bubble pushed from one place to another. Instead, political control over interpretation allows experimentation to achieve the proper balance of dispersion and power, as well as independence and responsiveness.

B. Juristocracy for the Most Vulnerable?

What about, then, the vulnerable and weak and courts that protect their rights or otherwise serve progressive understandings of justice? A characteristic assumption is that independent courts are the only credible source of hope for the marginalized, so that political control of legal interpretation would make their situation even worse. After arguing against that assumption, this Section proceeds to envision partial retention of courts for good causes to the extent plausible, before insisting that political struggle to relocate more legal interpretation (including adjudication) outside Article III ought to be the privileged goal for progressives.

In their recent article, Professors Cox and Kaufman remark that “impartial adjudication—once the backbone of the administrative state—has given way to a much more politicized vision of agency courts.”³²⁹ In voicing this complaint, they rightly highlight the hierarchical selectivity of conservative judges and scholars. The goal, Cox and Kaufman suggest, is to reserve Article III for privileged parties and their rights, most especially liberty and property rights, while assenting (as they must, barring the massive expansion of the federal judiciary) to the massive allocation of adjudication to the administrative state, where justice for the vulnerable and weak goes to die³³⁰— for example, immigration courts. Indeed, contrary to the image of conservatives at war with the administrative state, a closer look inspires fear that the Supreme Court in fact has an “increasing interest in insulating administrative courts from judicial oversight.”³³¹

The realities of the contemporary state and its routine victims, however, are not cause to idealize federal court protections. Conservatives today are hardly the only defenders of the privileged and their “private rights,” whose protection the federal judiciary has favored since its inception. Although Cox and Kaufman are convincing that administrative courts are not always uplifting, misplaced nostalgia for nonpolitical adjudication hardly follows from it, let alone the struggle to expand the remit of the federal judiciary it implies.³³²

Our perspective is, thus, the opposite: The judiciary is historically anything but an impartial political actor. Liberals should not defend the vulnerable and weak through the same Article III adjudication tactics conservatives reserve for the powerful and wealthy. Liberals should, instead, acknowledge that control of statutory interpretation is political through and through and struggle to make our politics friendlier to the nonprivileged—through electoral victory, legislative specification of law, and demystification of who makes law and how, once the legislature is done.

This does not necessarily mean that progressives should entirely repudiate courts. But progressives must reckon with how poorly courts have historically and recently served the very project of securing justice and rights that remains the most frequently cited argument in their defense. Our point is hardly just that courts do not “bring about social change”³³³ in a progressive direction—an assertion now so obvious that it is surprising it was once seen as debatable or wrong, before the liberal infatuation with courts collapsed in our time. Rather, it is that the record of courts in defending progressive ideals is so mixed.³³⁴

That liberals have failed to construct a progressive judiciary even at the height of their political era, while spending most of their time dreaming of doing so someday, suggests a grim verdict. It’s time to give up the ghost: It makes far more sense to both push legal interpretation even further outside Article III courts and push it in a progressive direction as a more overt political project. This general conclusion, still doubted sometimes by those who idealize courts for their enforcement of constitutional rights and values, is made even more powerful when we question who should control statutory meaning.

That said, as sketched above, nothing forbids what retention—of otherwise disempowered judicial control of statutory interpretation—remains (or becomes) credible or feasible. The concern to advance justice or protect rights for vulnerable and weak peoples, which no government institution has accepted to a great extent, is worth developing legally anywhere propitious to do so. This applies most obviously to extant rights enforcement within Article III afforded by various statutes, as well as more general private rights of action—especially when Congress has itself created them, like those authorized by 42 U.S.C. § 1983 for constitutional violations.³³⁵ The conservative drift of the judiciary has foreclosed or limited remedies in these matters. But disempowerment and eventual reallocation strategies can certainly be tailored to preserve court authority whenever it is plausible that there is no better alternative. How soon a world after courts is reached will have to be made conditional on achievement of existing Article III protections for justice and rights in comparable or superior form outside Article III first. Obstructions to progressive outcomes for the vulnerable and weak (roadblocks characteristically set up in the name of justice and rights for the powerful and wealthy) can be cleared, while leaving the federal courts alone when the obstructions they impose for better causes are allowed pending successful attempts to make them redundant and unnecessary.

Perhaps the actual objection to a politics of disempowering or dispensing with courts is that it is unimaginable that certain vulnerable and weak parties—and not just immigrants or prisoners—will ever receive majoritarian protection; their condemnation to be eternal wards of judges is their best hope, even if it is only when they are lucky.³³⁶ This anxiety lays behind Cox and Kaufman’s disheartening study of immigration authority. It may not be fair to generalize from so tragic an example of mass victimhood about the prospects of political control of law. But Cox and Kaufman are right that if “political control of immigration courts has always been the norm,” then “exporting the pathologies of immigration law to the rest of the administrative state” would be depressing indeed.³³⁷ Yet this objection, although understandable, pretends that there is an alternative to curing those pathologies politically. There isn’t.

No one is suggesting that the vulnerable and weak will automatically achieve justice, or see it done on their behalf, under a system in which legal evolution is a more openly political matter. But the real question is whether adding more Article III process is the likeliest path to better outcomes, in the immigration context as elsewhere. If the intuition is that adding a modest veto gate before deportation could only improve things, a supermajority rule imposed on removal decisions would have the same effect. More generally, if the baseline for the treatment of the vulnerable and weak is so unacceptable when legal evolution occurs under judicial control’s disguised politics, then making the future of law a necessarily and openly political struggle at least clarifies what is at stake for all concerned, without placing false hope in courts.

C. The Problem of Democratic Deficits

The last objection is that, insofar as democratic law is the goal, the political branches are not themselves (very) democratic. There is some irony, after all, in holding up the administrative state as a source of democratic legitimacy. Because agencies are staffed with “unelected bureaucrats,” conservatives increasingly complain that the executive branch has usurped congressional lawmaking authority.³³⁸ On the left, many object to the purportedly “technocratic” decisionmaking of agencies like the Federal Reserve, demanding “popular” control of monetary policy and similar areas.³³⁹

Some complaints about administration’s democratic bona fides are more forceful than others. As Professor Katharine Jackson has written, insistence by conservatives that agency officials limit themselves to faithful implementation of congressional will rests on an implausible, “transmission belt” picture of lawmaking, in which all significant policy judgements can be made by Congress ex ante—a picture belied by the reasons canvassed in Part I concerning the predictable underspecification of statutory law.³⁴⁰ Worries about the obfuscation of ideology in economic regulation is more persuasive (at least, to us). Still, our case for allocating power away from the judiciary and toward the executive accepts, for the most part, the administrative state with all its democratic imperfections.

As a threshold matter, our claims about the democratic pedigree of administrative interpretation are comparative rather than absolute. Administrative agencies, even in their present form, are responsive to democratic inputs in myriad ways. Executive-branch agencies are, most obviously, headed by actors appointed and removable by the President, and subject to centralized review.³⁴¹ Add to this congressional influence through oversight and (in most cases) budgetary control over both ordinary and “independent” agencies, and the input of elected officials on agency decisions is, at the very least, not insignificant. Agencies are also structured (to varying degrees) to receive input from interested parties, whether through notice-and-comment rulemaking or formal adjudication. In this respect, agencies also realize some amount of participatory democracy, participation facilitated by agencies’ candor about the (bounded) political character of their decisions.³⁴²

Whatever their (many) failings, then, agencies have a democratic advantage over life-tenured judges, even conceding that in some cases, like the Federal Reserve, that advantage is slim. To say that agencies are more democratic, though, is not to say that they should be empowered—and courts fully disempowered—as a matter of course. As in the constitutional context, disempowering reforms lend themselves to piecemeal implementation. Disempowerment strategies can be made conditional on superior democratic credentials in empowered decisionmakers. When a particular agency exhibits a toxic or reactionary culture, progressives specifically might reasonably refrain from empowering that agency immediately—and might go so far as to disempower it—to prioritize reform of the agency in the short term. In other words, like many conservatives have done, progressives might disempower courts (and empower agencies) selectively, at least as an initial step toward more expansive small-d democratic reform.³⁴³

More generally, our call to disempower courts in statutory cases should be understood as part of a larger democratizing agenda. It is of a piece, among other things, with scholarship advocating that the administrative state itself be made more democratic. Professor Karen Tani and others have observed that administration has, at key moments, been a space in which social movements have been able to advance through direct participation an egalitarian vision—for example, promoting a conception of Social Security and other federal benefits as entitlements rather than charity.³⁴⁴ And activist-scholars like Professor Sabeel Rahman have argued that there are various ways to promote and expand participation in regulation by affected parties, and, in particular, to “institutionaliz[e] effective contestation,” such that “civil society groups, social movements, and grassroots constituencies” could offset influence already exerted by monied interests.³⁴⁵ As Rahman in particular has emphasized, promoting this sort of direct participation requires confronting not only the sort of juristocratic meddling we criticize here, but also the more “technocratic” conception of administration that even many liberals embrace.³⁴⁶

Ultimately, then, an agenda to envision a legal world after courts combines with a larger democratizing agenda for law and legal institutions across the board. The first agenda will often be conditional on the second, but the converse is also true.

Conclusion

Of their “double axes”— constitutional invalidation and statutory interpretation—with which judges “have battered their way to supremacy,” as Learned Hand once memorably put it,³⁴⁷ it is not obvious which edge is sharper. Invalidation is rare while interpretation is quotidian, and although one is more confrontational, flagrant, and spectacular, the other is endemic and very possibly more damaging.

If so, the court reform debate of recent years³⁴⁸ may have gotten off on the wrong foot due to its overwhelming focus on the power of invalidation rather than interpretation. This Article begins to correct that mistake. But rather than accept glumly the rule of judges in statutory lawmaking as inevitable, our ideal response is the same as constitutional lawmaking: to devise and revive tools to chasten the federal judiciary’s power or to extinguish it altogether.

Experimentation with such tools—which, as we have shown, has been deeply rooted in American history from the first—is no small matter. At stake is who rules. Asking who interprets the law in a democracy amounts to asking whether the people, through their most accountable servants, make the rules.

* Professor of Law, Harvard Law School.
** Chancellor Kent Professor of Law and History, Yale Law School.
Thanks for comments to Aslı Bâli, Christine Desan, Blake Emerson, William Eskridge, David Froomkin, Amy Kapczynski, John Manning, and especially Nicholas Parrillo and Daphna Renan, as well as others in the audiences at the Boston College Law School Faculty Workshop, the Harvard Public Law Workshop, the UCLA Law School Faculty Workshop, and the Yale Legal Theory Workshop. For their help in finalizing the text, we are grateful to Dana Abelson, Nina Gayleard, and Erin Mizraki, as well as the editors of the Michigan Law Review, especially David Canada and Ethan Greenberg.