Tease: to tantalize especially by arousing desire or curiosity often without intending to satisfy it
Merriam-Webster Dictionary (3rd definition)
The Constitution forbids Congress from delegating legislative authority to the Executive Branch, but the Court has not struck down a statute on non-delegation grounds since 1935 in Panama Refining Co. v. Ryan, and A.L.A. Schechter Poultry Corp. v. United States. But recently, the Supreme Court gave very clear and strong indications that it was interested in re-visiting its approach to non-delegation, and yet when it was presented with two cases in which the issue was clearly and directly presented, it denied review in both: American Institute for International Steel v. United States, and Center for Biodiversity v. Wolf. Why? Was it because the issue was not properly presented, was there some procedural problem, did the Court simply change its mind, or was there something else that caused the Court to deny review with no dissents noted in either? Only the Justices know the reasons, but it matters to those who are considering whether to bring further delegation challenges. There may be no knowable answer, but the facts, especially those from 2018 to date, provide the data from which to make a semi-educated guess.
Until 2018, the Court’s most recent delegation decision was its 2001 decision in Whitman v American Trucking Ass’n, in which the Court unanimously rejected a delegation challenge to a provision of the Clean Air Act, overturning a contrary ruling of the D.C. Circuit. The Court agreed that the doctrine was alive, but that this statute had met the “intelligible principle” test that has been used since 1928 to decide these questions. To most observers, the doctrine has seemed to exist in theory only, although it has been occasionally used as a reason to construe a statute narrowly, as the Court did with the Occupational Safety & Health Act in Indus. Union Dep’t AFL-CIO v. Am. Petroleum Institute.
Then on March 5, 2018, the Court granted review in Gundy v. United States, 17-6086, in which the petitioner argued that the provision of the Sexual Registration and Notification Act that gave the Attorney General the power to make new registration requirements applicable to convictions that occurred before the effective date of the law amounted to a delegation of legislative authority. The provision at issue was unlike those in the Clean Air Act or other regulatory provisions where the delegation was needed because the issues were technical or complicated, or the facts or the science changed over time. The retroactivity issue was very simple, but it had significant consequences for those convicted of the covered offenses, as well as the states that would have to administer the new requirements. The reason that Congress did not specify whether the statute was to have retroactive application was that the House of Representatives and the Senate could not agree on the answer, and so they punted it to the Attorney General to decide on retroactivity, without providing any explicit criteria or other direction as to how the Attorney General should make that determination. The grant of certiorari, which was limited to only Question 4, was striking in another respect: as the brief in opposition of the Solicitor General told the Court, all eleven circuits that had considered the delegation claim had rejected it and the Court had denied review in fifteen cases raising the issue.
By coincidence, that same month, President Trump imposed a 25% tariff on imported steel products (and a 10% tariff on aluminum imports) under section 232 of the Trade Expansion Act of 1962. That provision allows the President to “adjust the level of imports” for any product whose importation “may threaten to impair the national security,” which is defined to include any injury to the economy as a whole or to any segment (industry) in it. In addition to tariffs, the adjustment may also (or in addition) take the form of quotas, embargoes, and/or licensing fees, with no limits whatsoever on their amounts. Nor does the statute impose any limits on the duration of the adjustment, and there is no right to judicial review under section 232, and the Administrative Procedure Act does not apply because the President is not an agency covered by it.
For many years, I have tried, without success, to revive the non-delegation doctrine in cases such as Bowsher v. Synar, and Mistretta v. United States. When I read about the President’s import tariffs, I contacted my GW colleague who teaches international trade, Steve Charnovitz, to tell me which law the President used to do that. He pointed me to section 232 and also to a program that he and Professor Tim Meyer of Vanderbilt had done on section 232, and I was on my way. Working with trade lawyers Gary Horlick, Don Cameron, and Will Planert, as well as Steve and Tim, we decided to proceed on a delegation challenge to section 232. Our clients were the American Institute for International Steel (AIIS)—a 120-member trade association, whose members include importers who pay the tariffs, as well as many other entities who were adversely affected by reduced imports or the higher prices resulting from the tariffs—and two of its members. A major factor in our decision to sue was the grant of review in Gundy.
In late June 2018, we filed suit in the U.S. Court of International Trade (CIT), which was the only available forum. The CIT agreed that our case raised a sufficiently serious challenge to warrant appointing a three-judge panel to rule on the only claim in the case: that section 232 violated the prohibition on delegation of legislative authority in Article I. The Government raised no procedural issues and even agreed that we had standing and that there were no disputed issues of material fact that would prevent the court from reaching the merits. We lost in the CIT, but the opinions suggested that the court agreed with AIIS that section 232 raised significant delegation concerns. As Judge Katzmann’s separate opinion dubitante observed: “If the delegation permitted by section 232, as now revealed, does not constitute excessive delegation in violation of the Constitution, what would?” But we lost in the CIT and in the Federal Circuit, because both courts concluded that they were bound the Supreme Court’s decision in a 1976 ruling in Federal Energy Administration v. Algonquin SNG, Inc.
In that case, the Supreme Court did not consider a direct constitutional challenge to section 232, like the one we brought by AIIS. Instead, the Court was only reviewing the D.C. Circuit’s decision that section 232 did not, as a matter of statutory interpretation, authorize the use of license fees as a remedy. In the course of defending that interpretation, the plaintiffs urged the Court to construe section 232 narrowly to avoid a delegation challenge, and the Court responded by concluding that there was no serious delegation issue presented. We argued that Algonquin was distinguishable on several grounds, but neither the CIT nor the Federal Circuit agreed. We were confident that Algonquin would not be a significant barrier if the Supreme Court took the case because it could easily limit the decision or overrule it, especially because the case had only been cited by the Court once in more than 40 years, and even then not for the delegation ruling.
Meanwhile, after we lost in the CIT, the Court decided Gundy, ruling in favor of the Government in June 2019. Justice Kagan wrote for the plurality, concluding that, despite the absence of any express criteria or other direction to the Attorney General on the issue of retroactivity, the statute should be interpreted to require retroactive application unless the Attorney General determined it was not “feasible” to do so. This, Justice Kagan concluded, eliminated any delegation problem by providing an intelligible principle to guide the Attorney General’s exercise of discretion.
Justice Gorsuch wrote a lengthy dissent, joined by Chief Justice Roberts and Justice Thomas, in which they rejected Justice Kagan’s reading of the statute and explained in detail the ways in which prior decisions of the Court had distorted the application of the intelligible principle test and had allowed Congress to turn over major policy issues to the Executive Branch in violation of Article I. In response, Justice Kagan did not suggest that the delegation doctrine was dead and agreed that, if the statute were as broad as the dissent suggested, the Court would have to face the delegation issue. Justice Alito concurred in the result (Gundy loses) but did not join the plurality; instead, he noted that he would be open to reconsidering the delegation doctrine in an appropriate case.
Justice Kavanaugh did not participate in Gundy, which was argued several days before he was sworn in as a Justice in October 2018. His absence was noted by the petitioner in Gundy, who promptly asked for a rehearing with Justice Kavanaugh sitting. In addition, there were a number of other cases raising the same delegation issue as Gundy that had been held or were in the pipeline, for which Justice Kavanaugh would be available. It was not until mid-November that the Court denied rehearing in Gundy and denied review in the other cases. Those denials were expected, but what was not expected was Justice Kavanaugh’s two-page statement in respect of denial of certiorari in Paul v. United States, in which he too expressed an interest in considering the delegation doctrine, while seeming to suggest a different approach—grounded in the major questions doctrine—from that in Justice Gorsuch’s dissent in Gundy.
For these reasons, we were cautiously optimistic that the Court would grant review when we sought cert in AIIS in March 2020. All we needed were four votes to hear the case, and the three dissenters in Gundy and the separate expressions of interest in the question by Justices Alito and Kavanaugh, appeared to give us one more than we needed. The delegation issue was teed up by the fact that, unlike Gundy, the Government did not dispute our construction of the applicable statute. Despite paying lip service to the constitutional requirement that section 232 must cabin the President’s discretion in some respect, the Government was never able to identify a single act that the President could not take regarding imports that would violate section 232, including restricting imports of peanut butter or denying income tax deductions for the 25% tariffs paid. That gave us a way to limit the impact of a favorable ruling on the statutes governing other administrative agencies.
The petition also cited recent Court decisions, such as Sessions v. Dimaya, holding portions of federal statutes void for vagueness in which Justices Gorsuch and Thomas analogized that doctrine to the non-delegation doctrine because in both Congress had failed to limit the reach of the law and had, in effect, assigned legislative powers to the other branches. Petitioners also observed that this case was similar to United States v. Lopez, in which the Court held that the Commerce Clause could not be used to sustain a federal law that outlawed the possession of guns near a school. In Lopez, the Court concluded that upholding the statute would result in there being no limits on a congressional power that was intended to be limited. As applied here, Lopez required that the President’s authority under section 232 had to have some limits, but the Government was unable to identify any.
Despite the Court’s ability to overrule or limit Algonquin, the Government’s opposition relied principally on that case. The opposition did not suggest any procedural or other barriers that would prevent the Court from reaching the merits. It did argue that section 232 could be supported by the President’s foreign affairs powers, but at most that would be a defense on the merits. Moreover, Algonquin did not rely on that rationale, and we replied that the President has no inherent power to impose tariffs absent congressional authorization. But alas we were mistaken in our optimism, and the Court denied cert without a dissent.
Unbeknown to us as we were litigating AIIS, there was another case raising a different delegation issue that ended up being considered by the Court the week after ours. The challenge in Center for Biodiversity v. Wolf, No. 19-975, was to the law that authorized the Department of Homeland Security to waive more than forty federal statutes that would otherwise have to be satisfied before building any portion of a border wall with Mexico. All judicial review except for constitutional questions was expressly forbidden, and cases went directly from the district court to the Supreme Court on certiorari. Earlier and less expansive versions of the waiver law had been previously challenged, and the Court had denied several of them over a decade before the current suits were filed in March and October 2018. DHS prevailed in both cases below in orders entered in September 2019, relying on the prior decisions involving substantially the same claims.
In the petition filed in February 2020, the Center relied heavily on Gundy and Paul and the extraordinary transfer of power from Congress to the Executive, which was empowered to render more those laws inoperative for any reason it wanted and with no judicial review. The Government’s opposition relied heavily on the pre-Gundy denials of cert (the functional equivalent of Algonquin in AIIS), and, as in AIIS, did not suggest that there were factual or procedural barriers to reaching the merits. Again, the Court denied review on June 29, 2020, also with no noted dissents.
What happened and why? The Court might not have wanted to be in the position of having to overrule Algonquin because the Court has been sharply divided over the application of stare decisis in cases such as Wayfair and Ramos, not to mention its recent abortion case. There were, however, several ways to avoid expressly overruling Algonquin while still reaching the merits of the delegation challenge presented. And in Biological Diversity, there were only cert denials in the way, and they pre-dated Gundy.
In theory, there are other reasons why each case could have been rejected as “unsuitable” for re-visiting or even re-applying the intelligible principle test, but even the sophisticated oppositions of the Solicitor General suggested none of these. Some Justices might have been troubled by the fact that AIIS involved the President, in contrast to a federal administrative agency, but the issue in non-delegation cases is whether Congress violated Article I, not whether the President acted lawfully. And Biological Diversity challenged waivers, which might have been viewed as a poor case to deal with the issue of giving agencies too much power to impose burdens on regulated entities. Or, in AIIS, the Court might have concluded that the law had been around for over 50 years and that this was the only time that it had been used in a manner anywhere as expansively as this. The Justices also might have realized that some of the laws governing the Federal Reserve are very open-ended, and feared that those powers were essential in times of economic crisis. And both cases go to specific policies on which Trump had campaigned—raising tariffs on imports and building the Mexican border wall—two political hot potatoes.
Or it may have denied review in both cases for any other reason, or for no reason, or because the Justices did not want to stir up the non-delegation controversy because they had no good answer to the question of what test would replace or refine the intelligible principle doctrine in its current incarnation. Until recently, any of those answers would have been reasonable. But after Paul and Gundy (in which Justice Gorsuch went to some length in his dissent to sketch out his view of a viable approach to a revived delegation doctrine), a majority of the Court virtually invited litigants to bring on delegation challenges, which makes the denial of cert in both cases seem inexplicable. Tease may not be the right word for what a majority of the Court did in those cases, but it is not obviously incorrect.
What happens to non-delegation from here? First, because we do not know why the Court said no (and almost certainly will not know for many years, if ever), rational litigators do not have much to go on. Second, litigants should assume that any arguably controlling precedents, such as Algonquin in AIIS, will be a barrier to obtaining a favorable decision by lower courts that relies on Gundy and Paul. Third, although Whitman teaches that an agency cannot cure a delegation problem, future litigation should focus on newly enacted statutes so that affected parties and the lower courts have not already become comfortable dealing with a very broad law in the ordinary case, before confronting a non-delegation challenge.
The real question is whether any rational party would undertake suits like these two in the face of what happened in those cases. I would have to think long and hard before undertaking such an endeavor, despite having done all the groundwork in the AIIS case. Perhaps the next case will be like Gundy, a criminal case where counsel is undeterred by considerations that would cause civil litigants to back away from raising the issue. And if that happens, the result may end up like Gundy, with no new learning on the non-delegation question. All in all, a very unsatisfactory outcome, one that is perhaps even worse than losing these cases in the Supreme Court on the merits.
Post-Script: Peeling Back a Little of the Court’s Secrecy
There are significant pieces of evidence that might help in answering the WHY question, short of learning what the Justices said at the conference when they decide which cases to hear. There are the law clerks’ memos, which recommend grants or denials, but they will not, and should not, be available until the Justices’ papers are opened-up decades from now.
There is one record for each conference whose disclosure could be routinely made public perhaps a year after it was created: the discuss list. Not all cases scheduled for a given conference are actually discussed; in fact, the opposite is more likely to be true, so that very few cases are debated by the Justices in person, although a single Justice can make that happen for any case. For these two cases, and also for others, I would be very interested to learn that a case I thought was certworthy did not make the discuss list, and vice versa. And if I found out that my “good” cases did not make that list, I might re-assess and file fewer petitions, or at least different ones, which might benefit my clients and the Court, with next to no interference with the Court’s deliberative process.
Alan B. Morrison is the Lerner Family Associate Dean for Public Interest Law at George Washington University Law School where he teaches constitutional law and civil procedure. He was lead counsel for the plaintiffs in the case of American International Institute for Steel v United States discussed in this essay.