The Living Anti-Injunction Act

In the coming months, the Internal Revenue Service is likely to issue a slew of new regulations interpreting the December 2017 federal tax reform legislation. These regulations are likely to define the scope of the new deduction for pass-through entities; determine the reach of the new base erosion tax on multinational enterprises; fill in the details of the new “opportunity zone” program aimed at encouraging investment in low-income communities; and address a wide range of other important matters.[1] Inevitably, some taxpayers will object to these regulations and will seek to challenge the new rules in court. When, where, and how they can do so will depend upon the way courts construe the 150-year-old Anti-Injunction Act (AIA).

For decades, individuals and entities wishing to contest their tax liabilities have had a choice among three paths: (1) file a prepayment petition in the U.S. Tax Court; (2) pay the tax and then sue for a refund in federal district court; or (3) pay the tax and then sue for a refund in the U.S. Court of Federal Claims.[2] What they could not do is seek an injunction preventing the Internal Revenue Service from assessing or collecting the tax in question. Standing in their way would be the AIA, which provides, in relevant part, that “no suit for the purpose of restraining the assessment or collection of any tax shall be maintained in any court by any person.”[3]

All that is now in doubt. In 2016, the U.S. Chamber of Commerce sued the IRS in the U.S. District Court for the Western District of Texas, seeking to set aside a Treasury tax regulation[4] that determines the circumstances under which a domestic entity that switches its legal domicile to a foreign country becomes subject to a special tax.[5] The IRS argued that the AIA clearly barred the chamber’s action. In a decision that surprised many observers (including me[6]), the district court said last fall that the AIA presented no barrier to the chamber’s claim for equitable relief. According to the court, the regulation “is not a tax,” but instead “determin[es] who is subject to taxation.”[7] The court then proceeded to the merits and agreed with the chamber that the regulation should be set aside.

As far as judicial decisions on matters of tax procedure go, this one was a bombshell. A headline in the trade publication Tax Notes announced that the ruling “throws [the] door open” to more challenges to IRS rules.[8] Tax scholar Andy Grewal noted that the district court’s decision “breaks from the common (though not necessarily correct) understanding” of the AIA.[9] Fellow tax scholar Bryan Camp went one step further and argued that the decision was not only a departure from precedent but also a clear misinterpretation of the AIA. In his view, “[t]his is exactly the kind of suit that the Anti-Injunction Act is supposed to stop.”[10] The IRS has appealed from the district court’s ruling to the U.S. Court of Appeals for the Fifth Circuit.[11]

With the fate of the AIA hanging in the balance, now is the perfect moment for a thoughtful and thorough treatment of the statute that traces the law’s evolution from its origins to the present day. And in Restoring the Lost Anti-Injunction Act, Professor Kristin Hickman and Gerald Kerska provide exactly that.[12] Indeed, their new article on the AIA is quite possibly the most comprehensive analysis of the Act ever written. “Timing matters,” Hickman and Kerska write in their opening sentence,[13] and while the authors are referring to the timing of judicial review, their own timing is impeccable.

Hickman and Kerska’s analysis also provides a thought-provoking counterweight to the conventional wisdom that Chamber of Commerce v. IRS marks a sharp break from the past. The narrow interpretation of the AIA adopted by the district court in the Chamber of Commerce case is, in their view, largely consistent with the “lost” history of the Act.[14] According to Hickman and Kerska, the AIA historically applied only after a taxpayer filed a return and federal tax officials began their assessment and collection efforts. Pre-enforcement judicial review of a tax regulation would, on this reading, fall outside the statute’s scope.[15]

Whether or not one ultimately agrees with Hickman and Kerska’s conclusion, their article is likely to become the jumping-off point for future debates about the AIA. I, for one, was impressed by Hickman and Kerska’s historical and doctrinal heavy lifting but was unpersuaded by their bottom line. This essay briefly summarizes Hickman and Kerska’s case for a narrower reading of the AIA and then responds with three criticisms of the authors’ argument. Specifically, I argue (1) that the history of the AIA is at best inconclusive as to whether the statute should be construed broadly or narrowly; (2) that developments in federal tax and administrative law since 1867 do not weigh decisively in favor of a narrow interpretation of the statute; and (3) that the AIA has come to play an important role—unacknowledged in Hickman and Kerska’s otherwise comprehensive analysis—in protecting an under-resourced IRS from an onslaught of administrative law challenges across a wide range of litigation forums. I end by arguing that any further narrowing of the AIA should be done by Congress—not by the courts—and should be accompanied by an increase in IRS resources and additional limits on taxpayer forum shopping.

I. Lost and Found?

Hickman and Kerska begin their argument for a narrower AIA with a deep dive into the statute’s history. As they write, “historical analysis provides a powerful tool for resolving the AIA’s meaning and scope.”[16] Yet this “powerful tool” turns out to be a double-edged sword: the lessons that the authors draw from the AIA’s history can be deployed to argue for either a narrow or expansive reading of the statute.

The story starts with the Revenue Act of 1862,[17] the first federal income tax law to take effect.[18] As Hickman and Kerska explain, this statute—enacted in the midst of the Civil War—empowered the President to divide the country into districts and to appoint “assessors” and “collectors” to administer the new tax law in each district.[19] The job of the assessors and their assistants was to receive returns, conduct investigations to determine whether taxpayers had understated their liability, publish tentative assessments, resolve appeals, and then make out a list of amounts due. Section 19 of the Revenue Act of 1862 instructed collectors to publish those lists and, if need be, seize and sell the property of tax delinquents.[20]

The AIA enters the narrative a half decade in. Styled as an amendment to be appended to Section 19, it provided (in language that has changed little in the years since) that “no suit for the purpose of restraining the assessment or collection of tax shall be maintained in any court.”[21] Neither the sponsor of the amendment, Senator William Fessenden of Maine, nor any other member of the House or Senate said anything on the record about the provision’s purpose, leading one later commentator to conclude that the Act’s “legislative history is shrouded in darkness.”[22] But according to Hickman and Kerska, “Congress did not need to be more specific about the AIA’s scope because the meaning of the new restriction on judicial review was obvious from its statutory context.”[23] As they see it, the location of the new language at the end of Section 19 meant that the AIA was intended as “a limited remedy for judicial obstruction” of the “particular procedures” for assessment and collection prescribed by that Section.[24]

In the century and a half since the statute’s passage, Congress has carved out a number of specific exceptions to the AIA’s coverage. A few of those carve-outs now allow for prepayment petitions in Tax Court;[25] innocent spouse relief;[26] and injunctions in cases where a taxpayer seeks a hearing before the IRS seizes her property[27]a preparer seeks to delay collection of penalties against her,[28] or a person other than a delinquent taxpayer seeks to block the sale of property in which she holds an interest.[29] As Hickman and Kerska observe, “the only amendments to the AIA have come when Congress wanted to expand the availability of judicial review and, correspondingly, to make clear Congress’s intention to limit the AIA’s reach.”[30] The authors appear to interpret this as an indication that Congress favors a narrower AIA.

Three further developments play an important role in Hickman and Kerska’s narrative. The first involves the Tax Injunction Act of 1937 (TIA), which provides that federal district courts “shall not enjoin, suspend or restrain the assessment, levy or collection of any tax under State law where a plain, speedy and efficient remedy may be had in the courts of such State.”[31] In the 2015 case Direct Marketing Association v. Brohl, the Supreme Court held that the TIA does not prevent a federal court from enjoining a Colorado law requiring out-of-state retailers to share certain tax-related information with Colorado tax authorities.[32] The Supreme Court also said that it “assume[d] that words in both [the Anti-Injunction and Tax Injunction] Acts are generally used in the same way” and so would interpret the two statutes in tandem.[33] According to Hickman and Kerska, “the reasoning of Direct Marketing is different from and difficult to square with at least some of the Court’s past AIA precedents.”[34] A more circumscribed construction of the Anti-Injunction Act would “bring[] the AIA in alignment” with the post-Direct Marketing TIA.[35]

Second, Hickman and Kerska note that the Administrative Procedure Act of 1946 (APA) and the Supreme Court cases construing it have established a general presumption in favor of pre-enforcement judicial review of final agency action.[36] As the Supreme Court said in Abbott Laboratories v. Gardner, “judicial review of a final agency action by an aggrieved person will not be cut off unless there is persuasive reason to believe that such was the purpose of Congress.”[37] Interpreting the AIA to preclude pre-enforcement review of Treasury regulations puts it in some tension with the APA’s presumption of reviewability. Hickman and Kerska seek to ease that tension with a narrower reading of the AIA.[38]

Third, the authors emphasize that the federal tax laws do much more today than they used to do.[39] Important antipoverty programs, such as the Earned Income Tax Credit[40] and the Low-Income Housing Tax Credit,[41]are run through the Internal Revenue Code. Congress also uses the tax system to subsidize—and to regulate—health insurance, retirement saving, higher education, and charitable giving, among countless other tax expenditures. Hickman and Kerska fear that a robust application of the AIA will thus interfere with judicial oversight over a wide swath of the modern administrative state.

Drawing from the “lost history” of the AIA as well as these more recent developments, Hickman and Kerska propose a new “engagement test” that would “limit the AIA’s scope to those cases in which the IRS has initiated enforcement proceedings of one manner or another against a particular taxpayer” or the taxpayer has filed a return.[42]According to Hickman and Kerska, this new test would restore the Anti-Injunction Act to its “original scope” and harmonize it with the Tax Injunction Act and Administrative Procedure Act.[43] And they add that the test would be “very easy” to apply in the “vast majority” of cases, thus bringing clarity to what is now a morass of conflicting case law.[44] While they suggest that courts can adopt the test on their own, Hickman and Kerska also urge Congress to codify their new engagement test, and they propose legislative language to that effect.[45]

II. Assessing the Engagement Test

Hickman and Kerska’s “engagement test” has undeniable appeal. It provides a plausible interpretation of the AIA’s text, and by narrowing the statute’s scope, the test would ease the discomfort that many (including me) feel when legitimate challenges to Treasury regulations are tossed aside on jurisdictional grounds. And apart from the merits of the test, there is much to admire about the enterprise in which Hickman and Kerska engage. This is the sort of “practical” legal scholarship that jurists such as Judge Harry Edwards of the D.C. Circuit have urged academics to produce more often.[46]Hickman and Kerska’s article accomplishes exactly what Judge Edwards said that practical scholarship should do: it “gives due weight to cases, statutes and other authoritative texts, but also employs theory to criticize doctrine, to resolve problems that doctrine leaves open, and to propose changes in the law.”[47] Their careful and powerful argument is likely to elicit attention from judges and their clerks who come across AIA cases on their dockets.

Yet Hickman and Kerska’s inferences from the AIA’s history are also open to question. As the authors note, “[i]n 1867, when the AIA was adopted, the only circumstances in which a taxpayer might have sought injunctive relief from assessment or collection would have occurred when revenue officials acted to enforce the tax laws against particular taxpayers.”[48] Nearly eight decades before the APA and a full century before Abbott Laboratories v. Gardner, the notion that the 1867 law later would preclude pre-enforcement challenges to Treasury regulations would have seemed foreign to the Reconstruction-era Congress.

But if members of Congress in 1867 could have peered far into the future, it is not clear whether they would have wanted the AIA to apply to pre-enforcement judicial review of Treasury regulations. In Hickman and Kerska’s view, Congress wanted the statute to apply only to injunctions against assessors and collectors who were enforcing the tax laws against particular taxpayers. Hickman and Kerska also note, though, that the statute originally “forced aggrieved taxpayers to pay their taxes as assessed and sue the government for a refund” rather than pursue alternative remedial paths.[49] The alternative of a pre-enforcement challenge to a Treasury regulation was not one that existed at the time, but if it had, perhaps Congress would have wished to cut that route off too. In other words, we know that Congress wanted aggrieved taxpayers to sue for a refund rather than to seek injunctive relief against assessors and collectors, but that tells us little about whether Congress wanted to allow other end runs around the refund remedy.

The postenactment legislative history of the AIA also is amenable to competing inferences. On the one hand, the fact that Congress has narrowed the statute’s scope again and again might suggest that it disfavors an expansive reading. On the other hand, Congress clearly knows how to cut back on the AIA when it wants to, and the fact that it has stood by as courts have construed the statute expansively might suggest that Congress acquiesces to the broader interpretation. In all likelihood, very few members of Congress in the past century and a half have arrived at any opinion whatsoever as to whether the AIA should preclude pre-enforcement judicial review of Treasury regulations. The fact that Congress has carved out a number of other exceptions to the AIA does not mean that the legislative branch favors the particular narrowing that Hickman and Kerska propose.

The Supreme Court’s interpretation of the TIA in Direct Marketing Association v. Brohl also does not yield clear lessons with respect to pre-enforcement judicial review and the AIA. The similar language in the two statutes—”restrain[] the assessment . . . or collection of any tax”[50]—demands some limiting principle; otherwise, a suit to stop the construction of a state highway could be barred by the AIA or TIA because it interferes with the ability of tax authorities to travel around and do their jobs. The Court in Direct Marketing concluded that the Colorado law, which required retailers to share tax-related information with the state but did not impose any tax on them, was too attenuated from “assessment” and “collection” for it to fall within the TIA’s protection.[51] But that does not tell us whether an order that sets aside a regulation determining actual tax liabilities operates as a “restrain[t]” on assessment and collection.

The rise of pre-enforcement judicial review under the Administrative Procedure Act fails to illuminate the Anti-Injunction Act’s scope any further. Hickman and Kerska write that “Congress, in Section 559 of the APA, expressly instructed courts to read the APA and specific statutes like the AIA so as to give maximum effect to both.”[52] But that is plainly not what the APA mandates. Section 559 says that the provisions of the APA “do not limit or repeal additional requirements imposed by statute.”[53] The AIA is one such additional requirement, instructing taxpayers as to when and where they can seek relief. Section 559 goes on to say that any “[s]ubsequent statute may not be held to supersede or modify [the APA], except to the extent that it does so expressly,”[54] but the Anti-Injunction Act, which precedes the Administrative Procedure Act by nearly eighty years, is not a “subsequent statute.”

The octopus-like extension of federal tax law’s tentacles into new areas of American life likewise tells us little about the AIA’s reach. To be sure, Congress circa 1867 could not have imagined that programs like the Earned Income Tax Credit and the Low-Income Housing Tax Credit would be run through the Internal Revenue Code. But members of Congress most certainly did know about the existence of the AIA when they first enacted the Earned Income Tax Credit in 1975[55] and the Low-Income Housing Tax Credit in 1986.[56] There are many reasons why lawmakers might have chosen to place these provisions in Title 26 and to assign administrative responsibility to the IRS, but the fact that the AIA would shield regulations under these provisions from pre-enforcement judicial review might have been one attraction. At the very least, if one believes that Congress carefully placed the AIA where it did in 1867 so as to send a signal regarding the provision’s scope, it becomes difficult to argue that Congress scattered other provisions throughout the Internal Revenue Code by sheer accident or happenstance.

None of this is to suggest that the inferences drawn above from the Anti-Injunction Act’s history, the Supreme Court’s Tax Injunction Act case law, and subsequent developments in federal administrative and tax law are more plausible than the conclusions that Hickman and Kerska reach. Rather, the point is that the materials upon which Hickman and Kerska rely are inconclusive. Neither the partisans for a broader reading of the AIA nor the proponents of a narrower interpretation can claim that the historical origins, statutory context, or subsequent developments in administrative law and Tax Injunction Act jurisprudence confirm the correctness of their position. At least as I see it, the debate over the AIA’s proper scope ultimately turns on normative arguments that are based on contemporary concerns and conditions. The AIA’s “lost history” can inform this debate but cannot resolve it.

III. A Modern Anti-Injunction Act

Against a present-day backdrop, the AIA stands out as peculiar in several respects. First, the fears that seem to have motivated the statute’s enactment appear outmoded today. In light of our modern pay-as-you-go tax system as well as the United States’ access to deep and liquid capital markets, it is hard to imagine any injunction seriously disrupting the flow of federal revenue.[57] And insofar as the law was intended to protect individual local tax collectors from vexatious litigation,[58] there is something strange about using it to shield the IRS as a whole from suit.

Meanwhile, the statute sometimes dictates not merely the time and forum of taxpayer challenges but whether such challenges will be pursued at all. Consider the inversion-related regulations at issue in Chamber of Commerce v. IRS.[59] The special tax on domestic companies leaving the country is sufficiently steep that the very possibility of having to pay it will deter many firms from moving their legal domicile abroad. The in terrorem effect of certain Treasury regulations may be so great that the rules will remain immune from challenge unless they can be contested in a pre-enforcement action. While the inversion regulations strike me as an appropriate exercise of IRS authority, there is certainly something disconcerting about the notion that the IRS could issue legally defective rules and escape judicial oversight.

Yet even if the AIA has outlived its original purpose, and even if it yields normatively unattractive consequences in certain circumstances, the statute still serves at least two useful ends. First, it relieves some of the immense pressure on the IRS’s already-strained regulatory resources. As Hickman has argued elsewhere, temporary Treasury regulations, IRS revenue rulings, and other guidance documents issued by the IRS may be vulnerable to APA challenges on the grounds that these pronouncements did not go through the notice-and-comment process required for so-called “legislative rules.”[60] Yet as Hickman also acknowledges, broad application of the notice-and-comment requirement to temporary Treasury regulations and other IRS pronouncements would make it more difficult for the agency to respond to taxpayers’ need for guidance and could at times be “ridiculously wasteful.”[61] And this is at a time when the IRS has precious few resources to waste: the agency’s workforce is only two-thirds of what it was a quarter century ago,[62] and the Trump administration’s most recent budget proposes base funding for the agency that, in real terms, is down by more than one-fifth since 2010.[63] By delaying challenges to IRS pronouncements, the AIA gives the agency additional time to complete the notice-and-comment process for final rules while also allowing it some flexibility in issuing temporary regulations and other stopgap measures.

Second, and in a similar vein, the AIA protects the IRS from forum shopping by plaintiffs who otherwise would seek a nationwide injunction in the friendliest district court that they could find. Without the AIA, sophisticated taxpayers and the interest groups that represent them would enjoy a “general hunting license” to fire at the IRS in different jurisdictions until one of their shots strikes flesh.[64] This is not a problem unique to the IRS: other commentators have noted that the increasingly widespread use of nationwide injunctions poses a growing challenge to administrative agencies of all sorts.[65] But the fact that this problem plagues other administrative agencies is not a reason to foist it upon the IRS as well. Narrowing the AIA without also reining in the practice of nationwide injunctions would make an already-significant problem that much worse.[66]

Against these benefits must be weighed the cost of delaying a day in court for taxpayers with valid grievances. But existing Supreme Court case law construing the AIA ameliorates some of the statute’s more draconian effects. The Supreme Court has held that the AIA applies “only when Congress has provided an alternative avenue for an aggrieved party to litigate its claims on its own behalf;”[67] if there is no way for a party to challenge an IRS determination in a later Tax Court proceeding or refund suit, the party can seek a pre-enforcement remedy. The Supreme Court also has carved out a limited exception allowing pre-enforcement injunctions when the challenger can show that she would suffer “irreparable injury” from a delay and that the IRS could not possibly prevail on the merits.[68] As noted above, this still deters some taxpayers from challenging certain IRS positions indefinitely. For example, a company is unlikely to undertake a merger so that it can move its legal domicile overseas unless it can be sure that it will avoid the special tax that is at issue in Chamber of Commerce.[69] But in most cases in which a taxpayer seeks a deduction or contests the inclusion of an item in income, the statutory scheme ultimately allows her a judicial hearing and the possibility of full relief.

To be sure, the AIA is a rather roundabout way of writing a statute to achieve the goals I have laid out for it. A more direct approach would be for Congress to (1) fund the IRS appropriately and (2) establish limits on forum choice that mitigate the risk of nationwide injunction shopping. The latter objective might be accomplished through a jurisdictional statute that allows pre-enforcement challenges to Treasury regulations exclusively in the U.S. District Court for the District of Columbia, with appeal to the D.C. Circuit. In contrast, a ban (with limited exceptions) on equitable actions against the IRS appears to be overbroad.

Yet in our far-from-first-best world, with a woefully underfunded IRS and few apparent limits on the ability of district courts to issue nationwide injunctions, the notion of narrowing the AIA so as to allow pre-enforcement judicial review of tax regulations seems to me like a risky gambit. Better, in my view, for courts to defer to Congress and for Congress to pair any amendment to the AIA with a boost in IRS funding and a forum provision like the one described above. To whittle down the AIA without simultaneously bolstering the IRS’s defenses would be to expose the tax authority to an onslaught that could overwhelm it.

There are commonalities between this argument and elements of Hickman and Kerska’s case for a narrower AIA. Although their title highlights their appeal to history, their analysis accounts for the more recent evolution of the federal tax system. But while they emphasize the expansiveness of the current Code, my focus is on the hollowed-out agency that has been tasked with interpreting and administering our tangle of tax statutes. To be sure, I cannot claim that my view of the AIA as a shock-absorber for the IRS is deeply rooted in the statute’s nineteenth-century history. But it is attentive to the reality of twenty-first-century tax administration.

Where does that leave us as to the AIA’s reach? In my view, the Supreme Court’s AIA case law supplies a serviceable test. First, a court considering a pre-enforcement challenge to a federal tax would ask whether the challenger would have a subsequent opportunity to litigate the claim on her own behalf in the Tax Court or in a refund suit. If no, the suit could proceed.[70] If yes, the AIA would bar relief unless the challenger could show both “irreparable injury” and “certainty of success on the merits.”[71] On this view, preemptive strikes against Treasury regulations would in most cases fail. They would succeed only when Congress has provided the challenger with no other avenue for redress or when the IRS’s action is both indefensible and irremediable.

In sum, the flurry of regulatory activity that we are likely to see soon from the IRS makes the question at the heart of Hickman and Kerska’s article—whether the AIA bars pre-enforcement judicial review of tax regulations—vitally important. But the same factor that makes their analysis so relevant should also give us pause regarding the solution that they propose—a solution that would place further stress on an IRS that already appears to be buckling under the burden of the new tax law.[72] Hopefully there will arrive a time in the not-too-distant future when resource constraints are less binding and process values can be vindicated. If at that point Congress revisits the statute, Hickman and Kerska’s “engagement test” strikes me as a viable template for an updated AIA. Again, timing is everything. And the time for the “engagement test” has yet to come.