The Label Test: Simplifying the Tax Injunction Act after NFIB v Sebelius

Introduction

In National Federation of Independent Business v Sebelius1 (“NFIB”), the Supreme Court maintained both its jurisdiction over the case and the constitutionality of the Affordable Care Act2 (ACA) by threading the needle between the Anti-Injunction Act3 (AIA) and Congress’s taxing power under the Constitution.4 The legal implications of the majority opinion, however, have yet to work themselves out in the field of tax law. The AIA and the Tax Injunction Act5 (TIA) protect federal and state taxes, respectively, from precollection injunctions in federal court. In other words, litigants must pay first and then challenge the taxes in court. Because the text of each act mentions only “taxes,” courts have interpreted the acts to mean that they do not protect other government charges. This means that courts must determine when a government charge is a “tax” and when it is something else. The main alternatives are “regulatory fees” (or simply “fees”) and “penalties,” and the problems of delineating these categories are the subject of Part II. In short, taxes are seen as being imposed broadly for the purpose of general revenue while penalties and fees are imposed more narrowly to incentivize certain behaviors or defray costs arising from regulations. Historically, courts deciding cases under the TIA have tended to minimize the importance of the statutory label applied to a charge when deciding how to categorize it,6 but the logic of NFIB in interpreting the AIA turns on the fact that Congress labeled the “individual mandate” a “penalty” rather than a “tax.”7 The majority construed this language as evidence of a purposeful decision by Congress not to take advantage of the protection against injunctions offered by its own creation, the AIA.8

The application of NFIB to future cases under the AIA should be straightforward: if Congress labels a charge a “penalty” or “fee,” then it is deemed to have intended that the AIA not protect that charge. On the other hand, the logic of NFIB, in which a single legislative body excepts one of its laws from another of its own laws, is one step removed from the TIA, which applies to laws enacted by state legislatures. But just as Congress can control the application of its own law, it can also give states the ability to control the application of federal law by “incorporating” state law.9 Despite the federal–state distinction between the AIA and the TIA, the text and effects of the statutes are substantially identical, and cases that treat the definitions and applications of the statutes are used as interchangeable precedent by default in current jurisprudence. Given this background, the logical extension of the Supreme Court’s decision in NFIB would be to look to a state-law label to define a charge under the TIA. This Comment argues that applying NFIB’s label test to the TIA is justifiable, based on the act’s legislative history and the practice of using the AIA in TIA cases, and beneficial, in that it offers a bright-line test for determining when an exaction is a “tax” for purposes of the TIA.

Currently, courts use a variety of tests to separate taxes from nontaxes. But federal courts struggle to make a consistent distinction, especially when faced with exactions that both raise revenue (like a tax) and penalize very specific behavior (like a penalty). For instance, circuits are split over whether “tax delinquency penalties,” applied to individuals who fail to pay their taxes, should be categorized as taxes or penalties under the TIA. But under NFIB, the state-law “penalty” label would be dispositive. The simplicity of this new test saves litigants from current uncertainty and federal judges from navigating a growing body of conflicted jurisprudence.

Part I of this Comment outlines the legislative history of the AIA and TIA, as well as the Supreme Court’s understanding of their related legislative purposes. Part II reviews the current inter- and intracircuit splits over how to tell when a government charge is a “tax” protected under the TIA. Part III examines two recent Supreme Court decisions, Direct Marketing Association v Brohl10 and NFIB, in order to evaluate possible solutions. While the legislative history and purpose behind the TIA are important, Brohl precludes a solution based on legislative purpose alone. On the other hand, NFIB offers the possibility of a much simpler test. Part III then offers an analysis of why this label test should be transplanted from NFIB, an AIA case, into TIA jurisprudence.

I. Background Law and Legislative History

This Part introduces the parallel statutes of the Anti-Injunction Act and Tax Injunction Act. Each statute protects federal and state taxes, respectively, from equitable interference by federal courts. Of particular importance is the legislative history behind the TIA, which documents a congressional concern for the integrity of state budgets. The Supreme Court has also expressed a concern about federal-court respect for the state-court system, drawn from the text of the TIA, which states that federal intervention is allowed only if a “speedy and efficient remedy” is not available in state court.

A. The Anti-Injunction Act: Protecting Federal Taxes

Article III of the Constitution grants Congress the power to create “inferior courts” in which to vest the judicial power of the United States.11 Congress, in turn, has granted the district courts original jurisdiction to hear cases arising under federal law12 and cases with diverse parties.13 But what Congress has given, Congress can take away.14 In 1867, Congress enacted the Anti-Injunction Act, barring federal courts from interfering in cases involving federal taxes before they are collected. The statute mandates that “no suit for the purpose of restraining the assessment or collection of any tax shall be maintained in any court by any person.”15 In other words, the AIA institutes a timeline in which taxpayers pay the tax first and then file suit to recover the money. Before then, federal courts cannot hear AIA cases.16

Legislative history on the AIA is sparse, as the Supreme Court has acknowledged: “The Anti-Injunction Act apparently has no recorded legislative history.”17 The text of the AIA was adopted, without debate or comment,18 as an amendment proposed by Senator William Pitt Fessenden to the Revenue Act of 1866.19 While newspapers recorded the amendment’s passage, no reports of the Senate Finance Committee or Conference Committee have ever been recovered, nor is there any mention in the personal correspondence of Fessenden.20 On the other hand, an 1895 treatise claimed that the amendment was a response to applications for injunctions against the newly instituted income tax.21 Regardless of its lack of legislative history, the Supreme Court has stated that “its language could scarcely be more explicit. . . . [T]he principal purpose of this language [is] the protection of the Government’s need to assess and collect taxes as expeditiously as possible with a minimum of pre-enforcement judicial interference.”22

B. The Tax Injunction Act: Protecting State Taxes

In 1937, Congress passed a derivative statute to the AIA to protect state tax revenues from federal injunctions. The wording of the TIA was similar, but not identical, to the wording of the AIA: “The 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.”23 According to the conference report accompanying the bill, the Senate Committee on the Judiciary was most concerned with stabilizing state and local government finances, especially in protracted diversity cases involving large foreign corporations.24 The committee claimed that state and local governments had been forced to settle suits for less than the amount of tax owed, not because the lawsuits were strong but because the government entities needed the money.25 The bill was also envisioned as a way of equalizing the “highly unfair picture” in which citizens of the state would have to “pay first and then litigate,” while diverse plaintiffs could withhold money until cases were ended.26 The bill passed without controversy.27

Various Supreme Court cases evaluating the scope of the TIA have emphasized not only its text but its spirit and underlying purpose, and the Court has even relied upon these extratextual characterizations in its TIA jurisprudence. The Court has stated that the text of the law “reflects a congressional concern to confine federal-court intervention in state government.”28 This congressional concern mirrors an equivalent judicial concern of respect for state courts. When using the TIA to block a challenge to Oklahoma taxes imposed on foreign-based motor carriers, the Court stated, “We have long recognized that principles of federalism and comity generally counsel that courts should adopt a hands-off approach with respect to state tax administration.”29 The principle of “comity” is shorthand for the Supreme Court’s presumption in favor of state-court jurisdiction when faced with a “conflict between the state’s interest in having the issues adjudicated in a state forum and the individual’s interest in a federal forum.”30 The Court stressed that “the principle of comity [ ] underlies [the TIA].”31 Furthermore, the Court has stated that it expressly relied on this extratextual purpose of the TIA to inform its decisions: “We subsequently relied upon the Act’s spirit to extend the prohibition from injunctions to declaratory judgments regarding the constitutionality of state taxes.”32

C. Litigation under the TIA

The Tax Injunction Act is relevant only for cases in which plaintiffs are challenging a state or local tax in federal court. First, plaintiffs would rely on a positive grant of federal jurisdiction to get into federal court. The vast majority of cases fall under federal question jurisdiction,33 in which taxpayers claim that a charge imposed on them violated some provision of the Constitution,34 was preempted by federal law,35 or even violated federal racketeering laws.36 If a plaintiff has not yet paid the tax, then they would be seeking equitable relief in the form of an injunction or declaratory judgment. The TIA forbids federal courts from dispensing such relief.37 If, however, the plaintiff can successfully argue that the state charge they are challenging is not a tax, then the suit can proceed in federal court.

States may raise the TIA as a defense against a suit for an injunction in federal court, but because the text of the TIA is addressed to the district courts (“The district courts shall not enjoin . . .”), it is also typically interpreted as a jurisdictional law.38 Therefore, courts must “investigate the application of the Tax Injunction Act regardless of whether the parties raise it as an issue.”39 Nor is the TIA subject to waiver by the state.40 That is, even if a state were happy to allow a federal court to proceed to the merits, the TIA would still bar the case.41

II. The Shortcomings of Current TIA Jurisprudence

Circuit courts have long struggled with the problem of how to handle the definition of “taxes” under the TIA, a complicated endeavor that produces inconsistent results. Part II.A presents the leading cases and tests that courts have used to mark the boundaries of taxes and nontax charges. Many courts continue to apply multifactor tests that attempt to capture various aspects of a “classic tax.” Judge Richard Posner, on the other hand, advocates for a simpler test, objectively comparing the cost of the regulation to the charge imposed. A third group of courts abandons any pretense of formal tests or facts, instead pursuing a more holistic inquiry into the legislative “purpose” behind the charge.

Penalties pose an especially difficult taxonomic problem, because they may raise general revenue but seem to have the main purpose of regulating specific behavior. Part II.B presents the circuit split over tax delinquency penalties as an example of current problems in TIA jurisprudence, in which similar laws are categorized in inconsistent ways.

A. Circuit Courts Have Utilized a Variety of Tests to Categorize Charges

Most of the cases in the following sections were decided in the last twenty-five years, despite the fact that the TIA was enacted in 1937. Cases before the 1990s are fewer and further between. As the following cases show, courts began to categorize government exactions as taxes or as regulatory fees or penalties after the 1990s, leading to significant growth in the case law. Generally, courts agree that “taxes” are levied for the purpose of raising general revenue. As such, courts experience difficulty when determining whether a government charge that only incidentally raises revenue is a “tax.”

1. Courts in early cases reduced the inquiry to a three-factor test.

An early, seminal case discussing the distinction between taxes and nontaxes under the TIA is the First Circuit’s 1992 decision in San Juan Cellular Telephone Co v Public Service Commission of Puerto Rico,42 which sketched out the conceptual relationship between taxes and fees. When the plaintiff telephone company challenged a 3 percent charge on private telephone providers, the Puerto Rican government responded that the charge was a tax and that an injunction was not allowed.43 In addition to TIA precedent on the distinction between taxes and nontax charges (of which there was relatively little at the time),44 the court drew on a swath of public utilities cases. The distinction between taxes and nontaxes is important in the public utilities context because government agencies have the power to regulate such utilities, which includes levying regulatory charges, but not imposing taxes.45

From the public utilities cases, the court imported a “spectrum” framework into the realm of the TIA, with “a paradigmatic tax at one end and a paradigmatic fee at the other.”46 According to the court, “The classic ‘tax’ is imposed by a legislature upon many, or all, citizens,” and is “spent for the benefit of the entire community.”47 This definition of “tax” emphasizes broad incidence and broad benefits. On the other hand, the court defined a “regulatory fee” as a charge imposed not by a legislature but by an individual agency on a narrower group of citizens, namely, “those subject to its regulation.”48 Such regulatory fees may be for “deliberately discouraging particular conduct by making it more expensive” or for defraying the cost of associated regulation, such as when the Nuclear Regulatory Commission charges companies for performing environmental reviews.49 Therefore, fees tend to be imposed on a narrower subset of people, for a narrower purpose than raising general revenue. Finally, the court noted that for close cases in the middle of the spectrum, the focus of the taxonomic inquiry is on the ultimate use of the collected money, whether it is used for a general benefit or simply for defraying the cost of a regulation.50

In Bidart Brothers v California Apple Commission,51 the Ninth Circuit attempted to reduce San Juan Cellular to a three-factor test.52 The court identified three factors to distinguish taxes from fees: (1) the entity that levies the charge, (2) the parties on whom it is levied, and (3) the ultimate use of the money.53 The charge is a “fee” if the entity imposing it is a regulatory agency, the burdened population is small, and the funds raised from the charge are used for the benefit of the people or organizations that are regulated. It is a “tax” if it is imposed by the state, on many, and for the benefit of many. Applying its test to the facts, the court found that charges imposed by the California Apple Commission (an agency) on apple producers (a narrow group) to pay for promoting apple sales (a narrow use) were fees, and thus not protected by the TIA.54 Therefore, the federal district court had jurisdiction over the apple producer’s suit for injunctive relief.55

2. Judge Posner promotes a single-factor “user fee” test.

Despite the apparent simplicity of Bidart’s three-factor test, which continues to be cited as good law, courts have resisted a uniform approach to categorizing charges under the TIA. Judge Posner, for instance, has been the strongest advocate of reducing the problem to a single factor: comparing the amount of the government charge to the cost of the government service provided or regulation applied to the charged party, though his approach differs from other Seventh Circuit decisions during the same time period.56 Posner’s single-factor test can be seen simply as a refinement of San Juan Cellular, which stated that “[c]ourts facing cases that lie near the middle of this spectrum have tended . . . to emphasize the revenue’s ultimate use, asking whether it provides a general benefit to the public . . . or defrays the agency’s costs of regulation.”57 In effect, Posner takes the third factor from the Bidart test, isolates it as the only important factor, and construes it mathematically. Thus, when “users of the public way” impose calculable costs on the government, the government can impose a “user fee” up to the amount of those costs without such charges being characterized as “taxes.”58

Posner first addressed the tax–nontax distinction in Diginet, Inc v Western Union ATS, Inc,59 although it was not a TIA case. Illinois law prohibits cities from taxing without permission from the state, but it allows them to regulate, with any assessments that might entail.60 Drawing exclusively on cases involving municipal regulation of public utilities, Posner wrote that the rule for distinguishing a tax from a “user fee” is whether “the fee is a reasonable estimate of the cost imposed by the person required to pay the fee.”61 In contrast, a “tax” generates revenue beyond the cost of the service or regulation and can be used to “offset unrelated costs.”62 Posner was quick to point out that this test does not depend on how the city labels the charge, so a city cannot “circumvent this limitation by calling a tax something else.”63

Under the facts of the case, Western Union had laid a fiber-optic network in underground ducts owned by the City of Chicago, after receiving permission from the Department of Public Works.64 When higher officials in the city government noticed the installation, they imposed a “franchise fee” of 3 percent of network revenues.65 Under Posner’s construction, such a charge could be characterized as a fee if it offset costs that the fiber-optic network imposed on the city, but Posner, speaking for the panel majority, found no evidence of any costs imposed on the city.66 It was a pure play for revenue. Therefore, the “franchise fee” was, “by this test, a tax, not a user fee.”67

In Empress Casino Joliet Corp v Balmoral Racing Club, Inc,68 Posner, writing for an en banc majority, applied his “user fee” test to the TIA specifically, but also acknowledged the “fuzzy” problem of penalties, which tend to function like taxes. Because the TIA forces litigants into an alternate forum—state courts rather than federal courts—Posner called for a “crisp rule distinguishing taxes from other exactions,” and he voiced his dismay with previous cases “flirt[ing] with open-ended, multifactor tests.”69 In contrast, Posner posited that there are three categories of government charges: penalties, used to punish rather than generate revenue; fees, which function as prices for government services or regulations; and taxes, a catchall category.70 If a charge is neither a fee nor a penalty, it must be a tax. Whether a charge is a fee is easily determined under the “user fee” test from Diginet.71 But distinguishing taxes from penalties is more difficult, as both raise revenue beyond any costs imposed on the government. While Posner admitted that “a tax might be so totally punitive in purpose and effect that, since nomenclature is unimportant, it should be classified as a [penalty]” instead of a tax, he did not discuss how to tell them apart.72 Finally, he maintained that how a legislature labels a charge should hold little weight because the word “‘[t]axation’ is unpopular these days,” and legislatures have other motives for their labels besides the TIA.73

3. Courts in more recent cases tend to focus on the legislative purpose behind the enactment.

In spite of Posner’s best efforts, courts in the Seventh Circuit and many others have failed to coalesce around a single standard of analysis. In general, more recent decisions tend to be concerned with divining the legislative purpose behind a charge’s enactment. For example, in interpreting San Juan Cellular, the Fourth Circuit stated that when a tax falls in the middle of the spectrum, “the most important factor becomes the purpose behind the statute.”74 While this construction relies on many of the same factors as traditional multifactor tests—and even Posner’s “user fee” test—this emphasis on legislative purpose is more opaque as to which factors courts will find to be the most salient.

Consider Hager v City of West Peoria,75 a Seventh Circuit case written by Judge Daniel Manion, in which the court considered heavy-truck permit fees. First, the court applied the “user fee” test, concluding that the truck fees would not bring in more revenue than would be necessary to repair the damage they did to roads.76 At this point, if the panel had included Posner, he likely would have declared the charge a “fee” and been done with it; the court did reach that conclusion, but not before considering more holistic evidence of “why the money [was] taken.”77 In particular, the court relied on testimony from the mayor that the city “intended to regulate” the street on which trucks were driving, as well as the fact that “[t]he stated purposes in [the text of the ordinance] are public safety and highway maintenance, not revenue collection.”78 Therefore, the court concluded, “We thus have no doubt that the ordinances were passed to control certain activities, not to raise revenues,” meaning that the charge was not a tax.79 These quotations show that what mattered most to the court was the legislative purpose behind the ordinance—why was it passed?

The Hager court’s focus on purpose led to the same conclusion that Posner’s “user fee” test or San Juan Cellular’s multifactor test might have (as the charge was imposed on a narrow group for a narrow purpose). The hardest cases, however, arise not when charges look like regulatory fees, but when they look like penalties. The primary purpose of a penalty is to control undesirable behavior. The primary purpose of a tax is to raise revenue. But when penalties and taxes each do both, the line between them can be very fuzzy. Under an inquiry into legislative purpose, if the point is to determine why a legislature imposed a particular charge, courts must choose between competing explanations of revenue raising and incentivizing.

For instance, in Kathrein v City of Evanston,80 the Seventh Circuit examined an assessment that the city had imposed on building demolition, which developers were using to clear room for new residential projects. Despite the charge being labeled a “tax,” the court held that it was a “regulatory device” because it created a very specific incentive structure for developers.81 The court’s logic focused on determining the underlying purpose of the law, asking whether it was to raise revenue as a tax or to disincentivize certain behavior as a penalty.82 The court acknowledged that some taxes on undesirable behavior are still taxes.83 Ultimately, however, the court found evidence of the law’s purpose in the low amount of overall revenue collected, the segregated fund into which the money was put, and the law’s ability to affect developers’ behavior by significantly decreasing the profitability of new developments.84 Therefore, the court found it was a penalty, unprotected by the TIA.

The Tenth Circuit has also faced the problem of penalties under the TIA. In Chamber of Commerce of the United States v Edmonson,85 the court confronted an Oklahoma law that required businesses to withhold taxes from independent contractors whose work eligibility was unverified, or else be liable for the amount they should have withheld.86 Like Kathrein, the court’s analysis focused on the “purpose” of the statute, but it also cited cases that line up perfectly with Posner’s “user fee” test.87 The court looked beyond the use of the money, however, and listed other cases in which the use of the money was less important than the “incentive structure and avowed purpose.”88 In other words, the opinion makes room for decisions like Kathrein, in which a statute looks more like a regulatory tool than a revenue generator even though the money is not used for a specific related purpose. And it is on that basis that the court decided that the Oklahoma statute was a penalty, because it provided a clear incentive structure not to hire contractors with unverified work eligibility, despite the fact that the penalty raised revenue for the general treasury.89

4. The Fifth Circuit focuses on the legislative history and purpose of the TIA.

Finally, the Fifth Circuit is unique in that it tends to emphasize the broadness of the TIA’s protection, which leads it to be more lenient in classifying charges as “taxes.” In Home Builders Association of Mississippi, Inc v City of Madison,90 the court considered a per-unit fee imposed on developers building new residential units in order to support expanding city services. The court cited all of the relevant precedent, including San Juan Cellular and Hager, boiling past cases down to the basic proposition that “the classic tax sustains the essential flow of revenue to the government, while the classic fee is linked to some regulatory scheme.”91 The court also drew on its own precedent in Tramel v Schrader92 for the proposition that “a broad construction of ‘tax’ was necessary to honor Congress’s goals in promulgating the Tax Injunction Act, including that of preventing federally-based delays in the collection of public revenues by state and local governments.”93 The Fifth Circuit, in other words, is sensitive to interruptions in state revenue. Drawing on information from the preamble of the ordinance at issue in the case, which indicated that the tax would be used to fund a variety of services—from street improvements to firefighting94 —the court found that the tax was not linked to any specific regulatory scheme and was instead intended to fund general improvements. Thus, it was protected by the TIA.95

B. Tax Delinquency Penalties Present a Clear Circuit Split

As certain as death and taxes is the fact that some people will not pay their taxes. In response, many state governments impose what are known as “tax delinquency penalties.” The previous Section noted that courts tend to struggle with penalties, which raise revenue like taxes but have a primary purpose of regulating behavior. Therefore, as one might predict, litigants who wish to challenge tax delinquency penalties face uncertainty under the TIA. A current circuit split on these laws provides an excellent example of why TIA jurisprudence is in need of reform.

The first court to address the issue of tax delinquency penalties was the Seventh Circuit in RTC Commercial Assets Trust v Phoenix Bond & Indemnity Co.96 This case yielded a sparse opinion in which the court decided that the purpose of tax delinquency penalties is to penalize, not raise revenue. Although the court cited intracircuit precedent in Diginet and Hager, it made no effort to apply the logic of either case, whether through the “user fee” test or a deeper inquiry into legislative purpose.97 Instead, the court summarily concluded that “[s]tates do not assess penalties for the purpose of raising revenue.”98 The reason given was that “[i]n a Utopian world where all citizens complied fully with their obligations, no penalties at all would be collected,” which suggests that penalties are not “calculated to generate revenues.”99 This may be an unrealistic view of penalties. For instance, it is academically confirmed (and probably generally assumed) that police departments count on a constant supply of traffic tickets to make their budgets.100

The Fifth Circuit took a similarly surface-level approach to the problem in Washington v Linebarger, Goggan, Blair, Pena & Sampson, LLP101 (“Washington I”), but it reached the opposite conclusion: the purpose of tax delinquency penalties is to raise general revenue, and therefore they are protected as taxes under the TIA.102 The court stated that tax delinquency penalties are “inexorably tied to the tax collection itself” and “sustain[ ] the essential flow of revenue to the government.”103 Therefore, the court found that the penalty was properly characterized as a tax for purposes of the TIA. This decision is consistent with the Fifth Circuit’s sensitivity to the broadness of the TIA in protecting state revenues, as discussed in the previous Section regarding Home Builders.104

The dissent, on the other hand, applied a multifactor test similar to the Bidart test: a charge is a “fee” if it is imposed “(1) by an agency, not the legislature, (2) upon those it regulates, not the community as a whole, and (3) for the purpose of defraying regulatory costs.” In this case, the first two factors cut in opposite directions: tax delinquency penalties are imposed by the legislature but on only a small group of people. For the dissent, the deciding factor was that the “purpose” of the charge was to “control the behavior of delinquent taxpayers.”105 Therefore, the dissent would have characterized this charge as a fee or penalty, not covered by the protection of the TIA.

The Fifth Circuit doubled down on its decision in Washington v New Orleans106 (“Washington II”). After her claim was dismissed from federal court, plaintiff Denise Washington successfully challenged Louisiana’s delinquent tax penalty in state court as unconstitutional under the Louisiana Constitution (proving, incidentally, that “a plain, speedy and efficient remedy may be had in the courts of such State”).107 She then refiled a putative class action, asking the federal court to reconsider its decision in light of the state court’s determination that it was an unconstitutional charge. The federal district court rejected her claim, reiterating its first decision that “the penalty was ‘inexorably tied’ to tax collection.”108 The court then resummarized its reasoning: “Because the penalty directly sought to ‘sustain the essential flow of revenue to the government,’ it falls within the broad scope of § 1341.”109

Most recently, the Ninth Circuit agreed with the Fifth Circuit that tax delinquency penalties are taxes in a short unpublished opinion in Huang v City of Los Angeles.110 Without stating its reasoning, the court concluded that “[a]pplying Bidart, the business taxes assessed by the City of Los Angeles, as well as the penalties added thereto for delinquent payment, are ‘taxes’ under the TIA.”111 One could, however, attempt to re-create the court’s logic under the Bidart three-part test. The entity that imposed the citywide tax was the city itself (indicative of a tax), the charge fell on only a small number of delinquent payers (fee), and the charge presumably funded general government activity rather than being dedicated to a specific program (tax). With two out of three factors favoring a tax, the court could then have decided that the TIA barred the case.

* * *

In summary, circuit courts are split on how tax delinquency penalties should be characterized for purposes of the TIA. The Seventh Circuit holds that tax delinquency penalties are not taxes, and therefore not subject to the TIA, while the Fifth and Ninth Circuits hold that they are. These cases prove that, under current TIA jurisprudence, courts are free to select one of a number of different approaches, each emphasizing different factors and all failing to provide a systematic way to distinguish taxes from fees and penalties. The next Part, however, offers a much simpler test.

III. Recent Supreme Court Cases Provide an Opportunity to Reexamine Current Law

The current state of the tax–nontax distinction under the TIA is so fragmented in part because the Supreme Court has never addressed the question of which test or factors should be used to determine whether a government charge is a tax. Within the last decade, however, the Supreme Court has decided two cases with potentially significant consequences for the TIA: NFIB v Sebelius in 2012 and Direct Marketing Association v Brohl in 2015.

Part III.A introduces Brohl, in which the Supreme Court rejected an expansive reading of the text of the TIA based on its historical purposes in favor of a highly textual interpretation. While Brohl does not deal with the definition of “tax” specifically, the decision tends to foreclose any solution relying exclusively on legislative history favoring state protection. On the other hand, Part III.B discusses NFIB, a case under the AIA, which endorses a construction under which a charge qualifies as a “tax” if it is labeled as such. Part III.C makes the legal case under current Supreme Court precedent for why the new AIA label test should be applied to the TIA, and Part III.D discusses its practical effects, including the simplicity that the label test would bring to courts and litigants and the political pressures that could keep states from abusing newfound power.

A. Brohl Precludes an Isolated Appeal to Legislative Purpose

One possible solution to the problems of defining “tax” under the TIA would be to follow the lead of the Fifth Circuit in Home Builders, Washington I, and Washington II. In each case, the Fifth Circuit looked to the legislative purpose and history of the TIA to arrive at a broad definition of “tax.” According to that court, the TIA ought to be interpreted broadly in order to fulfill Congress’s intent to avoid federal interference in state and local revenue collection.112 The Fifth Circuit also highlighted the Supreme Court’s previous statements about the Act, stating that the “statutory text should be interpreted to advance its purpose of ‘confin[ing] federal-court intervention in state government.’”113 The Fifth Circuit’s approach, at its strongest, deems the vast majority of charges to be “taxes” for purposes of the TIA. There may still be penalties so punitive and rare, or fees so tiny and incidental, that they might fall outside such a broad definition of “tax.” Regardless, such an approach—legitimately rooted in the legislative purpose of the TIA—could simplify the majority of TIA cases.

Brohl, however, shows the weakness of this approach. In Brohl, the Court addressed a different part of the statute, dealing not with the definition of “tax” but of “assessment, levy, or collection.” Without even mentioning legislative history, the Court reached its decision by analogizing to other sections of the tax code. The Colorado statute at issue required out-of-state sellers, who are not required to collect state sales taxes, to provide a notice to buyers that buyers are liable for Colorado use taxes and to provide names to the state of all buyers who purchased more than $500 worth of goods.114 Retailers were subject to a $10 fine for each buyer not notified and each buyer not reported to the state.115 The Direct Marketing Association (DMA), a trade association for out-of-state retailers, challenged the notice requirements in federal court under the Commerce Clause.116

When the Tenth Circuit first heard Brohl, the parties did not anticipate that the TIA would be an issue. The court, however, raised its own inquiry of the TIA sua sponte and decided that DMA was asking it to “enjoin, suspend or restrain the assessment, levy or collection” of Colorado taxes.117 To reach that conclusion, the Tenth Circuit cited the legislative purpose of the act, much like the Fifth Circuit’s approach, recognizing the TIA’s “roots in equity practice, in principles of federalism, and in recognition of the imperative need of a State to administer its own fiscal operations.”118 Based on these principles, the Tenth Circuit gave the word “restrain” a broad enough reading to include Colorado’s notice requirements and their accompanying charges, barring the suit in federal court under the TIA.119

The Supreme Court granted certiorari to reverse the Tenth Circuit, finding that the suit was not barred by the TIA because the Colorado notice and reporting requirements were not an “assessment, levy, or collection.”120 Rather than basing its decision on the history or purpose of the Act, the Court looked to the federal tax code for statutes that would help define “assessment, levy, or collection.” The Court found that the gathering of information, which included the private reporting of tax information to the government, was a phase that happened before “assessment,” which was the “official recording of a taxpayer’s liability.”121 Therefore, plaintiffs were free to challenge the charge in federal court. Brohl does not address the definition of “tax” under the TIA, but it does provide a recent look at how the Supreme Court approaches the TIA. While the legislative history and purpose continue to be important considerations in TIA jurisprudence, this decision helps to show that invocations of purpose alone are not a strong enough approach to redefine “tax” for purposes of the TIA.

B. NFIB Provides a Clear Test Based on Labels

In one of the landmark Supreme Court cases of recent memory, the NFIB Court upheld the constitutionality of the ACA without resorting to the Commerce Power.122 In an opinion written by Chief Justice John Roberts, the Court held that the “individual mandate,” a charge imposed by the statute, was a “tax” for purposes of the Constitution, and therefore constitutional under the taxing power, but not a “tax” for purposes of the AIA because Congress had labeled it a “penalty.” Therefore, the suit was not barred by the AIA.

In order to keep health insurance costs down by ensuring large insurance pools, Congress included the individual mandate,123 a charge imposed on citizens who choose not to purchase health insurance, which Congress labeled a “penalty.”124 Plaintiffs claimed that Congress did not have the power under the Constitution to enact the individual mandate. The Court held that the charge was a proper exercise of Congress’s power to tax under the Taxing and Spending Clause125 rather than an exercise of Congress’s power to regulate interstate commerce.126 But the individual mandate had yet to take effect, and because it was upheld as an exercise of the taxing power, the suit might have been barred by the AIA’s prohibition on “restraining the assessment or collection of any tax.”127

Several circuit courts heard cases on the ACA, but none held that the individual mandate was an exercise of the taxing power while also avoiding the jurisdictional bar of the AIA. The litigation also caused an intercircuit squabble over whether labels should govern the operation of the AIA, which was a standard that the courts had never endorsed, at least explicitly. The Sixth Circuit weighed in first in Thomas More Law Center v Obama,128 holding that the individual mandate was a “penalty” because Congress had labeled it as such: “While the Anti-Injunction Act applies only to ‘tax[es],’ Congress called the shared-responsibility payment a ‘penalty.’”129 The Sixth Circuit then upheld the constitutionality of the ACA under the Commerce Clause.130 In contrast, the Fourth Circuit found in Liberty University, Inc v Geithner131 that the individual mandate was a “tax,” such that the case was barred by the AIA.132 The Fourth Circuit then singled out the Sixth Circuit: “[N]o federal appellate court, except the Sixth Circuit in Thomas More, has ever held that the label affixed to an exaction controls, or is even relevant to, the applicability of the AIA.”133 The DC Circuit, however, rode to the Sixth Circuit’s rescue in Seven-Sky v Holder,134 in which the court held that the label “penalty” was a “deliberate” choice by Congress.135 The court argued that while labels may not have been an explicit standard, they explained past AIA jurisprudence: “[A]side from the Fourth Circuit’s recent decision, no court has ever held that ‘any tax’ under the Anti-Injunction Act includes exactions that Congress deliberately called ‘penalties.’”136

The Supreme Court settled the issue with two separate conclusions. First, the Court held that the test for “tax” under the taxing power need not be the same as the test for “tax” under the AIA. Congress cannot change whether something is a tax for constitutional purposes. For instance, Congress might hypothetically impose a criminal penalty on a defendant who had been acquitted in court, but call it a “tax” in an attempt to circumvent the Double Jeopardy Clause.137 Such a label would obviously not protect the penalty from a constitutional challenge. Second, the Court held that labels were important tools for guiding the application of the AIA. The majority opinion noted that both the AIA and the ACA are “creatures of Congress’s own creation.”138 Therefore, if Congress wants to exempt the ACA from application of the AIA, it can, and in this case it did, implicitly, by labeling the individual mandate a “penalty.”139 This result does not mean that Congress is always bound by its labels. Congress could have been more explicit about its intent by, for example, including the following in the statute: “[T]his penalty is a ‘tax’ for purposes of the AIA.” However, Congress did no such thing, so the Court determined that the label was determinative of whether Congress wanted the AIA to apply.

C. The NFIB Label Test Should Be Applied to the TIA

NFIB stands for the proposition that courts must look at the label applied to a government charge to determine whether Congress meant for that charge to qualify as a “tax” under the AIA. If applied to the TIA, this label test would dramatically simplify the tax–nontax inquiry, for the benefit of both potential litigants and judicial efficiency. If the state statute labels a charge a “tax,” then that tax would receive the protection of the TIA. This Section explores several reasons why NFIB’s label test ought to be applied to the TIA despite the fact that NFIB is an AIA case. Most importantly, it argues that federal courts have a long-established practice of using AIA as persuasive precedent in TIA cases. Finally, it argues that such a reinterpretation of the TIA is supported by its history and purpose.

The reason that courts do not already look to state law to define “tax” is because the TIA is a federal law. This line of reasoning goes back to the Fifth Circuit decision in Tramel, one of the earliest disputes over the word “tax” under the TIA. In Tramel, the plaintiffs argued that Texas state courts had determined that the charges at issue were not taxes and thus that their suit should remain in federal court. The Fifth Circuit stated that “[t]he proper question is not what the Texas courts have said the Texas legislature meant when it used the term [tax] but what Congress meant when it used the term.”140 Since Tramel, circuit courts have repeated the proposition that the word “tax” is governed by federal law.141

That is a reasonable reading of the TIA, but not a necessary one. Just as Congress can choose how its own laws interact, it could choose to place discretion over the interaction of state and federal law in the hands of states by “incorporating” state-law labels.142 This Comment argues that such a reading of the TIA is supported by the text of the statute, its legislative history, and its purpose as explained by the Supreme Court. The impetus for reinterpreting the TIA is NFIB itself. By articulating a label test for the AIA against a background of interpreting the AIA and TIA in unison, the Supreme Court has implicitly endorsed referring to state law to define “tax” under the TIA.

1. AIA and TIA precedent are used interchangeably.

Due to the parallel natures of the TIA and the AIA, there is a substantial amount of crossover in their interpretation—that is, courts tend to regard a decision about one act as persuasive precedent, at the least, for a case under the other. The most explicit endorsement of this interpretive method comes from Brohl, in which the Supreme Court stated:

In defining the terms of the TIA, we have looked to federal tax law as a guide. Although the TIA does not concern federal taxes, it was modeled on the Anti-Injunction Act (AIA), which does. . . . We assume that words used in both Acts are generally used in the same way, and we discern the meaning of the terms in the AIA by reference to the broader Tax Code.143

Relying on Brohl, the Ninth Circuit has declared that “the Court construes the two Acts in tandem.”144 Brohl endorsed the practice of using one act to interpret the other as it appeared in Hibbs v Winn,145 a TIA case in which Justice Anthony Kennedy wrote a dissent that looked to the AIA for aid in defining “assessment.”146 The current also runs the other way, from TIA to AIA. Because the older AIA has no legislative history, the Supreme Court has quoted the Senate report on the TIA in order to state the purpose of the AIA.147

Lower courts have followed suit. For example, the Ninth Circuit cited an AIA case to support its conclusion that an injunction against paycheck withholding is equivalent to an injunction against tax collection itself under the TIA.148 Similarly, the Eastern District of Kentucky used Winn, a TIA case, to support its decision that a requested injunction against allegedly unconstitutional tax credits was not barred by the AIA.149 Other examples come from the DC Circuit,150 the Southern District of Alabama,151 and the District of New Jersey.152 Using AIA and TIA precedent interchangeably is pedestrian enough that courts sometimes fail to mention that they are citing precedent arising under separate statutes, or explicitly state that they will not be distinguishing the cases.153 Furthermore, this practice of reading the statutes together even extends to a third statute, the Butler Act,154 which accomplishes the same function in the jurisdiction of Puerto Rico. The most obvious case is San Juan Cellular, a Butler Act case that courts interpreting the TIA have cited repeatedly.155 This is by no means, however, an isolated incident.156

The practice of interpreting the TIA and AIA in tandem follows from their similar language and purposes, and some opinions are explicit about these similarities. The Supreme Court itself has stated: “The Tax Injunction Act was thus shaped by state and federal provisions [such as the AIA] barring anticipatory actions by taxpayers to stop the tax collector from initiating collection proceedings.”157 In using bankruptcy decisions under the AIA to decide a TIA case, the Southern District of New York wrote, “the two statutes should be interpreted in a harmonious manner.”158 Because the acts are to be interpreted in tandem, differences in their texts have been used to justify contrasting interpretations. In Hobby Lobby Stores, Inc v Sebelius,159 then-Judge Neil Gorsuch, in a concurrence, compared the statutes, finding that differences in their texts justified a jurisdictional reading of the TIA (“The district courts shall not enjoin”) in contrast to a nonjurisdictional reading of the AIA (“no suit . . . shall be maintained”).160 While differences in the statutes do not affect the analysis in this Comment, due to their nearly identical relevant provisions, this contrasting interpretation proves the rule that, in general, the acts mean the same thing. Furthermore, Gorsuch stated that, if anything, the difference was evidence that the TIA provided even greater protection to state tax laws than the AIA did to federal laws.161

2. The label test is consistent with the text and purpose of the TIA.

The Supreme Court has endorsed interpreting the AIA and TIA in a parallel manner, and the previous Section argued that, under this default rule, NFIB’s label test should apply to the TIA. One could argue that the label test makes more sense as applied to the AIA than the TIA, because Congress can control labels elsewhere in federal law, but it can’t control state law.162 This Section lays out additional reasons why the default rule should still control. The label test would allow states to take shelter under the TIA by choosing whether to label charges as “taxes.” This broad protection is consistent with the text, legislative history, and underlying purpose of the act as articulated by the Supreme Court.

a) The text of the TIA.  First, the text of the TIA is consistent with looking to state statutes to decide which charges are “taxes” for purposes of the Act. The Act states that “[t]he district courts shall not enjoin, suspend or restrain the assessment, levy or collection of any tax under State law.”163 The phrase “under State law” is ambiguous. It certainly means that the law protects state taxes, as opposed to federal taxes. But it could also mean, more specifically, that the law applies to taxes labeled as such “under state law.”

A plain-text reading of the statute does not favor one interpretation over the other. At the very least, allowing state-law labels to determine which charges are protected by the TIA is not in conflict with the text. Furthermore, a construction of the statute that looks to state statutes to define “tax” is consistent with how Congress has used “under state law” previously. Consider, for instance, 5 USC § 8478, in which an exception to Federal Deposit Insurance Corporation insurance is available only “if the bank or institution meets bonding requirements under State law which the Secretary of Labor determines are at least equivalent to those imposed on banks by Federal law.”164 This passage means that the bonding requirements are not only derived from state law but also defined by state law, just like “tax” would be if the label test were applied to the TIA.

b) Legislative history and purpose.  Congressional reports accompanying the TIA, as well as the text of the act, make it clear that the legislative history and purpose of the statute is consistent with broader protections for state tax laws. Although Brohl ruled out an approach that relies exclusively on these considerations, they still provide good support for the new label test. Part I of this Comment documented Congress’s concern for state revenues at the time that it passed the Act. The Senate report accompanying the bill noted that many states already had statutes mandating that taxpayers in state courts “pay first and then litigate.”165 The law was proposed for two reasons: because the idea that persons and corporations outside the state could sue in diversity without paying first was “highly unfair” and because the ability of foreigners, especially large foreign corporations, to withhold money during lengthy litigation would “seriously disrupt State and county finances.”166

Adopting the label test would put more power in the hands of states to take advantage of the TIA. These cases inevitably pit the state interest in tax collection against the taxpayer’s interest of litigating in federal court before payment. But the Senate report comprehended this battle of rights in enacting the statute and concluded that taxpayers were already well protected: “It should be emphasized that the bill does not take away any equitable right of the taxpayer or deprive him of his day in court.”167 Specifically, the report noted that the text of the act allowed for federal jurisdiction in the absence of a “plain, speedy, and efficient remedy” in state court.168 If a taxpayer does not like the results in state court, “[a]n appeal to the Supreme Court of the United States is available as in other cases.”169

To Congress’s concern for the state and local fisc, the Supreme Court has added a strong concern for judicial federalism and comity: “[T]he statute has its roots in equity practice, in principles of federalism, and in recognition of the imperative need of a State to administer its own fiscal operations.”170 For this reason, Supreme Court precedent on this topic emphasizes the broad protection that the TIA is supposed to afford. The Court declared in Rosewell v LaSalle National Bank171 that “this legislation was first and foremost a vehicle to limit drastically federal district court jurisdiction to interfere with so important a local concern as the collection of taxes.”172 More recently, the Court reaffirmed its characterization in Rosewell by stressing its commitment to federalism: “The federal balance is well served when the several States define and elaborate their own laws through their own courts and administrative processes and without undue interference from the Federal Judiciary.”173 This understanding continues to govern federal courts; as the Sixth Circuit has written, “The TIA’s prohibition of federal court restraint of state or local taxation encompasses a broad array of state and local taxes.”174 The current crop of cases, under which states are forced to shoehorn their collections into a judge-made definition of “tax,” is inconsistent with the “drastic” limitation that the TIA was meant to impose on federal interference with state tax collections. NFIB, on the other hand, would allow states to take consistent advantage of the protection that Congress afforded them under the TIA.

D. Implementing the Label Test

This final Section discusses how the label test could simplify the work of litigants and courts in practice, as well as how the position of states and litigants would change under the new regime. While it is true that states would gain wide discretion to take advantage of the TIA, litigants would be freed from much uncertainty and would retain postcollection options to challenge state charges. Furthermore, state constitutional limits on taxation and political pressures would limit states’ abuse of the TIA.

1. A dramatically simpler test for courts and litigants.

In general, implementing the label test under the TIA should be simple. If the state or local statute labels the exaction a tax, it is a tax. That simplicity is its main benefit. Take, for example, the hallmark case of Bidart, in which the Ninth Circuit devoted more than a dozen paragraphs and over three thousand words to develop and apply a multifactor test.175 The court finally concluded that a one-quarter-cent charge per pound of apples was a fee, due to its origin from an agency, the narrowness of its incidence, and its purpose of promoting the apple industry.176 The new label test under NFIB would have been limited to the following query: How does the statute refer to the charge? The statute refers to the charge as an “assessment”177 and as a “fee,”178 but never as a tax. Therefore, the TIA would not bar the case from federal court. Such a simple test would not only aid judicial efficiency but also alleviate uncertainty for litigants looking to challenge state charges.

The label test is even simpler than Judge Posner’s “user fee” test, on two counts. First, the “user fee” test requires some estimate of the cost of the regulation, in order to compare it to the fee charged to the regulated person or firm. The cases specifically addressed by Posner were cases in which the cost of the regulation was zero, making it easy to decide that the charge was a tax.179 But in Hager, the city imposed a charge of $12.50 on trucks over eight tons and $20.00 on trucks over fifteen tons.180 A court would need expert testimony about how much damage heavy-truck traffic does to roads in order to see what costs it imposes on the city.181 In this case, the district court accepted that the charges “could conceivably exceed the costs imposed by heavy truck traffic,”182 while the Seventh Circuit concluded that the charges “could not exceed the amount necessary to pay for the road repair.”183 The label test, on the other hand, does away with the necessity of quantifying such costs as road damage. Because the charge was labeled a “permit fee,” this case would not be barred by the TIA.

The other difficulty with the “user fee” test is that it cannot account for the category of penalties. Posner admits that a charge “might be so totally punitive in purpose and effect that, since nomenclature is unimportant, it should be classified as a fine,”184 but he does not discuss where to draw the line. Penalties flummox courts because they raise revenue like a tax, but have a primary purpose of modifying behavior. Take, for instance, the current circuit split over state tax delinquency penalties, discussed in Part II.C. Under the label test, these charges would not be covered by the TIA. This conclusion might only take one paragraph of an opinion. If states felt that certain claims belonged in state court, they could modify the label attached to their law.

Litigants would also benefit from the label test’s mitigation of uncertainty. Take, for instance, the class action plaintiffs in Washington I and Washington II. These plaintiffs had to appeal to the Fifth Circuit twice in order to challenge the status of New Orleans’s tax delinquency penalty (or “tax collection penalty”) under the TIA. They lost in both cases, even though they garnered a vigorous dissent in the former.185 Under the label test they would have won, and the simplicity of the test would have ensured a high measure of certainty on the matter.

In summary, this cuts through the mass of tests and factors, which become irrelevant under the new approach. To borrow Posner’s memorable phrasing, courts can stop relying on precedent that forces them to ask “how close a ‘family resemblance’ the exaction bears to an exaction acknowledged by all to be a ‘tax.’ Is it a brother, or a third cousin?”186

2. States gain flexibility; taxpayers remain protected.

Functionally, the label test means that state and local governments can protect their revenue streams from federal injunctions by passing statutes with an appropriate label. This protection is consistent with the concerns of federalism and comity outlined in Parts I and III.C.2. But one might worry that states will abuse this function. States could label everything a tax, or they could attempt to evade federal review of certain issues by applying the labels to nontraditional categories, like a “tax” on protest rallies. This sort of abuse would run up against state and federal constitutional limitations and a well-developed literature on “tax aversion”—the phenomenon of disproportionate voter resistance to paying charges labeled “taxes.”

1. States and local governments gain flexibility.  The label test is an undeniable victory for states, and it provides an even greater benefit to local governments. The bulk of state taxes may already be protected under the TIA because the largest taxes, such as income and use taxes, are already labeled as such.187 This fact helps ensure that immediate adoption of the label test would not be disruptive. But the label test does protect states’ ability to proactively decide to collect taxes free from federal-court interference, should the state government determine that is necessary in particular instances.

   The label test is of bigger aid to cities, counties, and other local government entities, some of which may depend less on traditional “taxes.”188 For instance, in Home Builders, the City of Madison, Mississippi, charged a $700 “municipal impact fee” for each dwelling constructed by developers.189 It used the money to fund “essential municipal services and facilities in the rapidly-growing city,” including street improvements and police and fire departments.190 Fortunately for the city, its case was before the Fifth Circuit, but a stricter circuit court might have found that the “impact fee” was indeed a “fee” because it was imposed on a very narrow class of individuals (developers) in order to cope with the cost of those individuals’ activities (developing new residences). A federal injunction, especially when tied to long-running litigation, could have crippled the expansion of Madison’s essential services, contrary to the legislative intentions of the TIA. Under the label test, Madison would have the power to take advantage of the TIA’s protection by labeling its ordinance a tax.

2. Statutory labels are the only route sanctioned under NFIB.  The Supreme Court’s decision in NFIB makes clear that Congress could “describe something as a penalty but direct that it nonetheless be treated as a tax for purposes of the Anti-Injunction Act.”191 The states, however, would have no such power to simply designate when the TIA should apply. As the earlier discussion of Tramel made clear, the TIA is still a federal statute. Therefore, the federal government has the exclusive power to decide how state law will affect the application of federal law. In NFIB, the Supreme Court explicitly sanctioned “statutory text” as the means by which to determine whether the law should apply. Since Tramel, state labels have been seen as “relevant but not dispositive.”192 Now, they are dispositive. State courts may continue to determine whether charges are taxes for the purpose of state law,193 but nothing in NFIB gives any reason to think that these state-court decisions are of any consequence for determining what is a “tax” under the TIA, just as Congress, and not federal courts, determine which taxes ought to be protected by the AIA.
3. Anti-tax political sentiments restrain abuse.  It is true that, over time, state and local governments could use the label test to narrow the charges that could be challenged before payment in federal court. The likelihood of state abuse is mitigated, however, by the exception clause in the TIA itself and, perhaps more importantly, the bad political optics of slapping the label “tax” on everything.

   First, it is worth noting that the Act itself is a policy decision by Congress to stop federal courts from interfering with state monetary collection. As the Senate report remarks, the TIA does not “deprive [the taxpayer] of his day in court.”194 The law blocks federal intervention only before the taxes are collected. Furthermore, the law itself contains a safeguard exception under which parties are blocked from federal court only “where a plain, speedy and efficient remedy may be had in the courts of such State.”195 This exception would take effect if citizens challenged the taxes in state courts, and the state-court system mishandled or buried their claims. In that case, litigants would again be able to turn to federal courts for an injunction against the collection of future taxes, until the state sorted the claims out.

   Furthermore, state and local governments will be restrained from making everything a statutory “tax” because of the political reality that citizens find taxes to be “more painful than other categories of expenses and losses.”196 There is an “extensive literature” documenting this phenomenon, termed “tax aversion.”197 In one study, researchers conducted a survey in which they asked participants about how they thought beneficiaries should pay for certain services, such as education, mail delivery, or trash collection.198 In short, they found that “labels mattered.”199 Subjects rated tax payment schemes as less desirable than user fees, even in situations in which “the economics were identical.”200 Subjects were especially affected by how payment schemes already were in their jurisdictions, accepting taxes that were already being imposed on them, but preferring new fees to new taxes when no taxes were in place.201 This “‘no new taxes’ bias,” as the authors nickname it,202 means that state and local governments would suffer political penalties not only for passing new penalties, but also for relabeling fees as taxes by statute.

   Not only that, but taxpayers may spend more time and money on tax avoidance than pure economic rationality would predict.203 In another study, subjects stated that they were willing to travel thirty minutes to a different jurisdiction to avoid 8 percent in sales taxes at higher rates than those who would travel to get a 9 percent discount.204 The word “tax” simply has irrationally negative implications in the minds of citizens. This means that states will change the label to “tax” only when the fiscal protection of the TIA outweighs the political ramifications. These are exactly the cases in which one might hope that state and local treasuries do receive protection.

Conclusion

In terms of applying the Tax Injunction Act, the real benefit of the NFIB label test accrues not to the states, but to federal judges who have to apply the TIA and to litigants whose cases it may affect. NFIB is clear in its statement that labels matter. Applying that decision to the TIA dramatically simplifies the test for determining whether a government charge is a tax for purposes of the act, and it also allows states to take advantage of the protection of the TIA by labeling charges as taxes in their statutes. This construction of the TIA upholds the historical reasons for enacting the law—namely, “to limit drastically federal district court jurisdiction to interfere with so important a local concern as the collection of taxes.”205 Any temptation a state may have to label every charge a tax is fortunately counteracted by the political consequences of imposing “taxes,” leaving the law in balance.

• 1567 US 519 (2012) (“NFIB”).
• 2Patient Protection and Affordable Care Act, Pub L No 111-148, 124 Stat 119 (2010).
• 3Act of Mar 2, 1867 § 10, 14 Stat 471, 475; Internal Revenue Code of 1954 § 7421, 68A Stat 1, 876, codified at 26 USC § 7421. This statute is sometimes also referred to as the Tax Anti-Injunction Act, but the majority of courts refer to it as the Anti-Injunction Act. The AIA should not be confused with the statute of the same name, enacted as part of the Judiciary Act of 1789 § 5, 1 Stat 333, 334–35, codified at 28 USC § 2283 (preventing federal courts from issuing injunctions against ongoing proceedings in state courts).
• 4See US Const Art I, § 8, cl 1 (“The Congress shall have power to lay and collect taxes, duties, imposts and excises.”).
• 550 Stat 738 (1937), codified at 28 USC § 1341.
• 6See Part I.B.
• 7See NFIB, 567 US at 543–46.
• 8Id.
• 9See note 142.
• 10135 S Ct 1124 (2015).
• 11US Const Art III, § 1.
• 1228 USC § 1331.
• 1328 USC § 1332.
• 14For discussion of the extent of such power, see generally Richard H. Fallon Jr, Jurisdiction-Stripping Reconsidered, 96 Va L Rev 1043 (2010).
• 15Act of Mar 2, 1867 § 10, 14 Stat at 475.
• 16For a discussion of the jurisdictional nature of the TIA and AIA, see note 41.
• 17Bob Jones University v Simon, 416 US 725, 736 (1974).
• 18See id, citing Note, Enjoining the Assessment and Collection of Federal Taxes despite Statutory Prohibition, 49 Harv L Rev 109, 109 & n 9 (1935).
• 19Act of Mar 2, 1867 § 10, 14 Stat at 475, amending the Revenue Act of 1866 § 19, 14 Stat 98, 152.
• 20Note, 49 Harv L Rev at 109 & n 9 (cited in note 17).
• 21Roger Foster and Everett V. Abbot, A Treatise on the Federal Income Tax under the Act of 1894 231 (1895) (“When the income tax was first imposed during the civil war, a number of applications were made for injunctions against its assessment or collection. To prevent this practice, Congress [passed the AIA].”).
• 22Bob Jones, 416 US at 736.
• 23TIA § 1, 50 Stat at 738, 28 USC § 1341.
• 24Amending the Judicial Code, S Rep No 75-1035, 75th Cong, 1st Sess 2 (1937).
• 25Id (“The pressing needs of these States for this tax money is so great that in many instances they have been compelled to compromise these suits, as a result of which substantial portions of the tax have been lost.”).
• 26Id.
• 27See S 1551, 75th Cong, 1st Sess, in 81 Cong Rec 8412 (Aug 6, 1937); S 1551 75th Cong, 1st Sess, in 81 Cong Rec 9008 (Aug 16, 1937).
• 28Arkansas v Farm Credit Services of Central Arkansas, 520 US 821, 826–27 (1997).
• 29National Private Truck Council, Inc v Oklahoma Tax Commission, 515 US 582, 586 (1995).
• 30Michael Wells, The Role of Comity in the Law of Federal Courts, 60 NC L Rev 59, 60 (1981).
• 31Franchise Tax Board of California v Alcan Aluminium Ltd, 493 US 331, 333 (1990) (emphasis added). See also Blatchford v Native Village of Noatak, 501 US 775, 784–85 (1991) (holding that the TIA is a “matter of comity”); Burlington Northern Railroad Co v Oklahoma Tax Commission, 481 US 454, 464 (1987) (“These are policy considerations which may have weighed heavily with legislators who considered the Act and its predecessors. It should go without saying that we are not free to reconsider them now.”); Fair Assessment in Real Estate Association, Inc v McNary, 454 US 100, 103 (1981) (“This legislation, and the decisions of this Court which preceded it, reflect the fundamental principle of comity between federal courts and state governments that is essential to ‘Our Federalism,’ particularly in the area of state taxation.”).
• 32National Private Truck, 515 US at 586 (emphasis added).
• 33See 28 USC § 1331 (“The district courts shall have original jurisdiction of all civil actions arising under the Constitution, laws, or treaties of the United States.”).
• 34See, for example, Kathrein v City of Evanston, 636 F3d 906, 910 (7th Cir 2011).
• 35See, for example, Chamber of Commerce of the United States v Edmondson, 594 F3d 742, 750 (10th Cir 2010).
• 36See, for example, Empress Casino Joliet Corp v Balmoral Racing Club, Inc, 651 F3d 722, 724–25 (7th Cir 2011) (en banc).
• 3728 USC § 1341.
• 38See California v Grace Brethren Church, 457 US 393, 417 n 38 (1982) (describing the TIA as a “jurisdictional bar”).
• 39Folio v City of Clarksburg, 134 F3d 1211, 1214 (4th Cir 1998).
• 40Trailer Marine Transport Corp v Rivera Vazquez, 977 F2d 1, 5 (1st Cir 1992) (“If this Court were nonetheless to reach the First Amendment issues presented in these appeals, the litigants would have sidestepped neatly Congress’ intent and our longstanding policy ‘to limit drastically’ federal interference in the administration of state taxes when a ‘plain, speedy and efficient’ state remedy is available.”), citing Grace Brethren Church, 457 US at 418–19.
• 41For a more in-depth discussion of the jurisdictional nature of the TIA, see generally Peter D. Enrich, Federal Courts and State Taxes: Some Jurisdictional Issues, with Special Attention to the Tax Injunction Act, 65 Tax Law 731 (2012). According to conventional wisdom, the AIA is also jurisdictional in nature, despite not being addressed to the district courts, but at least one scholar has argued that conventional wisdom is wrong. See generally Erin Morrow Hawley, The Equitable Anti-Injunction Act, 90 Notre Dame L Rev 81 (2014).
• 42967 F2d 683 (1st Cir 1992). This case actually concerns the Butler Act, which is Puerto Rico’s version of the TIA, but courts continue to cite it as if it were a TIA case. See, for example, Empress Casino, 651 F3d at 730; Bidart Brothers v California Apple Commission, 73 F3d 925, 930 (9th Cir 1996).
• 43San Juan Cellular, 967 F2d at 684–85.
• 44The first circuit-court case to examine the distinction between taxes and fees under the TIA was Tramel v Schrader, 505 F2d 1310, 1316 (5th Cir 1975) (finding city street-improvement assessments to be a “tax” falling under the TIA). For other circuit-court cases decided before San Juan Cellular, see generally Miami Herald Publishing Co v City of Hallandale, 734 F2d 666 (11th Cir 1984) (severing a city newspaper-licensing law’s revenue-raising components, which the TIA sheltered from federal-court challenge, from its regulatory components in order to consider constitutional objections to the regulatory provisions); Schneider Transport, Inc v Cattanach, 657 F2d 128 (7th Cir 1981) (ruling that state truck-registration fees represented taxes under the TIA rather than regulatory license fees); Robinson Protective Alarm Co v City of Philadelphia, 581 F2d 371 (3d Cir 1978) (holding that city collection of fees for the use of underground lines constituted a tax triggering the TIA).
• 45See John H. Ridge, Fees or Taxes: Rethinking the Bidart Test as Applied to Telecommunication Right-of-Way Charges, 19 J Multistate Taxn & Incentives 30, 30 (Sept 2009).
• 46San Juan Cellular, 967 F2d at 685.
• 47Id.
• 48Id.
• 49Id at 685–86, citing Mississippi Power & Light Co v United States Nuclear Regulatory Commission, 601 F2d 223, 228, 231–32 (5th Cir 1979).
• 50San Juan Cellular, 967 F2d at 685.
• 5173 F3d 925 (9th Cir 1996).
• 52Id at 930–31. The court considered importing a test from bankruptcy law, but it concluded that the bankruptcy test would make the reach of the TIA too broad. See id at 928–29, citing In re Farmers Frozen Food Co, 221 F Supp 385, 387 (ND Cal 1963).
• 53Bidart, 73 F3d at 931–33.
• 54Id.
• 55Id at 933.
• 56For a discussion of the Seventh Circuit decisions in Hager v City of West Peoria, 84 F3d 865 (7th Cir 1996), and Kathrein v City of Evanston, 636 F3d 906 (7th Cir 2011), see Part II.A.3. I have separated Posner’s approach from the rest of the Seventh Circuit in order to highlight his unique and influential approach to the problem. Though I conclude below that his approach fails to grapple with the category of penalties, his clarity and consistency are persuasive in a convoluted area of law.
• 57San Juan Cellular, 967 F2d at 685.
• 58Diginet, Inc v Western Union ATS, Inc, 958 F2d 1388, 1399 (7th Cir 1992).
• 59958 F2d 1388 (7th Cir 1992).
• 60Id at 1399.
• 61Id.
• 62Id.
• 63Diginet, 958 F3d at 1399.
• 64Id at 1391.
• 65Id.
• 66Id at 1392.
• 67Diginet, 958 F3d at 1399.
• 68651 F3d 722 (7th Cir 2011) (en banc).
• 69Id at 726–27.
• 70Id at 729–30.
• 71Id at 728.
• 72Empress Casino, 651 F3d at 729.
• 73Id.
• 74Valero Terrestrial Corp v Caffrey, 205 F3d 130, 134 (4th Cir 2000) (emphasis added), citing San Juan Cellular, 967 F2d at 685.
• 7584 F3d 865 (7th Cir 1996).
• 76Id at 870–71.
• 77Id at 871.
• 78Id.
• 79Hager, 84 F3d at 871.
• 80636 F3d 906 (7th Cir 2011).
• 81Id at 910, 912.
• 82See id at 912.
• 83See id.
• 84See Kathrein, 636 F3d at 912–13.
• 85594 F3d 742 (10th Cir 2010).
• 86Id at 750.
• 87Id at 761, citing Hill v Kemp, 478 F3d 1236, 1244–46 (10th Cir 2007), and Marcus v Kansas Department of Revenue, 170 F3d 1305, 1311–12 (10th Cir 1999).
• 88Chamber of Commerce, 594 F3d at 762.
• 89Id at 763.
• 90143 F3d 1006 (5th Cir 1998).
• 91Id at 1011.
• 92505 F2d 1310 (5th Cir 1975).
• 93Home Builders, 143 F3d at 1011.
• 94Id at 1012.
• 95Id.
• 96169 F3d 448 (7th Cir 1999).
• 97Id at 457.
• 98Id.
• 99Id.
• 100See Michael D. Makowsky and Thomas Stratmann, Political Economy at Any Speed: What Determines Traffic Citations?, 99 Am Econ Rev 509, 517 (2009) (finding that drivers are 26 percent more likely to be fined for a traffic stop when in a municipality that fits a proxy variable for fiscal distress); Thomas A. Garrett and Gary A. Wagner, Red Ink in the Rearview Mirror: Local Fiscal Conditions and the Issuance of Traffic Tickets, 52 J L & Econ 71, 86 (2009) (finding that “negative changes in local revenue from the previous fiscal year are significantly correlated with the change in the number of tickets issued”).
• 101338 F3d 442 (5th Cir 2003).
• 102Id at 444.
• 103Id, quoting Home Builders, 143 F3d at 1011.
• 104See Home Builders, 143 F3d at 1011–12. See also Part II.A.4.
• 105Washington I, 338 F3d at 446–48 (Duval dissenting).
• 106424 Fed Appx 307 (5th Cir 2011).
• 107Id at 309–10.
• 108Id at 310, quoting Washington I, 338 F3d at 444.
• 109Washington II, 424 Fed Appx at 311, quoting Home Builders, 143 F3d at 1011.
• 110637 Fed Appx 363 (9th Cir 2016), cert denied, 137 S Ct 294 (2016).
• 111Id at 364.
• 112Home Builders, 143 F3d at 1011, citing Tramel v Schrader, 505 F2d 1310, 1316 (5th Cir 1975).
• 113Washington I, 338 F3d at 444, quoting Arkansas v Farm Credit Services of Central Arkansas, 520 US 821, 826–27 (1997).
• 114Brohl, 135 S Ct at 1128.
• 115Id.
• 116Id.
• 117Direct Marketing Association v Brohl, 735 F3d 904, 909 (10th Cir 2013).
• 118Id at 910.
• 119Id at 912–13.
• 120Brohl, 135 S Ct at 1129–30.
• 121Id at 1130. It is worth noting that there is no universal federal definition of “tax” elsewhere in federal law like there is for “assessment, levy, or collection.” At least, no circuit court has so far claimed to have found one. So while the Supreme Court’s decision in Brohl helps show that legislative purpose alone will not suffice when defining words under the TIA, it does not offer a positive path toward a solution for the definition of “tax.”
• 122See Reid Pillifant, How John Roberts Saved Obamacare by Ignoring Obama’s ‘Tax’ Argument (Politico, June 28, 2012), archived at http://perma.cc/8P3H-ZP72; David Cole, Obamacare Upheld: How and Why Did Justice Roberts Do It? (The Nation, June 28, 2012), online at http://www.thenation.com/article/obamacare-upheld-how-and-why-did-justice-roberts-do-it/ (visited June 1, 2017) (Perma archive unavailable); Tom Scocca, Obama Wins the Battle, Roberts Wins the War (Slate, June 28, 2012), archived at http://perma.cc/2T3G-3NF8.
• 123See Susan Rupe, The Individual Mandate: It’s What Keeps Everyone in the Risk Pool (InsuranceNewsNet, Nov 18, 2016), archived at http://perma.cc/NFN4-9FTC.
• 12426 USC § 5000A(b)(1) (“If a taxpayer . . . fails to meet the requirement of [minimum essential coverage] for 1 or more months, then . . . there is hereby imposed on the taxpayer a penalty.”) (emphasis added).
• 125US Const Art I, § 8, cl 1 (“The Congress shall have Power To lay and collect Taxes, Duties, Imposts and Excises . . . .”).
• 126The Commerce Clause theory was advanced by Justice Ruth Bader Ginsburg in a concurring opinion, joined by three other members of the Court. See NFIB, 567 US at 589–646 (Ginsburg concurring in part, concurring in the judgment in part, and dissenting in part).
• 127NFIB, 567 US at 543.
• 128651 F3d 529 (6th Cir 2011).
• 129Id at 539 (citation omitted). For the relevant language from the ACA, see note 124.
• 130Thomas More, 651 F3d at 544.
• 131671 F3d 391 (4th Cir 2011).
• 132Id at 405.
• 133Id at 404.
• 134661 F3d 1 (DC Cir 2011).
• 135Id at 6.
• 136Id at 7.
• 137NFIB, 567 US at 544.
• 138Id.
• 139Id.
• 140Tramel, 505 F2d at 1315 n 7.
• 141See, for example, Robinson Protective Alarm Co v City of Philadelphia, 581 F2d 371, 374 (3d Cir 1978) (“The Tramel court denied any controlling effect to state courts’ distinctions between taxes and special assessments.”); Trailer Marine Transport Corp v Rivera Vazquez, 977 F2d 1, 5 (1st Cir 1992) (“Puerto Rico’s decision to call the fee a ‘contribution’ or ‘premium’ . . . rather than a ‘tax’ may be pertinent but does not decide the matter, for it is federal law that determines what constitutes a tax for this purpose.”), citing Robinson Protective, 581 F2d at 374. See also, for example, Bidart, 73 F3d at 933 (“Regardless of the labels placed on the Commission’s duties and functions, its assessments are not ‘taxes’ within the meaning of the TIA.”); Diginet, 958 F2d at 1399 (“[The city] may not circumvent this limitation by calling a tax something else, such as a ‘franchise fee.’”).
• 142The idea that Congress can rely on state legislatures in the course of federal lawmaking is an uncontroversial proposition. Usually, this this is phrased as “incorporation” of state law. See, for example, Radha A. Pathak, Incorporated State Law, 61 Case W Reserve L Rev 823, 838 (2011) (“Any type of federal law—constitutional, statutory, or judicial—can borrow state law, and state law can be borrowed for either substantive or procedural purposes.”). Consider, for instance, the Air Transportation Safety and System Stabilization Act, passed by Congress in the wake of the September 11 attacks. This enactment created a federal cause of action for victims against airlines. But instead of re-creating all of tort and contract law, the federal statute simply incorporated all relevant state law: “The substantive law for decision in any such suit shall be derived from the law, including choice of law principles, of the State in which the crash occurred.” See Air Transportation Safety and System Stabilization Act § 408(b)(2), Pub L No 107-42, 115 Stat 232, 241 (2001). Under federal statutes like these, states can dramatically alter the substance of federal law by changing their own law. In contrast, this Comment argues merely that the TIA can and does rely on a single word in state law—the label given to a charge—when determining whether the statute is a tax, a much less dramatic incorporation of state law.
• 143Brohl, 135 S Ct at 1129 (citations omitted).
• 144Fredrickson v Starbucks Corp, 840 F3d 1119, 1122 (9th Cir 2016).
• 145542 US 88 (2004).
• 146Id at 115 (Kennedy dissenting), citing Laing v United States, 423 US 161, 170 n 13 (1976).
• 147See Enochs v Williams Packing & Navigation Co, 370 US 1, 7 & n 6 (1962) (relying on the Senate report on the TIA to support the assertion that, under the AIA, “the United States is assured of prompt collection of its lawful revenue”), citing S Rep No 75-1035 (cited in note 24).
• 148See Fredrickson, 840 F3d at 1122, citing United States v American Friends Service Committee, 419 US 7, 10 (1974).
• 149See American Atheists, Inc v Shulman, 21 F Supp 3d 856, 868–69 (ED Ky 2014).
• 150See Cohen v United States, 650 F3d 717, 730–31 (DC Cir 2011).
• 151See National Federation of Republican Assemblies v United States, 148 F Supp 2d 1273, 1286 & n 16 (SD Ala 2001).
• 152See Pazzo Pazzo, Inc v New Jersey, 2007 WL 4166017, *3 (D NJ).
• 153See, for example, Florida Bankers Association v United States Department of the Treasury, 799 F3d 1065, 1074 n 2 (DC Cir 2015) (Randolph concurring) (stating that the opinion would “refer to the AIA only” even though “some of the cases cited herein interpret the Tax Injunction Act”).
• 154Butler Act Amendments § 7, 44 Stat 1418, 1421 (1927), codified at 48 USC § 872 (“No suit for the purpose of restraining the assessment or collection of any tax imposed by the laws of Puerto Rico shall be maintained in the United States District Court for the District of Puerto Rico.”).
• 155See Part II.A.
• 156See, for example, Wal-Mart Puerto Rico, Inc v Zaragoza-Gomez, 834 F3d 110, 119 (1st Cir 2016) (“As acknowledged above, the Butler Act and the TIA have been construed in pari materia in our circuit, which has extended the TIA’s exception to the Butler Act.”) (quotation marks omitted).
• 157Jefferson County v Acker, 527 US 423, 435 (1999).
• 158McCrory Corp v Ohio, 212 BR 229, 232–33 (SDNY 1997).
• 159723 F3d 1114 (10th Cir 2013) (en banc).
• 160Id at 1158 (Gorsuch concurring), quoting 28 USC § 1341 and 26 USC § 7421 (emphasis omitted).
• 161Hobby Lobby, 723 F3d at 1158 (Gorsuch concurring).
• 162It would still have the power to create particular exceptions to state-law labels, though it would presumably never spend time on such a small matter in practice. Of course, it would also have the power to abolish the label test by statute.
• 16328 USC § 1341 (emphasis added).
• 1645 USC § 8478.
• 165S Rep No 75-1035 at 2 (cited in note 24). The report from the House Committee on the Judiciary incorporated the Senate report in its entirety. See generally Suits Relating to Collection of State Taxes, H Rep No 75-1503, 75th Cong, 1st Sess (1937).
• 166S Rep No 75-1035 at 2 (cited in note 24).
• 167Id.
• 168Id.
• 169Id. The statutory grant for the Supreme Court’s appellate jurisdiction over such cases comes from 28 USC § 1257(a), which states that the Supreme Court may review decisions “rendered by the highest court of a State” for which the validity of a federal statute is questioned or a state law is alleged to be in conflict with the Constitution or a federal law.
• 170Tully v Griffin, Inc, 429 US 68, 73 (1976).
• 171450 US 503 (1981).
• 172Id at 522.
• 173Arkansas v Farm Credit Services of Central Arkansas, 520 US 821, 826 (1997).
• 174Laborde v City of Gahanna, 561 Fed Appx 476, 479 (6th Cir 2014).
• 175See generally Bidart, 73 F3d 925.
• 176Id at 931–33.
• 177See Cal Food & Ag Code § 75630(a)–(d).
• 178See Cal Food & Ag Code § 75630(d).
• 179See Diginet, 958 F2d at 1392 (finding that the activity of laying fiber optic cables “imposes no costs, congestion or otherwise”); Empress Casino, 651 F3d at 730 (holding that the tax is an example of “a state’s taking money from one group of firms and giving it to another group,” untethered from any specific regulations).
• 180Hager, 84 F3d at 867.
• 181Id at 871.
• 182Id at 870.
• 183Id at 871.
• 184Empress Casino, 651 F3d at 729.
• 185See Washington I, 338 F3d at 445–48 (Duval dissenting); Washington II, 424 Fed Appx at 312.
• 186Empress Casino, 651 F3d at 728.
• 187See, for example, House Fiscal Agency, State of Michigan Revenue: State Source and Distribution *4 (Oct 2013), archived at http://perma.cc/9DF8-M7EV (containing a table that shows that in 2011–2012, Michigan received over 80 percent of its income from sources already labeled “taxes,” such as sales tax, income tax, tobacco tax, and business taxes); Jeffrey L. Barnett, et al, 2012 Census of Governments: Finance—State and Local Government Summary Report *7 (US Census Bureau, Dec 17, 2014), archived at http://perma.cc/7HVR-CBQL (showing that taxes and intergovernmental transfers made up more than 75 percent of state government revenues).
• 188See, for example, Kent County Office of the Administrator, 2016 Adopted Budget *22 (2016), archived at http://perma.cc/DQQ6-CDD3 (containing a chart showing that in the fiscal year 2016, Kent County, Michigan, received only 33.5 percent of its revenue from “taxes”). But see Barnett, et al, 2012 Census of Governments at *7 (cited in note 187) (showing that property and other taxes make up more than half of local government revenue). At the very least, because there are so many more local government entities than state governments, one would expect a large amount of variation.
• 189Home Builders, 143 F3d at 1009.
• 190Id.
• 191NFIB, 567 US at 544.
• 192Empress Casino, 651 F3d at 740 n 3 (Sykes dissenting). See also Hager, 84 F3d at 871; Kathrein, 636 F3d at 912 (“The statute or ordinance creating a charge may provide evidence of its purpose, though the enacting government’s characterization of a charge is not determinative.”); Trailer Marine, 977 F2d at 5 (“Puerto Rico’s decision to call the fee a ‘contribution’ or ‘premium’ rather than a ‘tax’ may be pertinent but does not decide the matter, for it is federal law that determines what constitutes a tax for this purpose.”) (citation omitted); Robinson Protective, 581 F2d at 374.
• 193See, for example, Tramel, 505 F2d at 1315 n 7.
• 194S Rep No 75-1035 at 2 (cited in note 24).
• 19528 USC § 1341.
• 196Christopher C. Fennell and Lee Anne Fennell, Fear and Greed in Tax Policy: A Qualitative Research Agenda, 13 Wash U J L & Pol 75, 75–76 (2003).
• 197See Jonathan S. Masur and Eric A. Posner, Toward a Pigouvian State, 164 U Pa L Rev 93, 141 & n 197 (2015). See also Hayes R. Holderness, The Unexpected Role of Tax Salience in State Competition for Businesses, 84 U Chi L Rev 1091, 1126–27 & n 139 (2017).
• 198Edward J. McCaffery and Jonathan Baron, Thinking about Tax, 12 Psychology, Pub Pol & L 106, 117–18 (2006).
• 199Id at 118.
• 200Id.
• 201Id.
• 202McCaffery and Baron, 12 Psychology, Pub Pol & L at 118 (cited in note 198).
• 203Fennell and Fennell, 13 Wash U J L & Pol at 76 (cited in note 196).
• 204Abigail B. Sussman and Christopher Y. Olivola, Axe the Tax: Taxes Are Disliked More Than Equivalent Costs, 48 J Mktg Rsrch S91, S92–93 (2011).
• 205Rosewell, 450 US at 522.