Severability — the notion that a court may excise an unconstitutional part of a statute while leaving valid portions intact — forms a core tenet of American constitutional law.1 Courts have long maintained a strong presumption of severability,2 premised on the rule that a court should not “nullify more of a legislature’s work than is necessary.”3 A court should sever4 a statute “[u]nless it is evident that the Legislature would not have enacted [the valid portions]” without the invalid portion.5 Last Term, in Murphy v. National Collegiate Athletic Association,6 the Supreme Court held that a provision of the Professional and Amateur Sports Protection Act7 (PASPA) violated anticommandeering doctrine by prohibiting states from authorizing private sports gambling.8 The Court declared inseverable portions of PASPA that prohibited states and private parties from organizing or promoting sports gambling.9 Despite purporting to conduct a traditional severability inquiry, the Court made a number of subtle but significant departures from that inquiry. These changes may foreshadow a shift toward saving fewer schemes.
Fear that sports gambling would become widespread spurred PASPA’s passage. Until the 1980s, only Nevada and New Jersey allowed casino gambling, but many states and Native American reservations soon legalized it.10 Although sports gambling — betting on amateur and professional sports — remained legal only in Nevada,11 it appeared likely to gain national acceptance12 and many worried that this would corrupt sports leagues and increase gambling addiction among youth.13
Congress responded to these concerns with PASPA,14 which banned states and individuals from operating sports-gambling schemes. First, PASPA § 3702(1) made it unlawful for “a governmental entity to sponsor, operate, advertise, promote, license, or authorize by law” a sports-gambling scheme.15 Second, § 3702(2) made it illegal for “a person to sponsor, operate, advertise, or promote, pursuant to the law . . . of a governmental entity” a sports-gambling scheme.16 The Attorney General or a sports organization could file suit to enjoin violations of the provisions.17 The law allowed Nevada to maintain sports gambling and provided New Jersey one year in which to legalize it, but New Jersey declined.18 Twenty years later, New Jersey’s legislature reversed course and passed a law authorizing sports gambling.19
Professional sports leagues and the National Collegiate Athletic Association (NCAA) sued to enjoin the law as a violation of PASPA; New Jersey countered that PASPA was invalid because it violated anticommandeering doctrine by prohibiting the State from legalizing sports gambling.20 That doctrine prohibits Congress from issuing direct orders to states because Article I does not enumerate that power and the Tenth Amendment reserves all unenumerated powers to the states.21 Although Congress can “requir[e] or prohibit[] certain acts,” it cannot directly “compel the States to require or prohibit those acts.”22
The district court found PASPA did not commandeer the State,23 and the Third Circuit affirmed.24 The panel reasoned that only an affirmative command, which forces the State to take new action (rather than refrain from acting), would constitute an anticommandeering violation, and PASPA did not impose such a command.25 Moreover, the court concluded, a repeal of New Jersey’s current prohibition on sports gambling would not constitute an “authoriz[ation]” in violation of PASPA.26
New Jersey took up the Third Circuit’s suggestion and repealed most of its prohibitions on sports gambling.27 The same plaintiffs sued and the district court enjoined the new law after finding it preempted by PASPA.28 The Third Circuit, in a divided panel29 and then sitting en banc, affirmed.30 The latter found the repeal constituted an authorization because it “selectively remove[d] a prohibition [and] permissively channel[ed] wagering activity,”31 but concluded that PASPA did not commandeer New Jersey because it issued no affirmative command.32
The Supreme Court reversed.33 Writing for the Court, Justice Alito34 held that § 3702(1) violated the anticommandeering rule by prohibiting states from authorizing sports gambling, and its remaining portions could not be severed.35 PASPA violated anticommandeering doctrine, the Court reasoned, by issuing a direct order to the states. Past anticommandeering cases had involved affirmative commands that required states to act. The respondents argued that PASPA, by contrast, simply required states to refrain from acting.36 The Court saw no difference: in both instances, Congress “issue[d] direct orders to state legislatures,” placing them under its “direct control.”37 That past cases confronted only “affirmative” commands, not prohibitions, was “happenstance.”38
The Court also rejected the argument that PASPA was a valid exercise of Congress’s power to preempt state law. Federal law preempts state law only where the federal law regulates private parties39 and exercises an Article I power outside the Supremacy Clause.40 Preemption did not apply because § 3702(1) regulated only states and had no other constitutional basis.41 Meanwhile, the subsection’s prohibition on “licens[ing]” ran afoul of the Constitution for the same reasons.42
Having found invalid the prohibitions on state authorization and licensing, the Court assessed the remaining provisions and concluded they could not be severed.43 Justice Alito stated that the remainder would not be severed if it was “evident that [Congress] would not have enacted those provisions which are within its power, independently of [those] which [are] not.”44 And the Court would inquire “whether the law remains ‘fully operative’ without the invalid provisions.”45
The Court began by striking § 3702(1)’s prohibition on state-“operated” sports-gambling schemes. The majority concluded it “seem[ed] most unlikely” that Congress would have wanted to prevent states from running sports-betting lotteries if they could authorize such betting in casinos because state lotteries were viewed as far more benign.46 Moreover, it would have been “unusual” for Congress to ban states from commercial activity allowed to private individuals.47 Thus, the Court declined to sever the prohibition on operation.48 Further, the Court could find no clear distinction between operation and “sponsor[ship]” or “promot[ion],” so it also declined to sever these from § 3702(1).49
The majority next declined to sever § 3702(2)’s prohibition on private individuals operating or promoting sports-gambling schemes “‘pursuant to’ state law.”50 First, the Court noted sections 3702(1) and 3702(2) were meant “to work together” to achieve PASPA’s goal of preventing “state legalization of sports gambling.”51 Second, the majority noted that most federal gambling law implements a “coherent” policy of letting states decide what to make illegal and then imposing federal liability only when gambling violates state law.52 But § 3702(2), by itself, “cease[d] to implement any coherent federal policy”53 because it would render legal under federal law sports gambling that was prohibited by state law, and illegal under federal law sports gambling allowed by state law.54 The Court doubted “that Congress ever contemplated . . . such a weird result” and declined to sever.55
“PASPA’s enforcement scheme reinforce[d] [the Court’s] conclusion”: injunctions by the Attorney General or sports organizations could stop a small number of violators — feasible if state prohibitions remained intact — but would fail if sports gambling were widely legalized.56
Justice Thomas concurred to express his concern with modern severability law.57 First, as a historical matter, he argued early American courts did not sever; they merely declined to enforce the unconstitutional law in the case before them.58 Second, modern severability precedent deviates from the typical textual focus of statutory interpretation to instead ask what Congress would have intended — but Congress likely had no intent because it “typically does not pass statutes with the expectation that some part will later be deemed unconstitutional.”59 Third, severability conflicts with standing doctrine by requiring courts to address provisions no present party has standing to challenge.60
Justice Ginsburg dissented.61 First, she argued PASPA did not commandeer the State because its provisions fell within Congress’s power to regulate interstate commerce.62 Second, she argued severability should save the remaining provisions. Justice Ginsburg remonstrated that the Court “ordinarily engages in a salvage”63 by “severing any problematic portions while leaving the remainder intact.”64 She argued that the majority made a critical “mistaken assumption that private sports-gambling schemes would become lawful” once the prohibition on state authorization fell.65 This was incorrect, she reasoned, because even if state law authorized a sports-gambling scheme, federal law would make that scheme illegal under § 3702(2).66 The majority’s mistaken assumption ran through its severability analysis: that mistake drove the decision to strike state “operat[ion],” which then led to decisions to strike state “sponsor[ship]” and “promot[ion],” which in turn led to the decision to strike the analogous prohibitions on private parties, and finally “advertis[ing].”67 She concluded there was no rational basis for inferring Congress would have preferred no statute at all to the severed statute in the context of its effort to “stop[] sports-gambling regimes.”68
Justice Breyer concurred in part and dissented in part.69 He concurred with Justice Alito that § 3702(1) violated the anticommandeering rule, but thought New Jersey’s victory should be “mostly Pyrrhic.”70 He argued Congress sought to “keep sports gambling from spreading.”71 While § 3702(1) did this impermissibly, Justice Breyer agreed with Justice Ginsburg that § 3702(2) was a constitutional exercise of Congress’s power to regulate interstate commerce that could stand alone as an independent means to prevent the spread of sports gambling.72
Although it purported to conduct a traditional severability analysis, the Court made subtle but important departures from that test; these deviations may signal a shift toward saving fewer regulatory schemes. Courts going back as far as Marbury v. Madison73 have maintained a robust, if at times implicit, presumption of severability,74 and the basic rules of severability doctrine have remained largely consistent for decades. The Murphy majority made three key departures from that doctrine. First, the Court focused on the alignment between separate statutory schemes, rather than the viability of the valid remainder, to conclude Congress would not have preferred severing. Second, the majority declined to sever in the context of uncertainty — but the traditional test would find uncertainty a reason to sever. Third, the Court eschewed the typical severability focus on Congress’s overall goal to instead take a step-by-step approach that compared the effect of the severed statute to the invalidated portion. Cumulatively, these departures led the Court to an outcome that defies the typical schema of severability decisions and creates new uncertainty for regulatory schemes.
Courts assess severability with a “well-established” framework: “Unless it is evident that the Legislature would not have enacted those provisions which are within its power, independently of that which is not, the invalid part may be dropped if what is left is fully operative as a law.”75 This test supports severing when: (1) the statute’s remaining, valid part is “[]capable of functioning independently”;76 and (2) Congress would have preferred “what is left . . . to no statute at all.”77
A strong presumption of severability buttresses this framework. Justices and commentators agree that courts maintain a robust presumption in favor of severing when severability is raised.78 Indeed, the Court has instructed, “‘[W]hen confronting a constitutional flaw in a statute, we try to limit the solution to the problem,’ severing any ‘problematic portions while leaving the remainder intact.’”79 The traditional two-part test embodies this presumption: it requires legislative intent contrary to severability to be evident — that is, obvious — for a court to decline severing on the basis of that intent. For example, Free Enterprise Fund v. Public Co. Accounting Oversight Board80 rejected arguments against severability because “nothing in the statute’s text or historical context ma[de] it ‘evident’ that Congress . . . would have preferred no [scheme] to [the severed one].”81 The Murphy majority departed from this presumption and every part of the framework.
Murphy deviated by focusing on PASPA’s “coherence,” rather than its ability to function independently. Justice Alito noted that most federal sports-gambling law creates a “coherent” scheme that respects states’ decisions by imposing federal liability only for gambling a state already prohibits.82 In contrast, he reasoned, § 3702(2), once severed from § 3702(1), would “cease[] to implement any coherent federal policy”83 because it would prohibit sports gambling that was legal under state law but allow sports gambling that state law prohibited.84 The Court concluded this was “exactly the opposite of the general federal approach to gambling,” and would “implement[] a perverse policy that undermines whatever policy is favored by . . . a State.”85 This analysis affronts traditional severability doctrine. First, the relevant inquiry is not whether the severed scheme differs from other statutory schemes. Rather, the focus is whether the scheme is “fully operative as a law”86; if it is “incapable of functioning independently,”87 this counts against congressional intent, but mere oddity does not render a law inoperative. Second, as Justices Ginsburg and Breyer argued, the new scheme would be capable of functioning independently.88 Murphy itself shows that sports leagues were capable of enforcing PASPA through suits — without imposing enforcement costs on the Attorney General. Third, the Court overlooked the obvious: Congress built that very “perverse policy” into the original scheme of PASPA. It always differed from the general federal gambling scheme because Congress wanted to override states’ decisions about sports gambling.89 The Court’s focus on “coherence” represents a critical departure from severability doctrine.
The majority further deviated from the presumption of severability by declining to sever despite uncertainty and a paucity of evidence about congressional intent. In striking PASPA’s prohibitions on private actors, the Court concluded: “We do not think that Congress ever contemplated [the severed scheme].”90 But the majority cited nothing in the statute’s “text or histor[y]”91 to support its analysis. The Court failed to demonstrate it was “evident” that Congress would have preferred no statute to the severed version. Moreover, Murphy encountered uncertainty about how many states would choose to legalize sports gambling — and implicitly concluded first, that a large number would, and second, that state legalization would render sports gambling legal generally.92 These assumptions contrast with the Court’s prior approach to uncertainty: erring toward severability, rather than assuming an outcome,93 or remanding for a lower court to further assess legislative intent.94 This represents a weakening of the presumption: “evident” requires strong affirmative proof Congress would not have wanted the remainder. But declining to sever despite uncertainty suggests a heightened demand for proof that Congress would have preferred the remainder. This shift further deviates from the doctrine.
Justice Alito’s step-by-step assessment of the purpose behind each portion of the severed scheme also departed from the traditional congressional-intent inquiry. The majority conspicuously omitted the rule that a court should maintain the severed statute if “the legislature [would] have preferred what is left . . . to no statute at all.”95 Not only do the concurrence and dissent reference this inquiry,96 but the entire Court relied on the test in two recent leading cases.97 In assessing the prohibition on state-operated sports gambling, the Court did not compare Congress’s preference for the severed scheme to its preference for no scheme at all. Rather, it assessed whether Congress would prefer to prevent state-run sports lotteries if states could authorize private sports gambling.98 This narrow focus on just the state-operation provision occludes the broader insight of the “no statute at all” inquiry, which asks not whether Congress would nitpick a particular provision, but whether it would prefer the cumulative effect of the severed scheme.
New York v. United States99 — the founding case of anticommandeering doctrine — reflected this focus on the cumulative effect when it concluded: “[W]here Congress has enacted a statutory scheme for an obvious purpose, and where Congress has included a series of provisions operating as incentives to achieve that purpose, the invalidation of one of the incentives should not ordinarily cause Congress’ overall intent to be frustrated.”100 New York and the “no statute at all” inquiry indicate all remaining, severable provisions should be assessed as a whole, rather than inspected piecemeal. Where individual parts may appear to head in different directions, observing them as a constellation may reveal an “overall intent.” After working piecemeal through provisions, the majority concluded that PASPA “aimed to prevent . . . state legalization of sports gambling.”101 To be sure, this was one objective of PASPA — but it overlooks the overall intent behind that goal: “to ‘keep sports gambling from spreading.’”102 Each of the remaining provisions served that intent, but the Court’s analysis led it to decline to sever them.
Murphy’s approach to the congressional-intent inquiry provides a guide for attacking the administrative state and casts uncertainty on other regulatory schemes. Rather than assess whether the remaining scheme would accomplish Congress’s overall goal, the Court characterized one of its methods of achieving that goal — preventing states from authorizing sports gambling — as Congress’s central intent. Despite Professor Michael Dorf’s conclusion that the congressional-intent inquiry “may carry no real force,”103 the Court may well weaponize that inquiry by characterizing the invalidated portion of a statute as the central purpose of that statute — thereby guaranteeing its remainder deviates from “congressional intent.” Indeed, opponents of the Affordable Care Act and the Consumer Financial Protection Bureau (CFPB) have already adopted this tactic, arguing the statutory portions they seek to have declared invalid are the central goals of the statutes and the remaining portions — including the entire CFPB — cannot be saved.104
By focusing on coherence, declining to sever in the face of uncertainty, and refocusing the congressional-intent inquiry, the Murphy Court departed from traditional severability doctrine. These deviations may lay the groundwork for a broader reformation of the Court’s approach to severability. If so, Murphy’s methodology bodes poorly for other regulatory schemes, whose opponents have already made severability a new front in their war on the administrative state.