The federal government relies on the efficient appointment of “Officers of the United States”^1×1. U.S. Const. art. II, § 2, cl. 2. to run smoothly.^2×2. See Myers v. United States, 272 U.S. 52, 117 (1926) (“The vesting of the executive power in the President was essentially a grant of the power to execute the laws. But the President alone and unaided could not execute the laws. He must execute them by the assistance of subordinates.”). A “structural safeguard[] of the constitutional scheme,” the Appointments Clause gives the President and the Senate each a role in the process.^3×3. Edmond v. United States, 520 U.S. 651, 659 (1997); see U.S. Const. art. II, § 2, cl. 2. One historically contested area within that division of power is the appointment and service of interim acting  officers when vacancies arise. In 1998, Congress passed the Federal Vacancies Reform Act^4×4. Pub. L. No. 105-277, div. C., tit. I, § 151, 112 Stat. 2681-611 (1998) (codified as amended at 5 U.S.C. §§ 3345–3349d (2012)). (FVRA), which in section 3345 allows three groups of people to serve as acting officers: (a)(1) first assistants to the vacant office; (a)(2) officials in other positions requiring Senate confirmation; and (a)(3) high-ranking employees within the given agency.^5×5. 5 U.S.C. § 3345(a)(1)–(3). To qualify for the third category, employees must have served for at least ninety days in their high-ranking post in the year before the vacancy. See id. § 3345(a)(3)(A). But with that permission also came a limitation: “[n]otwithstanding subsection (a)(1), a person may not serve as an acting officer for an office under this section” if nominated by the President for the permanent office, unless the nominee has served as a first assistant to that office for ninety days during the year preceding the vacancy.^6×6. Id. § 3345(b)(1).

Last Term, in NLRB v. SW General, Inc.,^7×7. 137 S. Ct. 929 (2017). the Supreme Court held that the FVRA’s limitation on nominees serving as acting officers applies to all three groups of people, not only the first assistants mentioned in subsection (a)(1).^8×8. Id. at 935. To get there, the Court relied on the ordinary meaning of the provision’s text over and against arguments from purpose, post-enactment practice, and even a semantic canon. The finding of textual clarity here is the latest data point for plotting what some have called the Roberts Court’s willingness to find textual ambiguity and embrace purposivism in major cases.^9×9. See, e.g., Ryan D. Doerfler, High-Stakes Interpretation, 116 Mich. L. Rev. (forthcoming Feb. 2018) (manuscript at 3), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2928333 [https://perma.cc/BC29-T3XX] (“As different scholars have noted, courts treat statutory text as more malleable in big cases.”). But the stakes in the battle over FVRA interpretation were arguably just as high as in some of the Court’s key purposivism cases. SW General thus points up the continued need for a better account from the Court of what kinds and magnitudes of consequences suffice to create ambiguity in otherwise clear texts.

In June 2010, the General Counsel of the NLRB resigned.^10×10. SW Gen., Inc. v. NLRB, 796 F.3d 67, 71 (D.C. Cir. 2015). By statute, the General Counsel position requires presidential appointment and Senate confirmation.^11×11. 29 U.S.C. § 153(d) (2012). President Obama assigned the role of acting General Counsel to Lafe Solomon, an office director within the NLRB.^12×12. SW General, 796 F.3d at 71. The President later nominated Solomon to fill the role permanently but was twice rebuffed by the Senate.^13×13. Id. Solomon continued to serve in an acting capacity throughout the nomination process until November 4, 2013, when a newly confirmed General Counsel took office.^14×14. Id.

During Solomon’s acting service, SW General, Inc., an Arizona-based ambulance services company, was charged with committing unfair labor practices under the National Labor Relations Act.^15×15. 29 U.S.C. §§ 151–169; see SW Gen., Inc., 360 N.L.R.B. 835, 837 (2014). Acting on Solomon’s legal behalf, a Regional Director filed the formal complaint in January 2013, alleging the company had unilaterally discontinued bonus payments to its employees — payments subject to mandatory bargaining.^16×16. SW General, 796 F.3d at 71–72. An Administrative Law Judge found the allegations true, and SW General filed exceptions with the NLRB.^17×17. SW General, 360 N.L.R.B. at 835, 844. The company argued the initial complaint was invalid because the limitation in the FVRA prohibited Solomon from serving as acting General Counsel after being nominated for the permanent position: he had never served as first assistant to the General Counsel, let alone for ninety days in the year preceding the vacancy.^18×18. SW General, 796 F.3d at 72. The Board adopted the order below without addressing the argument.^19×19. SW General, 360 N.L.R.B. at 835–37.

The D.C. Circuit vacated the Board’s order.^20×20. SW General, 796 F.3d at 69. Writing for a unanimous panel, Judge Henderson^21×21. Judge Henderson was joined by Judges Srinivasan and Wilkins. agreed with SW General that the FVRA’s limitation on acting service by nominees extends to all three groups named in the statute. Solomon fell within the third category as a high-ranking employee, and the text of the limitation applies broadly to “a person” attempting to serve as an acting officer “under this section.”^22×22. SW General, 796 F.3d at 73–74 (emphases omitted) (quoting 5 U.S.C. § 3345(b)(1) (2012)). Once nominated, then, Solomon could only continue serving as an acting officer if he had been a first assistant to the General Counsel for ninety days in the year before the vacancy, which he had not.^23×23. See id. at 78. Judge Henderson rejected the government’s arguments from legislative history and purpose, see id. at 76–78, for reasons the Supreme Court would largely echo. Judge Henderson then turned to the question of remedies. The FVRA provides an express exemption for the NLRB General Counsel, thus Solomon’s actions were “voidable,” rather than void ab initio.^24×24. Id. at 79; see id. at 78–79 (citing 5 U.S.C. § 3348(e)(1)). Nonetheless, the government’s arguments that the initial defect in the complaint had subsequently been cured lacked merit.^25×25. See id. at 79–82 (rejecting theories of harmless error and de facto officer doctrine). The government petitioned for review of whether the FVRA’s limitation applies only to the category of first assistants.^26×26. See Petition for Writ of Certiorari at I, SW General, 137 S. Ct. 929 (No. 15-1251).

The Supreme Court affirmed. Writing for the Court, Chief Justice Roberts^27×27. Chief Justice Roberts was joined by Justices Kennedy, Thomas, Breyer, Alito, and Kagan. opened by tracing the history of the “perceiv[ed] . . . threat to the Senate’s advice and consent power” that had led to the FVRA.^28×28. SW General, 137 S. Ct. at 936; see id. at 935–36. He then began analysis with the ordinary meaning of the limitation’s text. Like the court below, Chief Justice Roberts noted that “a person” and “this section” are broad terms that make no distinction between the three groups of acting officers mentioned in the FVRA.^29×29. Id. at 938–39. After all, high-level employees like Solomon are “person[s]” too.^30×30. See id. at 938. Congress could have easily cabined the limitation to apply only to subsection (a)(1)’s first assistants but chose broad language instead.^31×31. Id. at 939.

Chief Justice Roberts then rejected the government’s primary textual argument. In the government’s view, the opening phrase “notwithstanding subsection (a)(1)” in the restriction limited that restriction to apply only to (a)(1)’s category of first assistants, under the interpretive canon expressio unius est exclusio alterius.^32×32. Id. at 939–40. But that principle — expressing one item in a series implies exclusion of other unnamed items — “depends on context.”^33×33. Id. at 940 (quoting Marx v. Gen. Revenue Corp., 133 S. Ct. 1166, 1175 (2013)). And here, the relevant context is that of a dependent clause beginning with “notwithstanding.” Such clauses serve to “confirm[] rather than constrain[] breadth” by affirming that a given provision still applies despite conflicting with another.^34×34. Id. Chief Justice Roberts drew an analogy to a radio station that plays hits from the ‘60s, ‘70s, and ‘80s but announces: “Notwithstanding the fact that we play hits from the ‘60s, we do not play music by British bands.”^35×35. Id. The “notwithstanding” clause highlights the category where one might most expect to hear British bands (the ‘60s), but the general prohibition still applies to music from the ‘70s and ‘80s as well.^36×36. See id. Chief Justice Roberts also pointed to similar usages of “notwithstanding” within section 3345 and added that the government’s interpretation would render other portions of the statute superfluous. See id. at 941.

Though textual clarity made it unnecessary to consider extratextual evidence, Chief Justice Roberts concluded by refuting the government’s reliance on legislative history and post-enactment practice.^37×37. See id. at 941–43. Even if certain members of Congress aimed the FVRA’s limitation at first assistants only, the final draft of the statute removed language that would have unambiguously cabined it to that first category.^38×38. See id. at 942–43. And “‘[h]istorical practice’ [was] too grand a title” for instances where Congress did not object on FVRA grounds to nominees from the second and third categories serving as acting officers.^39×39. Id. at 943. First, those nominations represent less than two percent of officer nominations in the FVRA era.^40×40. Id. Second, members of Congress might have abstained from pointing out FVRA problems either because they approved of a given nominee or because they simply did not notice.^41×41. See id. Thus this case was no NLRB v. Noel Canning,^42×42. 134 S. Ct. 2550 (2014). where a “voluminous historical record” concerning recess appointments had weighed heavily in the Court’s analysis.^43×43. SW General, 137 S. Ct. at 943. Since the limitation on acting service by nominees extends to high-level employees and not only to first assistants, “Solomon’s continued service violated the FVRA.”^44×44. Id. at 944.

Justice Thomas concurred. He agreed with the majority’s construction of the FVRA’s limitation but wrote separately to address an underlying constitutional question.^45×45. See id. at 945 (Thomas, J., concurring). In his view, the Appointments Clause on its own “likely prohibited Solomon’s appointment” as acting General Counsel because Solomon did not receive Senate confirmation.^46×46. Id. Under both the clause’s original meaning and Court precedent, Solomon would be an officer of the United States, and a principal officer at that.^47×47. See id. at 946–48. For Justice Thomas, the formal label of “acting general counsel” did not exempt Solomon from the constitutional requirements: “[T]he structural protections of the Appointments Clause can[not] be avoided based on such trivial distinctions.”^48×48. Id. at 946 n.1.

Justice Sotomayor dissented.^49×49. Justice Sotomayor was joined by Justice Ginsburg. The “[n]otwithstanding subsection (a)(1)” clause in the FVRA’s limitation, she argued, cabins the potentially broad language of “a person” and “this section.”^50×50. id. at 950 (Sotomayor, J., dissenting). Under expressio unius est exclusio alterius, the clause singles out the category of first assistants as the sole target of that limitation.^51×51. Id. In Justice Sotomayor’s view, the majority’s assertion that “notwithstanding” “confirms rather than constrains breadth” begged the question: the scope of that introductory clause was the very thing at issue, and the clause names only subsection (a)(1).^52×52. Id. at 951. Nor does that clause merely identify (a)(1) as a provision where conflict with the limitation is especially likely, since the limitation would more naturally collide with the other two categories.^53×53. Id. On purpose and practice, Justice Sotomayor traced the impetus for the FVRA to acting service by agency outsiders to whom Congress had strongly objected during the nomination process, not to all acting service.^54×54. Id. at 953 (describing Bill Lann Lee’s extended service as Acting Assistant Attorney General after rejection by the Senate). See generally Steven J. Duffield & James C. Ho, Comment, The (Still) Illegal Appointment of Bill Lann Lee, 3 Tex. Rev. L. & Pol. 403 (1999). Finally, that the President “began violating the FVRA almost immediately after its enactment”^55×55. SW General, 137 S. Ct. at 953 (Sotomayor, J., dissenting). and was met with nothing but congressional silence underscored for Justice Sotomayor that the limitation in the statute applies only to first assistants.^56×56. See id. at 953–54.

The blend of plain meaning, historical practice, and congressional purpose within the FVRA dispute makes SW General a unique study in Roberts Court statutory interpretation. Here, the majority placed primacy on the ordinary meaning of the statute’s terms. “Person” has “a naturally expansive meaning,” and the commonsense referent of “this section” is the entirety of section 3345.^57×57. Id. at 938 (majority opinion). So clear was the text that that Court “need not [have] consider[ed] . . . extra-textual evidence.”^58×58. Id. at 942 (citing State Farm Fire & Cas. Co. v. United States ex rel. Rigsby, 137 S. Ct. 436 (2016)). And yet it is not always so. The Court at times has found statutes with accessible ordinary meanings ambiguous — a turn that some commentators label the Roberts Court’s tendency to treat major cases differently.^59×59. See, e.g., Doerfler, supra note 9 (manuscript at 3–4); id. at 3–5 (summarizing accounts of the “high-stakes/low-stakes disparity,” id. at 5); Richard M. Re, The New Holy Trinity, 18 Green Bag 2d 407, 409 (2015) (arguing the Court’s finding of ambiguity in key purposivist cases rested in part on “pragmatic judgments”); Note, The Rise of Purposivism and Fall of Chevron: Major Statutory Cases in the Supreme Court, 130 Harv. L. Rev. 1227, 1227 (2017) (“[T]he Roberts Court . . . has recently betrayed a purposivist orientation in major statutory cases.”). The major cases idea goes a step beyond the “new purposivism” Dean John Manning has described, by attempting to explain even statutory cases where the text has no “genuine semantic ambiguity.” John F. Manning, The New Purposivism, 2011 Sup. Ct. Rev. 113, 117; see Note, supra, at 1230. But more difficult is providing a coherent description of exactly what elements create enough textual ambiguity such that purpose has the final word. Some recent accounts have described the Court as taking a purposivist tack in cases involving manifest congressional plans,^60×60. See Abbe R. Gluck, The Supreme Court, 2014 Term — Comment: Imperfect Statutes, Imperfect Courts: Understanding Congress’s Plan in the Era of Unorthodox Lawmaking, 129 Harv. L. Rev. 62, 80–81 (2015) (describing the Court’s finding of ambiguity in King v. Burwell, 135 S. Ct. 2480 (2015), as sensitivity to the “comprehensive legislative plan” of Congress, Gluck, supra, at 81). sweeping practical consequences,^61×61. See Re, supra note 59, at 417 (arguing “purposive and pragmatic considerations” inform the Court’s “identification of textual clarity or ambiguity” and then in turn resolve those ambiguities). and longstanding historical practice.^62×62. See Curtis A. Bradley & Neil S. Siegel, After Recess: Historical Practice, Textual Ambiguity, and Constitutional Adverse Possession, 2014 Sup. Ct. Rev. 1, 45–47 (arguing historical practice contributes to ambiguity in constitutional text and favors readings consistent with practice); see also Doerfler, supra note 9 (manuscript at 16–17) (evaluating extension of Bradley & Siegel’s historical practice arguments from constitutional interpretation to statutory interpretation). SW General challenges the explanatory power of those theories because their key elements were arguably just as present in the FVRA dispute as when the Court has favored purpose over text in the past. If the Court does treat some statutory cases differently, then, it should clarify what elements suffice to shade otherwise clear texts with ambiguity. The current doctrinal puzzle leaves the Court with problematic discretion to choose which cases it considers “major.”

One account of the Court’s willingness to favor purpose over text looks to whether the apparent plain meaning of the statute might thwart a congressional plan. Professor Abbe Gluck argues judicial sensitivity to such plans “has appeared in more than 100 cases in the U.S. Reports”^63×63. Gluck, supra note 60, at 81. and that it makes sense of the Court’s opinion in King v. Burwell.^64×64. 135 S. Ct. 2480 (2015); see Gluck, supra note 60, at 87. Rather than a run-of-the-mill purposivism looking for congressional intent in extratextual sources, the idea is that a court can perceive a statute’s own objective “from [its] . . . interlocking textual components, not from legislative history or other subjective tools.”^65×65. Gluck, supra note 60, at 88. In King, the Court hewed closely to “a fair understanding of the legislative plan” to “improve health insurance markets” in holding the Affordable Care Act authorizes tax credits for federally established insurance exchanges.^66×66. King, 135 S. Ct. at 2496. It did so in spite of the petitioners’ admittedly “strong”^67×67. Id. at 2495. plain meaning reading that would likely have caused the “‘death spirals’ that Congress designed the Act to avoid.”^68×68. Id. at 2493.

SW General is similar in this regard, in that objective evidence of a congressional plan also emerges from the internal fit of the FVRA. As with the Affordable Care Act, Congress enacted the FVRA to address a specific and recurring problem: agencies need effective interim leadership, but Congress and the President may disagree on proper leaders.^69×69. The Court noted as much in its recitation of the interbranch conflicts that precipitated the FVRA. See SW General, 137 S. Ct. at 934–36. The FVRA strikes a balance by giving the President three categories from which to choose acting officers,^70×70. See 5 U.S.C. § 3345(a)(1)–(3)(2012) (first assistants, officials in other positions requiring Senate confirmation, and high-ranking employees). but for each it provides a check: Senate-confirmed officials, by definition, have some congressional imprimatur. And the high-ranking employee category contains its own requirement of ninety days prior service.^71×71. See id. § 3345(a)(3). That leaves first assistants, for whom the ninety-day requirement in the disputed limitation provides a bona fide first assistant requirement of sorts. Since the other two categories already have their own internal checks, the internal fit of the statute suggests the limitation need apply only to first assistants.^72×72. That Senate confirmation is the ultimate check on objectionable acting officers is confirmed by section 3345(b)(2)’s exception to the disputed restriction: the bona fide first assistant requirement does not apply to first assistants whose own roles required Senate confirmation. See id. § 3345(b)(2). But here, unlike in King, plain meaning led the Court to hamper the congressional plan — imposing the bona fide first assistant requirement on officials already held sufficiently accountable by the statute. Thus an account of the Court’s occasional purposivism as respecting the internal fit of statutes currently suffers from an inability to identify just how much “fit” is sufficient.^73×73. Accordingly, Gluck notes that the explanatory power of the congressional plan view may not hold for “smaller disputes,” and that “King may have been an easier case for use of the plan concept than the typical statutory interpretation dispute” because of the centrality of the legal question to the overall scheme. Gluck, supra note 60, at 96. If certain congressional plans do make the difference for the Court while others do not, that hierarchy is unclear for now.

A more obvious metric for high-stakes cases where purposivism might thrive is “pragmatic considerations.”^74×74. Re, supra note 59, at 417. Perhaps the Court is more likely to find ambiguity and accept purposive readings of statutes where the alternative would have sweeping consequences.^75×75. See id. (describing practical considerations and purpose as playing roles in initial determination of textual ambiguity under what Professor Richard Re has called a “New Holy Trinity” doctrine). Here, SW General and King would seem to part ways. As important as it is for the executive branch to have the most skilled people serving in acting roles while nominated, it is easy to argue that the life or death of the Affordable Care Act is of greater weight.^76×76. But see SW General, 137 S. Ct. at 947 (Thomas, J., concurring) (“The general counsel is effectively the Nation’s labor-law prosecutor . . . .”).

Yet that claim is harder to make with other key purposivism cases. At risk in Yates v. United States^77×77. 135 S. Ct. 1074 (2015). was expansive liability under a provision of the Sarbanes-Oxley Act.^78×78. 18 U.S.C. § 1519 (2012); see Yates, 135 S. Ct. at 1081 (plurality opinion). A plurality of the Court held that a criminal prohibition on destroying “any record, document, or tangible object”^79×79. Yates, 135 S. Ct. at 1078 (plurality opinion) (quoting 18 U.S.C. § 1519). to impede an investigation did not apply to the defendant’s throwing undersized fish overboard to avoid a penalty for his illicit catch.^80×80. Id. at 1078–79. The plurality credited the section’s caption and the context of corporate fraud that prompted Sarbanes-Oxley over the dictionary definition of “tangible object” as anything having physical form.^81×81. See id. at 1079, 1081–83 (citing Webster’s Third New International Dictionary 1555 (2002) (“object”) and Tangible, Black’s Law Dictionary (10th ed. 2014)). It is difficult to compare the stakes in Yates with those in SW General because the potential consequences were different in kind. Looming in one was an additional felony conviction and an expansive view of Sarbanes-Oxley liability, while the other has eliminated the possibility of many executive nominees leading their departments while the Senate deliberates. But that difficulty demonstrates the current limitations of a consequence-based approach to adopting purposivist readings in certain cases. Unless the Court unpacks how to prioritize qualitatively different outcomes, this theory sheds little light on what stakes are sufficiently high to open the purposivist door.

Professors Curtis Bradley and Neil Siegel argue that displacement of historical practice affects whether certain constitutional texts are ambiguous and can be construed on non-textualist grounds.^82×82. See Bradley & Siegel, supra note 62, at 45–47; see also Doerfler, supra note 9, (manuscript at 16–17) (considering extension of Bradley and Siegel’s argument to statutory interpretation). Perhaps conscious of this element in the statutory context, the Court took care in SW General to distinguish the history surrounding the FVRA from the historical pedigree of the recess appointments practice considered in Noel Canning.^83×83. See SW General, 137 S. Ct. at 943. The comparison between the two is natural for major cases analysis because the stakes are so similar: both recess and acting service appointments powers derive from texts regulating how the President orders the heads of executive departments. In Noel Canning, the Court relied on historical practice dating back to the Founding era to hold, inter alia, that “all vacancies” in the Recess Appointments Clause includes vacancies that arise while the Senate is still in session.^84×84. NLRB v. Noel Canning, 134 S. Ct. 2550, 2570–73 (2014). For their part, the opinions in SW General disclosed three different views of the importance of historical practice. The majority and dissent divided on the weight of two decades of congressional silence, while Justice Thomas’s constitutional analysis left historical practice largely untouched.^85×85. See SW General, 137 S. Ct. at 943; id. at 953–54 (Sotomayor, J., dissenting); id. at 945–49 (Thomas, J., concurring). It is certainly true that historical practice brings less to bear on a text written in 1998 than one drafted in 1787. But the line-drawing problem inherent in the idea that displacing historical practice can trigger purposivism suggests that approach will remain unhelpful until the Court clearly identifies what amount of practice counts. If historical practice indeed does raise the stakes of cases, the divided approaches to that practice in SW General reveal enormous discretion currently residing in the Court’s hands.

One could argue that various theories about purposivism in the Roberts Court are simply irrelevant to SW General because the language in the FVRA limitation was so clear that plain meaning had to control. The trouble here is that the Court has found ambiguity and favored purpose in cases with even more textual clarity. In Bond v. United States,^86×86. 134 S. Ct. 2077 (2014). the Court found ambiguous a criminal prohibition on using “chemical weapon[s],” as applied to a woman who used toxic chemicals in hopes of giving her husband’s mistress a rash.^87×87. Id. at 2085, 2090 (quoting 18 U.S.C. § 229(a)(1) (2012)). The statute defined “chemical weapon” as “[a] toxic chemical and its precursors, except where intended for a purpose not prohibited under this chapter.”^88×88. 18 U.S.C. § 229F(1)(A). “Toxic chemical” was also defined in the statute broadly enough to include the materials the defendant used. See id. § 229F(8)(A). Though the defendant’s activity seemed contained within the statute’s plain meaning, the Court found the provision ambiguous in light of “the deeply serious consequences of adopting such a boundless reading” and “the context from which the statute arose.”^89×89. Bond, 134 S. Ct. at 2090. On a spectrum of textual clarity, the language in the Bond statute is about as clear as a provision can be: Congress used specific multi-tiered definitions, ensuring its chemical weapons prohibition covered the field.^90×90. See id. at 2094 (Scalia, J., concurring in the judgment) (describing 18 U.S.C. § 229F). The Court’s fairly strong plain meaning arguments in SW General, then, do not make the case irrelevant for evaluating major cases theories. If anything, the differing responses to textual clarity in SW General and Bond underscore the need for better-refined explanations by the Court of when it embraces purpose over text.

This is not to say SW General disproves current accounts of the Court’s occasional purposivist leanings or that the Court has been inconsistent. If the major cases idea does describe the Court’s reasoning, it could be that the facts of SW General simply reflect a lower bound for each proposed element: it takes more objective evidence of congressional purpose, more sweeping practical consequences, or more historical practice before purpose ascends. And filling out doctrine takes time for a Court limited to resolving “Cases” and “Controversies.”^91×91. U.S. Const. art. III, § 2, cl. 1; see also Gluck, supra note 60, at 109 (“There is a sense that the Court is still working out its institutional position in the current statutory landscape.”). But if some uncertainty simply comes with the territory of finding the threshold for high stakes, it is magnified where high-stakes elements coalesce in the same case, yet plain text ultimately governs. Greater uncertainty in doctrine locates greater discretion with the Court.^92×92. See, e.g., Antonin Scalia, The Rule of Law as a Law of Rules, 56 U. Chi. L. Rev. 1175, 1177 (1989) (describing less predictable totality-of-the-circumstances approaches to judicial decisionmaking as “discretion-conferring”). As a case with many features of the Court’s prior purposivist moves, SW General reveals just how broad that discretion currently is. The normative drawbacks to this state of affairs are twofold: Discretion in the courts creates less predictability in the law, as legislators and the people who elect them cannot foresee how statutes will be interpreted.^93×93. See Note, supra note 59, at 1247; cf. William Baude & Ryan D. Doerfler, The (Not So) Plain Meaning Rule, 84 U. Chi. L. Rev. 539, 559 (2017) (arguing inconsistent thresholds of “plainness” for plain meaning approaches to statutory interpretation “can . . . exacerbate unpredictability”). Further, discretion in decisionmaking can even lend itself to judicial picking-and-choosing where the relevant issue is simply the importance of each case.^94×94. See Note, supra note 59, at 1247 (“With [textualism] uprooted, Justices may choose their methodology based on how important a given case is to them and which method suits their preferred outcome.”); cf. Brett M. Kavanaugh, Fixing Statutory Interpretation, 129 Harv. L. Rev. 2118, 2142–43 (2016) (reviewing Robert A. Katzmann, Judging Statutes (2014)) (arguing that inherent arbitrariness in determinations of ambiguity and clarity in statutory interpretation “threatens to undermine the stability of the law and the neutrality (actual and perceived) of the judiciary,” id. at 2143).

Plotting the line of best fit through cases where the Court chooses purpose over plain text is incremental by nature. SW General provides an opportunity to continue that project, and it does so by revealing the present shortcomings of some recent theories about what might make cases sufficiently “major.” If those theories are nonetheless roughly correct, broad discretion will remain with the Court until it further defines what counts when it comes to congressional plans, practical consequences, and historical practice. If the Roberts Court does in fact approach major cases with special solicitude for legislative purpose, it ought to continue refining the elements that raise the stakes.