When a court strikes down part of a statute as unconstitutional, it must decide what becomes of the wreckage. If the court determines “Congress would have enacted the other provisions in the statute without the unconstitutional provision,” the court may cut out the unconstitutional portion and leave the rest of the statute on the books.1×1. Rachel J. Ezzell, Note, Statutory Interdependence in Severability Analysis, 111 Mich. L. Rev. 1481, 1485 (2013) (citing Alaska Airlines, Inc. v. Brock, 480 U.S. 678, 684–85 (1987)). It is worth noting that there is some debate over whether severability operates at the level of construction or remedy. See Lisa Marshall Manheim, Beyond Severability, 101 Iowa L. Rev. 1833, 1844 (2016). Yet this determination may raise more questions than it answers: among them, how should judges choose between equally curative but different ways of repairing a constitutional defect? The Supreme Court has suggested that severability solutions that require too much “editorial freedom” are impermissible because to creatively edit a statute is to legislate.2×2. Free Enter. Fund v. Pub. Co. Accounting Oversight Bd., 561 U.S. 477, 510 (2010). Recently, in New Mexico v. Department of the Interior,3×3. 854 F.3d 1207 (10th Cir. 2017). the Tenth Circuit invoked this rhetoric to reject as too creative a line edit of the Indian Gaming Regulatory Act4×4. 25 U.S.C. §§ 2701–2721 (2012). (IGRA) that would have allowed the Department of the Interior (DOI) to prescribe gaming-approval procedures to replace an unconstitutional statutory procedure. Treating anti-editorial-freedom rhetoric as if it were a rule caused the Tenth Circuit to summarily reject a solution that may have avoided the ultimate outcome of the case: a balancing of tribal and state interests far removed from the one Congress enacted. A rejection of editorial freedom, in other words, pressured the court into disrupting a legislative scheme. This result suggests that a strict rule against editorial freedom might, in practice, betray its promise of safeguarding against judicial intrusion into the policymaking sphere.
In clashes between Indian tribes that wish to operate gaming enterprises and states that wish to regulate gaming within their borders, Congress plays referee. Sovereign tribes have the right to conduct gaming on reservations; states may not regulate tribal gaming absent the consent of the federal government.5×5. California v. Cabazon Band of Mission Indians, 480 U.S. 202, 207–10 (1987). In 1988, in an attempt to balance these interests,6×6. Specifically, Congress sought to “balance the need for sound enforcement of gaming laws and regulations with the strong Federal interest in preserving the sovereign rights of tribal governments.” S. Rep. No. 100-446, at 5 (1988), as reprinted in 1988 U.S.C.C.A.N. 3071, 3075. Congress passed IGRA. IGRA subjected gaming activity to the terms of tribe-state compacts.7×7. See New Mexico v. Dep’t of the Interior, No. 1:14-cv-00695, 2014 WL 10298036, at *1 (D.N.M. Oct. 17, 2014). To ensure that a “recalcitrant state”8×8. New Mexico, 854 F.3d at 1234. could not block gaming by simply refusing to negotiate a compact, IGRA imposed upon states a duty to negotiate in good faith.9×9. Id. at 1212 (citing 25 U.S.C. § 2710(d)(3)(A)). This duty had teeth: a tribe could sue to enforce the duty in federal court.10×10. Id. at 1211. A judicial finding of lack of good faith triggered a court-supervised mediation process.11×11. Id. at 1212. If the state still declined to consent to a compact, DOI would prescribe procedures to govern gaming.12×12. Id. In short, IGRA gave tribes a means of realizing their right to gaming even in the face of a state that refused to negotiate.
The Supreme Court upset this balance: in Seminole Tribe of Florida v. Florida,13×13. 517 U.S. 44 (1996). the Court held that IGRA violated the Eleventh Amendment by subjecting states to suit without their consent.14×14. Id. at 47; see also New Mexico, 854 F.3d at 1212. The Court in Seminole Tribe “expressly declined to consider the extent to which the portion of the statute concerning . . . gaming procedures remained intact following its holding.” Id. at 1231 (citing Seminole Tribe, 517 U.S. at 76 n.18). Now, by refusing to negotiate and asserting a sovereign immunity defense, a state could thwart a tribe’s gaming efforts. DOI attempted to recalibrate the scales by issuing regulations to provide tribes with an administrative remedy to substitute for the unconstitutional court-supervised remedy. Under 25 C.F.R. § 291, when a tribe’s suit against a recalcitrant state is dismissed on sovereign immunity grounds, the tribe may ask DOI to initiate a mediation process that mimics IGRA’s court-supervised process.15×15. 25 C.F.R. § 291.1–.11 (2016). The rule thus affords a tribe a path to gaming even when a state asserts a sovereign immunity defense.
In 2013, the Pueblo of Pojoaque tribe requested to negotiate a gaming compact with New Mexico.16×16. New Mexico v. Dep’t of the Interior, No. 1:14-cv-00695, 2014 WL 10298036, at *4 (D.N.M. Oct. 17, 2014). Over six months later, New Mexico had not agreed to negotiate.17×17. Id. The Pueblo brought suit in the District of New Mexico, alleging lack of good faith.18×18. Id. at *5. New Mexico asserted sovereign immunity; the court dismissed the case.19×19. Id. The Pueblo turned to 25 C.F.R. § 291: they proposed a compact to DOI, which in turn found the Pueblo eligible to establish gaming and requested comment from New Mexico.20×20. Id. During the comment period, New Mexico sued DOI, alleging that 25 C.F.R. § 291 exceeded DOI’s statutory authority and seeking an injunction prohibiting DOI from issuing gaming procedures for the Pueblo.21×21. New Mexico, 854 F.3d at 1214. The Pueblo successfully moved to intervene as a defendant.22×22. Id. The district court granted New Mexico’s summary judgment motion, finding under step one of the Chevron23×23. Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837 (1984). framework that the regulation contradicted IGRA’s clear language by allowing DOI to impose gaming procedures in the absence of a federal court finding of lack of good faith.24×24. New Mexico, 2014 WL 10298036, at *14. The court thus enjoined DOI from issuing gaming procedures for the Pueblo. Id. at *16.
The Tenth Circuit affirmed.25×25. New Mexico, 854 F.3d at 1211. Writing for the panel, Judge Holmes26×26. Judge Holmes was joined by Judges McHugh and Moritz. concluded that New Mexico had standing — it had suffered a procedural injury by being denied the right to a judicial determination of lack of good faith prior to DOI involvement in setting up gaming procedures.27×27. New Mexico, 854 F.3d at 1216–17. Alternatively, it had suffered an injury in the form of 25 C.F.R. § 291’s “forced choice”: either participate in the DOI process or lose the opportunity to provide input regarding the content of gaming procedures. Id. at 1218. The court further found New Mexico’s challenge to the § 291 process to be ripe. Id. at 1219–21. Turning to the heart of New Mexico’s argument, the court applied Chevron to determine that IGRA did not authorize 25 C.F.R. § 291.28×28. Id. at 1223. The explicit text of IGRA permitted DOI to issue gaming procedures only after a court determined that a state had conducted itself in bad faith and a court-appointed mediator had selected a compact to which the state had failed to consent.29×29. Id. at 1224–25. This “specific and detailed” process left no ambiguity that might permit DOI issuance of “gaming procedures outside of” these “narrow circumstances.”30×30. Id. at 1224. DOI’s rule allowed the agency to issue procedures outside these circumstances — it allowed DOI to appoint a mediator and issue gaming procedures absent a finding of bad faith — so it failed at Chevron step one.31×31. Id. at 1224–26 (citing 25 U.S.C. § 2710(d)(7)(A), (d)(7)(B)(iv) (2012)).
Finally, the court addressed an argument the Pueblo raised for the first time on appeal: stripped of the requirement that states be subject to suit by the tribes (gone after Seminole Tribe) and of the substitute regulatory remedy (gone as a result of this case), IGRA was now a statute Congress would not have enacted in the first place — it provided tribes no remedy to deal with recalcitrant states, and was thus at odds with the state-tribe balance Congress so carefully sought to strike.32×32. See Pueblo of Pojoaque’s Opening Brief at 21–24, New Mexico, 854 F.3d 1207 (10th Cir. 2017) (Nos. 14-2219, 14-2222), 2015 WL 1138164, at *21–24. The Pueblo argued that the court should engage in severability analysis in order to blue-pencil from the statute the provisions that made 25 C.F.R. § 291 problematic; that is, those that were premised on the ability of tribes to hale states into court. The Pueblo’s brief offered an edited version of the statute that struck out, among other things, provisions requiring (1) a court finding of bad faith prior to DOI involvement and (2) a failure of the parties to produce a compact within sixty days after such a finding before a mediator could be appointed.33×33. See id. at 25–28., 2015 WL 1138164, at *25–28. This way, the court could leave in place the possibility of an administrative remedy to preserve the tribe-state balance Congress sought to enact.
The court refused to conduct this “elaborate editing of IGRA”34×34. New Mexico, 854 F.3d at 1233. because “such editorial freedom . . . belongs to the Legislature, not the Judiciary.”35×35. Id. at 1234 (quoting Free Enter. Fund v. Pub. Co. Accounting Oversight Bd., 561 U.S. 477, 510 (2010)). The only options before the court were to leave the statute on the books absent the constitutionally offensive suit-subjecting provision or strike down the entire statute.36×36. Id. (“IGRA’s jurisdiction-granting clause falls alone or it falls with all of IGRA.”). Per the Pueblo, this choice put the court in a bind: intrude upon congressional intent by striking down all of IGRA, or intrude upon congressional intent by leaving on the books a hobbled statute Congress would not have enacted on its own.37×37. Id. Moreover, IGRA had a severability clause, which suggested a congressional preference for seeing an unconstitutional provision removed over seeing the entire statute fall. Id. at 1233. The court purported to find a way out of this bind. It determined that, even severed of the court-haling provision and in the absence of 25 C.F.R. § 291, “[IGRA] is capable of operating consistently with Congress’s intent,”38×38. Id. at 1234. because, contra the Pueblo, Seminole Tribe did not leave tribes wholly without an IGRA remedy. First, the court argued, states can categorically waive their sovereign immunity for IGRA purposes, effectively reinstating the pre–Seminole Tribe remedy.39×39. Id. at 1234–35. Second, states retain the option of waiving sovereign immunity in an individual IGRA case.40×40. Id. at 1235. Finally, the United States may sue a state as a tribal trustee, “sidestepping the sovereign immunity defense.”41×41. Id. Thus, even absent the 25 C.F.R. § 291 remedy, the court could avoid both striking down the whole statute and leaving on the books a statute that upset the tribe-state balance Congress crafted.
Severability doctrine has long instructed courts to “seek to determine what Congress would have intended in light of the Court’s constitutional holding.”42×42. Nat’l Fed’n of Indep. Bus. v. Sebelius, 567 U.S. 519, 586 (2012) (plurality opinion) (quoting United States v. Booker, 543 U.S. 220, 246 (2005)). The New Mexico panel’s summary rejection of the Pueblo’s line edits exhibits the influence of newer calls for judges, in order to avoid impermissible judicial “redraft[ing]” of legislation,43×43. David H. Gans, Severability as Judicial Lawmaking, 76 Geo. Wash. L. Rev. 639, 643 (2008). to steer clear of “editorial freedom”44×44. Free Enter. Fund v. Pub. Co. Accounting Oversight Bd., 561 U.S. 477, 510 (2010). when deciding what is to become of a statute afflicted with an unconstitutional provision. But New Mexico demonstrates how the rejection of editorial freedom may fall short of — and even hinder — such rejection’s laudable separation of powers goals. The panel’s rejection of the Pueblo’s edits as too creative failed to stop it from “legislating” by imposing a tribe-state balancing scheme quite at odds with the one Congress initially enacted. Worse, by treating the anti-editorial-freedom rhetoric of Free Enterprise Fund v. Public Company Accounting Oversight Board45×45. 561 U.S. 477. as an absolute rule, the New Mexico panel robbed itself of the chance to fully evaluate an option that may have allowed for greater harmony with Congress’s original scheme. To avoid unnecessary judicial disruption of statutory schemes, courts should, at most, treat complex severability solutions with skepticism, rather than dismissing them out of hand.
Under severability doctrine, courts look to congressional intent to determine which combination of statutory lines should be struck to cure the constitutional defect,46×46. A complex statute may present a court with “alternative route[s] to a constitutionalizing result.” Kenneth A. Klukowski, Severability Doctrine: How Much of a Statute Should Federal Courts Invalidate?, 16 Tex. Rev. L. & Pol. 1, 73 (2011); see, e.g., Free Enter., 561 U.S. at 509 (“[T]he language providing for good-cause removal is only one of a number of statutory provisions that, working together, produce a constitutional violation.”). and whether, given the invalidity of a provision found to be unconstitutional, other provisions must fall as well. The court may strike other provisions if (and only if) “it is evident that the Legislature would not have enacted those provisions . . . independently of” the unconstitutional provision.47×47. Free Enter., 561 U.S. at 509 (omission in original) (quoting New York v. United States, 505 U.S. 144, 186 (1992)); see also PHH Corp. v. Consumer Fin. Prot. Bureau, 839 F.3d 1, 37 (D.C. Cir. 2016) (Kavanaugh, J.) (admitting that severability doctrine commands judges to “speculate” with respect to congressional intent), rev’d en banc, 881 F.3d 75 (D.C. Cir. 2018). In other words, courts must engage in the “inherently counterfactual” exercise of determining whether Congress would have left provisions out of the statute (or passed it at all) had it known the unconstitutional provision was un-available.48×48. Kevin C. Walsh, Partial Unconstitutionality, 85 N.Y.U. L. Rev. 738, 744 (2010). Importantly, this unique inquiry into “what the legislature would have done, not what the legislature actually did,” id., does not submit easily to the traditional textualist objection that the only discoverable legislative intent is the “‘objectified’ intent” manifested in the text of a statute Congress actually did pass, Antonin Scalia, Common-Law Courts in a Civil-Law System: The Role of United States Federal Courts in Interpreting the Constitution and Laws, in A Matter of Interpretation 3, 17 (Amy Gutmann ed., 1997). Perhaps this is why Justice Scalia, along with his colleagues, “repeatedly . . . acknowledged the central role that legislative intent plays in severability analysis.” Manheim, supra note 1, at 1840. Any severability exercise that results in a regime Congress would not have enacted thus fails on the terms of the doctrine.49×49. The panel agreed with this premise. New Mexico, 854 F.3d at 1234 (describing its inquiry as “looking to whether Congress would prefer no IGRA to its current state and whether, in its current state, it is capable of operating consistently with Congress’s intent”).
When it cited Free Enterprise Fund to reject the Pueblo’s suggestions as impermissible “judicial editing,”50×50. Id. at 1234. the New Mexico panel marched in step with a trend toward rejecting severability remedies that employ too much “complexity or creativity.”51×51. Manheim, supra note 1, at 1848. This trend has played a significant role in such prominent recent decisions as National Federation of Independent Business v. Sebelius, 567 U.S. 519 (2012), see Manheim, supra note 1, at 1849 (noting Justice Scalia’s questions at oral argument), and PHH Corp. v. Consumer Financial Protection Bureau, 839 F.3d at 37–39. Importantly, however, Free Enterprise Fund did not announce a general rule against judicial editing of statutes; it merely rejected a complicated set of edits in favor of a simpler one (eliminating good-cause removal protection). See Free Enter., 561 U.S. at 509–10. At its core, the emerging disapproval of editorial freedom seeks to ensure that judges do not legislate.52×52. See Gans, supra note 43, at 643 (“[An] editorial role . . . requires courts to redraft legislation without any effective limits on the power of the pen or any of the tools that make for sound legislation. . . . The end result is lawmaking with a democratic deficit.”). By cabining judges’ ability to “rewrite” a statute to adjust for an unconstitutional provision, the anti-editorial-freedom approach promises to limit judges’ options so that Congress, not a court, performs any necessary rewrite of a statute plagued with constitutional problems.53×53. See Free Enter., 561 U.S. at 510 (“[S]uch editorial freedom . . . belongs to the Legislature, not the Judiciary. Congress of course remains free to pursue any of these options going forward.”); PHH Corp., 839 F.3d at 39 (arguing that editorial freedom takes courts “far beyond [the] judicial capacity”).
But in New Mexico, this approach to editorial freedom failed to stop a judicial upheaval of a carefully crafted legislative regime. The New Mexico panel invoked Free Enterprise Fund to categorically dismiss the blue-pencilling of the statute the Pueblo argued would save Congress’s intended tribal-state balance. The suggestion was complicated enough to constitute “judicial editing”;54×54. New Mexico, 854 F.3d at 1234. any potential it held for limiting the judges’ policy impact on the IGRA scheme could not be tapped. This left the court in a tight spot. Either strike down the statute — and deny states the “subordinate but significant role in regulating tribal gaming” IGRA gave them55×55. Texas v. United States, 497 F.3d 491, 494 (5th Cir. 2007). — or leave in place a tribal-state balance the court acknowledged might weaken the tribal “bargaining position” enough to “stymie Congress’s intent in IGRA.”56×56. New Mexico, 854 F.3d at 1234. The panel purported to evade this outcome by concluding that IGRA, even without the suit-granting provision and without 25 C.F.R. § 291, could preserve the intended balance because states can waive sovereign immunity for IGRA claims (categorically or in individual cases) and the United States can bring suits on behalf of tribes.57×57. Id. at 1234–35. But none of these measures construes the statute in a way that even approximates the balance that existed prior to Seminole Tribe — the balance Congress enacted in the first place. Sovereign immunity waiver is hardly a satisfactory means of dealing with states that refuse to cooperate: it depends on the cooperation of states.58×58. Moreover, the court was able to point to only one example of categorical waiver and one example of waiver for an individual IGRA claim having occurred. Id. Reliance on the United States is similarly unhelpful because there is no guarantee the United States will intervene, much less that it will side with tribes instead of states.59×59. Cf., e.g., United States v. Spokane Tribe of Indians, 139 F.3d 1297, 1298, 1302 (9th Cir. 1998) (denying the United States’ request for an order enjoining the Spokane Tribe of Indians from conducting gaming operations). Thus, the court left in its wake a regime Congress would not have enacted: one that conditions the tribal right to gaming on the conclusion of a compact without handing tribes a corresponding means of incentivizing states to negotiate a compact.60×60. See S. Rep. No. 100-446, at 13 (1988), as reprinted in 1988 U.S.C.C.A.N. 3071, 3083 (“The practical problem in formulating statutory language to accomplish the desired result is the need to provide some incentive for States to negotiate with tribes in good faith because tribes will be unable to enter into such gaming unless a compact is in place.”). With its hands tied by its treatment of rhetoric against editorial freedom as a rule, the court could not even explore an option that may have more faithfully adhered to severability doctrine’s command to approximate a scheme Congress would have enacted.
This realization points up two problems with the anti-editorial-freedom approach. First, by sanctioning the establishment of a scheme out of step with congressional intent, the avoidance of editorial freedom failed to achieve its goal of stopping judges from legislating. Worse, it made such an outcome more likely by removing a more intent-tracking option from the table. Instead of dismissing the Pueblo’s suggestion as violating a perceived rule against editorial freedom, the court should have weighed it against the alternatives (striking IGRA altogether or leaving an IGRA with weakened tribal remedies) according to the key metric of the severability framework: “what Congress would have intended in light of” the constitutional problem.61×61. Nat’l Fed’n of Indep. Bus. v. Sebelius, 567 U.S. 519, 586 (2012) (plurality opinion) (quoting United States v. Booker, 543 U.S. 220, 246 (2005)). Perhaps the court would have concluded that editing the statute to support a specific regulatory remedy was more divorced from congressional intent than the alternatives. Perhaps not. Yet treating anti-editorial rhetoric as a rule precluded this inquiry. That the anti-editorial-freedom approach prevents judges from evaluating options that may limit disruption of statutory regimes undermines its promise to patrol the line separating judging from legislating. Some scholars argue that limiting editorial freedom keeps judges out of the legislation business by ensuring that Congress, instead of a court, answers the complicated policy question of what should become of a severed statute that is, by definition, not the same statute Congress enacted.62×62. Gans, supra note 43, at 643–44. But this approach forces disruptive outcomes and places the burden on Congress to pick up the pieces. Until Congress does react to a severability decision, legislative supremacy is best served by a regime that allows courts to avoid interfering with policy determinations Congress has already made. Perhaps then, as Professor Lisa Marshall Manheim suggests, severability doctrine should be abandoned altogether because it limits the tools with which courts can engage in a “broader inquiry into legislative intent.”63×63. Manheim, supra note 1, at 1892. A more modest solution would work within existing severability doctrine by encouraging courts to treat a proposed severability fix that seems to require creativity with caution — complex editing may be a hallmark of a solution that does not track legislative intent64×64. It is reasonable to presume that Congress usually “does not intend wholesale revisions of a statute in response to that statute’s constitutional disruption.” Id. at 1887. — rather than categorically dismissing it. This approach would afford courts greater flexibility to achieve what severability doctrine demands: a statutory regime Congress would have chosen to enact had it known the unconstitutional provision was unavailable.
The realization that a prohibition on editorial freedom in severability analysis permits, and may encourage, rulings that leave in their wake statutory regimes Congress would not have enacted should give pause to anyone concerned about judicial intrusion into the legislative sphere. At the very least, courts should not adopt New Mexico’s approach of treating editorial freedom as an absolute disqualifier. An approach that restrains the judicial pen but sanctions great disruption of policy regimes is a hollow victory for the separation of powers.