`
`
`
`In the
`United States Court of Appeals
`For the Seventh Circuit
`____________________
`
`No. 17-2428
`PLANNED PARENTHOOD OF INDIANA
`AND KENTUCKY, INC.,
`
`Plaintiff-Appellee,
`
`v.
`
`KRISTINA BOX, Commissioner,
`Indiana State Department of Health, et al.,
`Defendants-Appellants.
`____________________
`
`Appeal from the United States District Court for the
`Southern District of Indiana, Indianapolis Division.
`No. 1:17-cv-01636-SEB-DML — Sarah Evans Barker, Judge.
`____________________
`
`ON REMAND FROM THE SUPREME COURT OF THE UNITED STATES
`DECIDED MARCH 12, 2021
`____________________
`
`Before KANNE, ROVNER, and HAMILTON, Circuit Judges.
`HAMILTON, Circuit Judge. This appeal returns to us on re-
`mand from the Supreme Court of the United States. In 2019,
`we affirmed the district court’s grant of a preliminary injunc-
`tion against enforcement of a new Indiana statutory re-
`striction on minors’ access to abortions. See Planned
`
`

`

`2
`
`No. 17-2428
`
`Parenthood of Indiana & Kentucky, Inc. v. Adams, 258 F. Supp. 3d
`929 (S.D. Ind. 2017), aff’d, 937 F.3d 973 (7th Cir. 2019), reh’g de-
`nied, 949 F.3d 997 (7th Cir. 2019). The State defendants peti-
`tioned for a writ of certiorari. The Supreme Court granted the
`petition, vacated our decision, and remanded for further con-
`sideration in light of June Medical Services LLC v. Russo, 140 S.
`Ct. 2103 (2020), which struck down a Louisiana law regulat-
`ing abortion providers, but without a single majority opinion.
`We apply the predominant and most sound approach to
`the “narrowest ground” rule in Marks v. United States, 430 U.S.
`188 (1977), for assessing the precedential force of Supreme
`Court decisions issued without a majority opinion. The opin-
`ions in June Medical show that constitutional standards for
`state regulations affecting a woman’s right to choose to termi-
`nate a pregnancy are not stable, but they have not been
`changed, at least not yet, in a way that would change the out-
`come here.
`The Chief Justice’s concurring opinion in June Medical of-
`fered the narrowest basis for the judgment in that case, giving
`stare decisis effect to Whole Woman’s Health v. Hellerstedt, 136
`S. Ct. 2292 (2016), on the essentially identical facts in June Med-
`ical. The Marks rule does not, however, turn everything the
`concurrence said—including its stated reasons for disagree-
`ing with portions of the plurality opinion—into binding prec-
`edent that effectively overruled Whole Woman’s Health. That is
`not how Marks works. It does not allow dicta in a non-major-
`ity opinion to overrule an otherwise binding precedent. We
`applied those binding standards from Whole Woman’s Health
`in our earlier decision, and that decision has not been over-
`ruled by a majority decision of the Supreme Court. We there-
`fore again affirm the district court’s preliminary injunction
`
`

`

`No. 17-2428
`
`3
`
`barring enforcement of the challenged law pending full re-
`view in the district court.
`I. Factual and Procedural Background
`Given the lengthy opinions already issued in this case, we
`summarize the issues leading up to this point. Indiana’s Sen-
`ate Enrolled Act 404, enacted in 2017, included amendments
`to Indiana’s judicial-bypass process. That process, required
`by Bellotti v. Baird, 443 U.S. 622 (1979), creates a narrow legal
`path for an unemancipated minor to obtain an abortion with-
`out parental consent. The minor must first find her way to a
`state trial court. She must then obtain a court order finding
`either that the abortion would be in her best interests or that
`she is sufficiently mature to make her own decision. Ind. Code
`§ 16-34-2-4(e). Senate Enrolled Act 404 amended the process
`in several ways, some of which the district court preliminarily
`enjoined. Only one amendment is at issue in this appeal: a
`new requirement that a minor’s parents be notified that she is
`seeking an abortion through the bypass procedure—unless
`the judge finds that such parental notice, as distinct from re-
`quiring parental consent, is not in the minor’s best interests.
`Ind. Code § 16-34-2-4(d). Maturity does not affect the new no-
`tice requirement.
`To support its motion for preliminary injunction, plaintiff
`offered evidence on the likely effects of the new notice re-
`quirement. The evidence took the form of affidavits from
`seven witnesses familiar with the actual workings of the judi-
`cial bypass process and the situations of and stresses upon
`minors seeking abortions or advice on abortions. The State de-
`fendants chose not to offer evidence at that stage of the case.
`They also did not challenge the reliability or credibility of
`plaintiff’s evidence.
`
`

`

`4
`
`No. 17-2428
`
`The district court issued detailed findings of fact and con-
`clusions of law finding that the new notice requirement was
`likely to impose an undue burden on the right to obtain an
`abortion for a significant fraction of minors for whom the re-
`quirement would be relevant. 258 F. Supp. 3d 929, 939–40. We
`affirmed, emphasizing the lopsided evidence showing both
`the likely burden and the absence of appreciable benefit from
`the new notice requirement. 937 F.3d at 989–90. We relied
`heavily on Whole Woman’s Health, guided by its application of
`the “undue burden” standard adopted in Planned Parenthood
`of Southeastern Pennsylvania v. Casey, 505 U.S. 833 (1992). We
`also relied on Whole Woman’s Health’s approval of a pre-en-
`forcement injunction against challenged laws likely to impose
`an undue burden. 937 F.3d at 979−80.
`In Whole Woman’s Health, the Supreme Court affirmed a
`district court decision striking down a so-called admitting
`privileges requirement. The challenged Texas law required a
`physician who performed an abortion to have admitting priv-
`ileges at a hospital within thirty miles of the abortion site. The
`Supreme Court based its decision on detailed factual findings
`showing both the burdens imposed by that requirement and
`the lack of accompanying benefits. 136 S. Ct. at 2310–14.
`In June Medical in 2020, the Court held unconstitutional a
`Louisiana admitting-privileges law that tracked nearly word-
`for-word the Texas law struck down in Whole Woman’s Health.
`A plurality of four Justices examined the detailed evidence
`and findings on the likely burdens and benefits of the Louisi-
`ana admitting privileges law, and, following the reasoning
`and holding of Whole Woman’s Health, the plurality voted to
`strike down the new law. 140 S. Ct. at 2122–32 (plurality opin-
`ion of Breyer, J.). Four Justices dissented in four opinions.
`
`

`

`No. 17-2428
`
`5
`
`Chief Justice Roberts also voted to strike down the Louisi-
`ana law, concurring in the judgment in a separate opinion that
`is the focus here on remand. He had dissented in Whole
`Woman’s Health. He wrote that he still disagreed with that de-
`cision, but he explained that principles of stare decisis called
`for the Court to adhere to that earlier result on the essentially
`identical facts. 140 S. Ct. at 2134, 2139 (Roberts, C.J., concur-
`ring in judgment). He then explained that he believed Whole
`Woman’s Health had erred by balancing the challenged law’s
`benefits against its burdens in evaluating its constitutionality.
`Id. at 2135–36. Both the plurality and the Chief Justice agreed,
`however, that enforcement of the Louisiana law was properly
`enjoined before it took effect.
`Shortly after issuing June Medical, the Court issued its or-
`der in this case granting the State defendants’ petition for a
`writ of certiorari, vacating our decision, and remanding for
`further consideration in light of June Medical. See Box v.
`Planned Parenthood of Indiana & Kentucky, Inc., 141 S. Ct. 187,
`188 (2020). Such a “GVR” order calls for further thought but
`does not necessarily imply that the lower court’s previous re-
`sult should be changed. Klikno v. United States, 928 F.3d 539,
`544 (7th Cir. 2019). Pursuant to Circuit Rule 54, the parties
`submitted their views on the remand.1
`
`
`1 The State defendants at the same time petitioned for immediate en
`banc consideration of this case. No member of this court has requested an
`answer to or a vote on that petition. This decision on remand is being is-
`sued by the panel that heard this appeal originally. The pending petition
`is denied.
`
`

`

`6
`
`No. 17-2428
`
`II. Marks v. United States and Narrow Opinions
`A. Marks and its Variations
`The remand poses questions about how to interpret and
`apply decisions by the Supreme Court issued without major-
`ity opinions. The Supreme Court’s leading guidance on the
`question is one sentence in Marks: “When a fragmented court
`decides a case and no single rationale explaining the result
`enjoys the assent of five Justices, ‘the holding of the Court may
`be viewed as that position taken by those Members who con-
`curred in the judgments on the narrowest grounds.’” 430 U.S.
`at 193, quoting Gregg v. Georgia, 428 U.S. 153, 169 n.15 (1976)
`(plurality opinion of Stewart, Powell, and Stevens, JJ.). In re-
`cent decades, plurality decisions have become more frequent,
`especially on some of the most controversial issues the federal
`courts face. Lower courts have tried to follow the Marks in-
`struction in a variety of scenarios, and scholars and lower
`courts have identified several distinct models for applying
`Marks.
`A helpful guide comes from Professor Ryan Sullivan:
`The first of these approaches interprets Marks as
`limited to a narrow subset of plurality decisions
`reflecting a clearly discernible “implicit consen-
`sus” or “common denominator” among the Jus-
`tices. The second approach understands Marks
`as an instruction to lower courts to identify the
`opinion in a plurality decision that reflects the
`judgment-critical vote—typically the fifth con-
`curring vote—and treat that opinion as the
`Court’s holding. The third and final approach
`looks for points of majority consensus among
`
`

`

`No. 17-2428
`
`7
`
`different factions of concurring and dissenting
`Justices on distinct legal issues raised by the
`plurality decision.
`Ryan Williams, Questioning Marks: Plurality Decisions and Prec-
`edential Constraints, 69 Stan. L. Rev. 795, 806−07 (2017).
`The parties’ positions here identify different approaches
`and set the stage for our consideration. Relying on the first
`model, which is predominant in precedent, plaintiff Planned
`Parenthood contends that the June Medical plurality and con-
`currence share the narrow, common ground that Whole
`Woman’s Health has stare decisis effect on essentially identical
`facts. Because that is all that they share, that is the holding of
`June Medical, which thus did not produce a majority to over-
`rule Whole Woman’s Health. Not having been overruled, the
`standards and principles of Whole Woman’s Health still govern
`here.
`The State invokes both the second and third models for
`applying Marks. Using the second model, the State says the
`June Medical concurrence provided the swing vote and the
`narrowest ground for the judgment—stare decisis for Whole
`Woman’s Health on identical facts. That much is clear. The
`State goes further, however, in asserting in effect that every
`word of the concurrence must therefore be treated as the bind-
`ing, precedential holding of June Medical, whether those addi-
`tional portions support the judgment or not. Under that ap-
`proach, we would give the concurrence the effect of overrul-
`ing Whole Woman’s Health except as to virtually identical facts.
`Invoking the third model, using all opinions to predict votes
`in a future case, the State also argues that the June Medical con-
`currence and the dissents agreed on enough common ground
`to predict reliably that a majority of the Court would overrule
`
`

`

`8
`
`No. 17-2428
`
`Whole Woman’s Health and strike down the Indiana statute
`challenged here.
`We first identify questions in applying Marks and then ad-
`dress the variations argued by the parties, albeit in a different
`order. We close by addressing a couple of additional argu-
`ments raised in the briefs. The Supreme Court has observed
`that the Marks rule is “more easily stated than applied” and
`that it has “baffled and divided” lower courts. Grutter v. Bol-
`linger, 539 U.S. 306, 325 (2003), quoting Nichols v. United States,
`511 U.S. 738, 745−46 (1994). We hope here to avoid adding ev-
`idence to support the “baffled” observation.
`To identify a few of the problems baked into the Marks
`rule, how do we measure narrow v. broad? Does Marks re-
`quire common ground among opinions, and what if there is
`none? What counts as a common ground? Is it simply the ex-
`istence of a shared outcome or does it require a shared ap-
`proach to resolving a given legal question? Is everything in
`the narrowest opinion controlling, or just the portion support-
`ing the judgment? Can a “narrow” non-majority opinion
`overrule a previously controlling precedent? Do dissenting
`opinions count at all in measuring precedential effect?
`B. Dissenting Opinions and the Prediction Model of Precedent
`The last question, about dissenting opinions, is the easiest
`to answer, at least for a lower court like this one. The answer
`resolves the State’s reliance on the third model, counting
`votes among all opinions. Dissenting opinions do not count
`in the Marks assessment. Marks itself wrote in terms of “those
`Members who concurred in the judgments” 430 U.S. at 193,
`
`

`

`No. 17-2428
`
`9
`
`quoting Gregg, 428 U.S. at 169 n.15. The weight of circuit and-
`scholarly authority has taken the Court’s instruction at face
`value.
`We have rejected using dissents in Marks assessments:
`“under Marks, the positions of those Justices who dissented
`from the judgment are not counted in trying to discern a gov-
`erning holding from divided opinions.” Gibson v. American
`Cyanamid Co., 760 F.3d 600, 620 (7th Cir. 2014); accord, e.g.,
`United States v. Heron, 564 F.3d 879, 884 (7th Cir. 2009) (stating
`that Marks applies to opinions of those “Members who con-
`curred in the judgment[]” of the Court); Manning v. Caldwell
`for City of Roanoke, 930 F.3d 264, 280 n.13 (4th Cir. 2019) (en
`banc) (same); United States v. Carrizales-Toledo, 454 F.3d 1142,
`1151 (10th Cir. 2006) (same); United States v. Alcan Aluminum
`Corp., 315 F.3d 179, 189 (2d Cir. 2003) (same); Rappa v. New
`Castle County, 18 F.3d 1043, 1057 (3d Cir. 1994) (same); United
`States v. Hughes, 849 F.3d 1008, 1012 (11th Cir. 2017), rev’d on
`other grounds, 138 S. Ct. 1765 (2018) (”When determining
`which opinion controls, we do not ‘consider the positions of
`those who dissented.’”), quoting United States v. Robison, 505
`F.3d 1208, 1221 (11th Cir. 2007); United States v. Epps, 707 F.3d
`337, 348 (D.C. Cir. 2013) (“Stated differently, Marks applies
`when, for example, ‘the concurrence posits a narrow test to
`which the plurality must necessarily agree as a logical conse-
`quence of its own, broader position.’”) (emphasis added and
`removed), quoting King v. Palmer, 950 F.2d 771, 782 (D.C. Cir.
`1991) (en banc); King, 950 F.2d at 783 (“[W]e do not think we
`are free to combine a dissent with a concurrence to form a
`Marks majority.”); cf. United States v. Davis, 825 F.3d 1014, 1025
`(9th Cir. 2016) (en banc) (“[W]e assume but do not decide that
`
`

`

`10
`
`No. 17-2428
`
`dissenting opinions may be considered in a Marks analysis.”);2
`United States v. Johnson, 467 F.3d 56, 65 (1st Cir. 2006) (“[W]e
`do not share the reservations of the D.C. Circuit about com-
`bining a dissent with a concurrence to find the ground of de-
`cision embraced by a majority of the Justices.”).
`Scholars have generally agreed that dissenting opinions
`do not actually count, while noting that courts are not entirely
`consistent on this score. Michael L. Eber, When the Dissent Cre-
`ates the Law: Cross-Cutting Majorities and the Prediction Model of
`Precedent, 58 Emory L.J. 207, 218 (2008); Maxwell L. Stearns,
`The Case for Including Marks v. United States in the Canon of Con-
`stitutional Law, 17 Const. Comment. 321, 328 (2000); Nina
`Varsava, The Role of Dissents in the Formation of Precedent, 14
`Duke J. of Const. Law & Public Policy 285, 298−99 (2019); Jon-
`athan H. Adler, Once More, with Feeling: Reaffirming the Limits
`of Clean Water Act Jurisdiction, in The Supreme Court and the
`Clean Water Act: Five Essays 81, 93−94 (L. Kinvin Wroth ed.,
`Vt. Law Sch. 2007).3
`
`
`2 The Davis en banc majority did not decide this question, but concur-
`ring and dissenting opinions disagreed on it. See 825 F.3d at 1029 (Chris-
`ten, J., concurring) (Marks limits review to the opinions of Justices who
`concurred in judgment); id. at 1031 (Bea, J., dissenting) (Marks permits
`counting votes, including from dissenting Justices).
`3 To be sure, some of these scholars have urged a different approach,
`arguing that lower courts should use a prediction model, taking dissenting
`opinions into account to predict how the Supreme Court will decide the
`next case, but they agree that the prediction model is rarely used by courts
`and even more rarely acknowledged. See, e.g., Evan H. Caminker, Prece-
`dent and Prediction: The Forward Looking Aspects of Inferior Court Deci-
`sionmaking, 73 Tex. L. Rev. 1, 74 (1994) (arguing that “prediction has a
`proper, albeit circumscribed, role to play in inferior court decisionmak-
`ing,” but “conced[ing] that others will disagree with this conclusion”);
`
`

`

`No. 17-2428
`
`11
`
`This aversion to dissenting opinions in applying Marks is
`consistent with our more general approach to Supreme Court
`precedent. We simply do not survey non-majority opinions to
`count likely votes and boldly anticipate overruling of Su-
`preme Court precedents. That is not our job. As we are fre-
`quently reminded, only the Supreme Court itself can overrule
`its own decisions. Rodriguez de Quijas v. Shearson/American Ex-
`press, Inc., 490 U.S. 477, 484 (1989); accord, State Oil Co. v. Khan,
`522 U.S. 3, 20 (1997) (“it is this Court’s prerogative alone to
`overrule one of its precedents”); Agostini v. Felton, 521 U.S.
`203, 237 (1997) (instructing courts of appeals to leave to the
`Supreme Court “the prerogative of overruling its own deci-
`sions”), citing Rodriguez de Quijas, 490 U.S. at 484; Scheiber v.
`Dolby Laboratories, Inc., 293 F.3d 1014, 1019 (7th Cir. 2002)
`(highlighting that in State Oil v. Khan, the Supreme Court
`“pointedly noted” that the Seventh Circuit had been correct
`in refusing to declare defunct the Court’s directly controlling
`precedent). Accordingly, we decline the State’s invitation here
`to add together the Chief Justice’s concurrence and the dis-
`senting opinions and declare Whole Woman’s Health over-
`ruled.4
`
`
`Eber, When the Dissent Creates the Law, 58 Emory L. J. at 232 (acknowledg-
`ing that “most judges do not endorse the prediction model of precedent,
`at least openly”) (footnotes omitted); Varsava, The Role of Dissents, 14 Duke
`J. of Const. Law & Public Policy at 321–22 (“Advocates of the predictive
`approach generally exclude dissenting opinions from the process, but the
`inclusion of dissents is a theoretical possibility.”) (footnotes omitted).
`4 We recognize that parties may decide to adopt the prediction model
`in making decisions about their conduct or in deciding how to litigate dis-
`putes. The prediction model has a distinguished pedigree: “The prophe-
`cies of what the courts will do in fact, and nothing more pretentious, are
`what I mean by the law.” Oliver Wendell Holmes, The Path of the Law, 10
`
`

`

`12
`
`No. 17-2428
`
`C. Logical Subsets and Nesting Dolls
`We turn now to the first model of the Marks rule, argued
`by plaintiff and consistent with the substantial weight of au-
`thority: look for a “narrowest ground” that is a logical subset
`of the reasoning in other opinions concurring in the judgment.
`The Marks rule is easiest to apply when the fifth vote comes in
`a concurrence that agrees with part of the plurality’s reason-
`ing, so that the narrower opinion may be described as adopt-
`ing a logical subset of a broader opinion’s reasoning. The of-
`ten-cited metaphor is Russian nesting dolls. We and other
`courts have often said that for the Marks rule to apply, there
`must be a genuine common denominator underlying the rea-
`soning of a majority of justices. E.g., Gibson, 760 F.3d at 619;
`Heron, 564 F.3d at 884; Rappa, 18 F.3d at 1058; King, 950 F.2d at
`781. That opinion—the narrowest one—“must represent a
`common denominator of the Court’s reasoning; it must em-
`body a position implicitly approved by at least five Justices
`who support the judgment.” King, 950 F.2d at 781 (emphasis
`added).
`Under this approach, when the reasoning underlying the
`decisive concurring opinion fails to fit within a broader logical
`circle drawn by the other opinions, Marks simply does not ap-
`ply. King, 950 F.2d at 782; accord, Alcan Aluminum Corp., 315
`F.3d at 189 (explaining that where no single standard “consti-
`tutes the narrowest ground for a decision on that issue, there
`is then no law of the land”).
`
`
`Harv. L. Rev. 457, 461 (1897). But in a hierarchical court system, lower
`courts do not arrogate to themselves the task of overruling precedents of
`higher courts.
`
`

`

`No. 17-2428
`
`13
`
`In the simplest scenario, four Justices agree on two
`grounds for the judgment, and the decisive vote is cast by a
`concurring Justice who agrees with only one of those. In such
`a case, the concurring opinion’s rationale provides the nar-
`rowest ground and is deemed controlling.
`If we were to depart from this predominant understand-
`ing of Marks and applied it in the absence of a common de-
`nominator, then a single approach to a given legal question
`lacking majority support, perhaps lacking support from more
`than one Justice, would become national law. See King, 950
`F.2d at 782. This would be true even if that single approach
`produced the critical fifth vote supporting the judgment of the
`Court. Id.; see also Gaylor v. Mnuchin, 919 F.3d 420, 433 n.9 (7th
`Cir. 2019) (stating that where one Justice’s concurring opinion
`reached the same result as the plurality opinion, but did so
`under a different constitutional clause, that concurring opin-
`ion was not a “logical subset” of the plurality opinion), quot-
`ing Gibson, 760 F.3d at 619; Heron, 564 F.3d at 884 (“When,
`however, a concurrence that provides the fifth vote necessary
`to reach a majority does not provide a ‘common denominator’
`for the judgment, the Marks rule does not help to resolve the
`ultimate question.”). If there is no common denominator, then
`there is no binding reasoning, just facts and a result.
`The Supreme Court itself appears to follow this approach.
`In King, the District of Columbia Circuit illustrated this point
`with Coolidge v. New Hampshire, 403 U.S. 443 (1971), where a
`plurality of four Justices wrote that evidence could be seized
`pursuant to the plain-view exception to the Fourth Amend-
`ment’s warrant requirement only when the evidence was dis-
`covered inadvertently. See King, 950 F.2d at 782. Four other
`Justices wrote that inadvertence was not necessary for a valid
`
`

`

`14
`
`No. 17-2428
`
`seizure of evidence in plain view. Coolidge, 403 U.S. at 492, 506,
`510, 516 (four opinions, each concurring in part and dissent-
`ing in part). Justice Harlan concurred in the judgment that the
`challenged search was unconstitutional, but he offered no ra-
`tionale for evaluating the inadvertence requirement laid out
`by the plurality. Id. at 490 (Harlan, J., concurring in judgment).
`In a later decision, Texas v. Brown, 460 U.S. 730 (1983), the
`Supreme Court said that the Coolidge plurality’s inadvertence
`requirement did not constitute binding precedent and should
`be understood only as “the considered opinion of four Mem-
`bers of this Court.” Brown, 460 U.S. at 737 (plurality opinion).
`Eventually, the Supreme Court rejected the inadvertence re-
`quirement altogether. King, 952 F.2d at 782, citing Horton v.
`California, 496 U.S. 128 (1990).
`In similar circumstances—where no opinion adopting a
`narrowest common denominator of the Court’s reasoning can
`be identified—this court and other circuits have explicitly de-
`clined to apply Marks. See, e.g., Gibson, 760 F.3d at 619–20 (de-
`clining to apply Marks to Eastern Enterprises v. Apfel, 524 U.S.
`498 (1998)); Heron, 564 F.3d at 884 (declining to apply Marks
`to Missouri v. Seibert, 542 U.S. 600 (2004)); Schindler v. Clerk of
`Circuit Court, 715 F.2d 341, 345 (7th Cir. 1983) (declining to
`apply Marks to Baldasar v. Illinois, 446 U.S. 222 (1980)); Davis,
`825 F.3d at 1021–22 (declining to apply Marks to Freeman v.
`United States, 564 U.S. 522 (2011));5 Alcan Aluminum Corp., 315
`
`5 In Hughes v. United States, the Supreme Court explained that
`in Freeman, “[n]o single interpretation or rationale commanded a
`majority” of Justices. 138 S. Ct. 1765, 1768 (2018). The Court
`acknowledged that some courts of appeals, including the Seventh
`Circuit, in applying Marks had adopted the reasoning of Justice So-
`tomayor’s solo opinion concurring in the judgment. Id. Other
`
`

`

`No. 17-2428
`
`15
`
`F.3d at 189 (declining to apply Marks to Eastern Enterprises);
`A.T. Massey Coal Co. v. Massanari, 305 F.3d 226, 237 (4th Cir.
`2002) (also declining to apply Marks to Eastern Enterprises).
`In other words, Marks does not command lower courts to
`find a common denominator—to find an implicit consensus
`among divergent approaches—where there is actually none.
`Cf. Grutter, 539 U.S. at 325 (discussing division among federal
`courts of appeals in applying Marks to Regents of the University
`of California v. Bakke, 438 U.S. 265 (1978)); Nichols, 511 U.S. at
`745 (discussing division among state and federal courts in ap-
`plying Marks to Baldasar). It is not our duty or function to
`bring symmetry to any “doctrinal disarray” we might en-
`counter in our application of Supreme Court precedent.6
`
`
`courts, however, also applying Marks, adopted the plurality’s rea-
`soning. Id. Hughes resolved the sentencing issue in Freeman but ex-
`plicitly declined “to reach questions regarding the proper applica-
`tion of Marks.” Id.
`6 These limits of Marks are recognized in legal scholarship. See,
`e.g., Richard Re, Beyond the Marks Rule, 132 Harv. L. Rev. 1943, 1982
`(2019) (“[I]nstead of finding Marks holdings in all, or even most,
`fractured Supreme Court decisions, the logical subset approach as-
`pires to recognize Marks holdings only when one opinion is logi-
`cally and therefore inescapably ‘narrower’ than any other.”); Lewis
`A. Kornhauser & Lawrence G. Sager, The One and the Many: Adjudi-
`cation in Collegial Courts, 81 Cal. L. Rev. 1, 46–48 (1993) (Marks is
`available only “where the rationales for the majority outcome are
`nested, fitting within each other like Russian dolls”); Stearns, The
`Case for Including Marks, 17 Const. Comment. at 328 n.26 (explaining
`that Marks does not apply where “[t]he majority on the Court’s
`judgment [is] composed of two minority camps, each reaching op-
`posite resolutions of the two dispositive issues, but also reaching
`
`

`

`16
`
`No. 17-2428
`
`The logical subset approach to Marks applies here. In June
`Medical, there is one critical sliver of common ground between
`the plurality and the concurrence: Whole Woman’s Health was
`entitled to stare decisis effect on essentially identical facts. 140
`S. Ct. at 2120 (plurality); id. at 2139 (concurrence). The Marks
`rule therefore applies to that common ground, but it applies
`only to that common ground. That application offers no direct
`guidance for applying the undue burden standard more gen-
`erally, let alone to the quite different parental notice require-
`ment in this case. That absence of guidance answers our ques-
`tion: the Marks rule tells us that June Medical did not overrule
`Whole Woman’s Health. That means Whole Woman’s Health re-
`mains precedent binding on lower courts.
`D. The Swing-Vote Model
`To avoid this result, the State also invokes the second ap-
`proach to Marks and plurality opinions, in which lower courts
`try to identify the decisive fifth vote on the Supreme Court
`and treat that vote’s reasoning as controlling, even if it repre-
`sents the views of only one justice. Courts and scholars have
`called this the “swing-vote” approach. The State argues here
`that we should adopt the strongest, most controversial ver-
`sion of this swing-vote approach, which “treats as binding all
`aspects of the opinion reflecting the median Justice’s views,
`
`the same judgment”); Joseph S. Cacace, Note, Plurality Decisions in
`the Supreme Court of the United States: A Reexamination of the Marks
`Doctrine after Rapanos v. United States, 41 Suffolk U. L. Rev. 97, 113
`(2007) (emphasizing King’s “Russian dolls” approach to Marks);
`Linda Novak, Note, The Precedential Value of Supreme Court Plurality
`Decisions, 80 Colum. L. Rev. 756, 767 (1980) (“Many of the most
`troublesome plurality [and concurring] opinions” do not “stand in
`a ‘broader-narrower’ relation to each other.”).
`
`

`

`No. 17-2428
`
`17
`
`including propositions that no other participating Justice ex-
`plicitly or implicitly assented to.” Williams, Questioning
`Marks, 69 Stan. L. Rev. at 815; see also Re, Beyond the Marks
`Rule, 132 Harv. L. Rev. at 1979 (“In general, the median opin-
`ion would be outvoted whenever at least five Justices in non-
`median opinions would converge on the same outcome.”).
`The State here argues that we should treat as binding every-
`thing in the Chief Justice’s June Medical concurrence, including
`its continued disagreement with Whole Woman’s Health,
`whether that position was essential to the June Medical judg-
`ment or not, giving that non-majority opinion the power to
`overrule binding precedent established in a majority opinion.
`This swing-vote model is not consistent with Supreme
`Court precedent or our circuit precedent, nor is it the predom-
`inant model in courts around the country. For example, in
`United States v. Santos, 553 U.S. 507 (2008), the Supreme Court
`split four-one-four on the decisive issue. Justice Scalia wrote
`a plurality opinion for four justices to affirm; Justice Stevens
`wrote a separate, narrower opinion concurring in that judg-
`ment. But Justice Stevens’ concurring opinion expressed
`views on future cases not before the Court. The plurality ad-
`dressed how Marks should apply. Justice Stevens’ reasoning
`was the narrowest in support of the judgment, but the plural-
`ity flatly rejected the idea that everything in Justice Stevens’
`opinion was binding, in terms directly applicable here:
`“JUSTICE STEVENS’ speculations on that point address a
`case that is not before him, are the purest of dicta, and form
`no part of today’s holding.” Id. at 523 (plurality opinion of
`Scalia, J.).
`And whatever strengths the swing-vote model might have
`in other situations, it is not an appropriate application of
`
`

`

`18
`
`No. 17-2428
`
`Marks in these circumstances: First, the stated disagreement is
`not essential to the concurrence’s bottom-line vote to strike
`down the Louisiana law. The portions of the concurrence go-
`ing beyond stare decisis did not support the judgment and are
`obiter dicta. Second, the dispute between the plurality and
`concurrence in June Medical was not about a new legal issue
`but about the scope and validity of a Court precedent. Apply-
`ing the swing-vote test to treat everything in the concurrence
`as a binding holding would allow less than a majority to over-
`rule a Court precedent that had been established by majority
`vote.
`To frame the issue in simple, logical terms, the June Medical
`plurality adopted two propositions that we can label A and B.
`Proposition A was that Whole Woman’s Health previously
`struck down a nearly identical Texas law, so stare decisis re-
`quired striking down the new Louisiana law. Proposition B
`was that the majority opinion in Whole Woman’s Health cor-
`rectly stated and applied the undue burden test for abortion
`regulations. The Chief Justice’s concurrence adopted Proposi-
`tion A, applying stare decisis. It rejected Proposition B, adopt-
`ing instead Not-B: the majority opinion in Whole Woman’s
`Health misstated and misapplied the undue burden test.
`Applying Marks, the best way to understand the two opin-
`ions together is that the plurality’s adoption of Proposition B
`and the concurrence’s adoption of Proposition Not-B are both
`obiter dicta. They were not necessary to the actual judgment
`striking down the new Louisiana law on stare decisis
`grounds, Proposition A, for which there were five votes.
`There was no majority to overrule Whole Woman’s Health, so
`that precedent stands as binding on lower courts unless and
`until a Court majority overrules it.
`
`

`

`No. 17-2428
`
`19
`
`E. Other Arguments
`Additional arguments raised here do not fit as neatly into
`the three principal models for applying Marks. The State ar-
`