Bristol Reg'l Women's Center v. Herbert Slatery, III

RECOMMENDED FOR PUBLICATION Pursuant to Sixth Circuit I.O.P. 32.1(b) File Name: 21a0043p.06 UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT ┐ BRISTOL REGIONAL WOMEN’S CENTER, P.C.; MEMPHIS │ CENTER FOR REPRODUCTIVE HEALTH, on behalf of │ itself and its patients, KNOXVILLE CENTER FOR │ REPRODUCTIVE HEALTH; PLANNED PARENTHOOD OF │ TENNESSEE AND NORTH MISSISSIPPI, formerly known > No. 20-6267 as Planned Parenthood of Middle and East Tennessee, │ and DR. KIMBERLY LOONEY, │ Plaintiffs-Appellees, │ │ │ v. │ │ HERBERT H. SLATERY, III, Attorney General of │ Tennessee, GLENN R. FUNK, District Attorney General │ of Nashville, Tennessee, AMY P. WEIRICH, District │ Attorney General of Shelby County, Tennessee; │ BARRY P. STAUBUS, District Attorney General of │ Sullivan County, Tennessee, CHARME P. ALLEN, LISA │ PIERCEY, Commissioner of the Tennessee Department │ of Health, and W. REEVES JOHNSON, JR., M.D., │ President of the Tennessee Board of Medical │ Examiners, in their official capacities, │ Defendants-Appellants. │ ┘ On Motion for Stay Pending Appeal United States District Court for the Middle District of Tennessee at Nashville; No. 3:15-cv-00705—Bernard A. Friedman, District Judge. Decided and Filed: February 19, 2021 Before: MOORE, WHITE, and THAPAR, Circuit Judges. _________________ COUNSEL ON MOTION AND REPLY: Sarah K. Campbell, Mark Alexander Carver, OFFICE OF THE TENNESSEE ATTORNEY GENERAL, Nashville, Tennessee, for Appellants. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 2 ON RESPONSE: Autumn Katz, Michelle Moriarty, Rabia Muqaddam, CENTER FOR REPRODUCTIVE RIGHTS, New York, New York, Maithreyi Ratakonda, PLANNED PARENTHOOD FEDERATION OF AMERICA, New York, New York, Scott Tift, BARRETT JOHNSTON MARTIN & GARRISON, LLC, Nashville, Tennessee, Michael J. Dell, Jason M. Moff, KRAMER LEVIN NAFTALIS & FRANKEL LLP, New York, New York for Appellees. MOORE, J., delivered the order of the court in which WHITE, J., joined. THAPAR, J. (pp. 20–33), delivered a separate dissenting opinion. _________________ ORDER _________________ KAREN NELSON MOORE, Circuit Judge. “Unfortunately, the teachings of precedent are not always as clear as we might wish.” Wright v. Spaulding, 939 F.3d 695, 699 (6th Cir. 2019) (Thapar, J.). This bit of wisdom rings particularly true for the resolution of the motion to stay pending appeal before us today, which targets the district court’s judgment declaring unconstitutional and permanently enjoining a Tennessee statute that imposes a waiting period of 48 or 24 hours on women seeking an abortion in the state. Defendants1 argue that a stay is warranted because two precedents, EMW Women’s Surgical Center, P.S.C. v. Friedlander, 978 F.3d 418 (6th Cir. 2020), and Planned Parenthood of Southeastern Pennsylvania v. Casey, 505 U.S. 833 (1992), will compel us to vacate the district court’s judgment and permanent injunction on appeal. Our dissenting colleague—in his zeal to uphold what appears to be yet another unnecessary, unjustified, and unduly burdensome state law that stands between women and their right to an abortion—agrees. We think, however, that this case is not so simple. Because neither Casey nor EMW has foreclosed Plaintiffs’2 arguments, we must decline Defendants’ invitation to follow them blindly. After all, we are bound by “complementary duties: adhering to precedent when an issue has already been decided and considering an issue with an open mind when it has not.” Wright, 939 F.3d at 702. Assessing the parties’ preliminary 1 The defendants in this case are various Tennessee officials charged with the enforcement of the state’s waiting period law. 2 Plaintiffs are Tennessee abortion providers. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 3 arguments with the requisite clear eyes, we conclude that a stay is unwarranted. Accordingly, we DENY the motion for a stay pending appeal.3 I. BACKGROUND At issue in this case is the constitutionality of Tennessee Code Annotated § 39-15- 202(a)–(h), which imposes informational and temporal requirements on women seeking an abortion in Tennessee. These requirements are asserted to “ensure” that the woman’s “consent for an abortion is truly informed consent.” Tenn. Code Ann. § 39-15-202(b). 3 Our dissenting colleague calls for “immediate correction” of this order. Dissent at 20. Yet we fail to see how en banc review of this stay order is warranted, or even available. Indeed, despite our dissenting colleague’s impressive string citation, he fails to reference a single case where this court has assembled en banc to review a pre- merits stay order. Rather, each involves en banc review of a panel opinion resolving a direct appeal of a preliminary injunction—something that the court unquestionably has the authority to do, but something that is inapposite in the context of a stay order. That the dissent fails to reference a case involving en banc review of a stay order—despite the similarity between the standards for addressing a motion to stay and reviewing a preliminary injunction—is unsurprising. It is one thing for the en banc court to take on a matter once it has run its course with the merits panel; it is something else entirely to burden the court and the parties with en banc proceedings on a pre-merits stay order while the panel and the parties proceed on the merits of the appeal. Such a course of action would be exceedingly wasteful, especially when the en banc court is able to review the panel’s resolution of the appeal itself in due course. Here, merits briefing is already underway and is scheduled to conclude in short order. An opinion resolving the merits will follow thereafter. With the benefit of complete briefing, we may rethink our reasoning and conclusions. Once an opinion on the merits issues, if either party desires en banc review, they are free to seek it. Any other approach would be an unwarranted waste of the Parties’ and this court’s resources, and would raise questions as to why this court was going out of its way to wield en banc review so indiscriminately and unnecessarily. See generally Neal Devins and Allison Orr Larsen, Weaponizing En Banc (February 9, 2021) (unpublished manuscript) (on file with authors). Available at SSRN: https://ssrn.com/abstract=3782576. In any case, this court’s Internal Operating Procedures preclude such a wasteful result. They specify that only “[p]etitions seeking rehearing en banc from an order that disposes of the case on the merits or on jurisdictional grounds are circulated to the whole court” with limited exceptions that do not apply here. See 6 Cir. I.O.P. 35(g). Otherwise, “[p]etitions seeking rehearing en banc from other orders will be treated in the same manner as a petition for panel rehearing: They will be circulated only to the panel judges.” See 6 Cir. I.O.P. 35(h). This is the approach that we have previously followed. See Order, No. 12-4264, Serv. Emps. Int’l Union Local 1 v. Husted (6th Cir. Dec. 5, 2012) (treating a petition for rehearing en banc of a stay order as a petition for panel rehearing pursuant to 6th Cir. I.O.P. 35(g), (h)), ECF No. 48. We see no reason why it should be any different when a judge takes it upon himself to call for en banc rehearing sua sponte. As for our dissenting colleague’s call for Defendants to attempt an end run around this court’s ordinary procedures by seeking initial hearing en banc, we think that Defendants are quite capable of making their own strategic decisions without our dissenting colleague’s assistance. Suffice it to say, there are good reasons for leaving our ordinary procedures in place and allowing an appeal to run its course before calling upon the full court to resolve an issue. The parties soon will have fully briefed the merits, expanding upon their initial arguments here. With the benefit of more detailed briefing and (if required) oral argument, the resulting opinion(s) will be all the more informed. Yes, as things stand, we think Defendants are unlikely to succeed on appeal, but that result is not preordained and will depend on the Parties’ full arguments. The same is true—we hope—for our dissenting colleague’s opposite conclusion, and we trust that he will keep an open mind. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 4 The informational component of § 39-15-202(a)–(h) prohibits a physician from performing an abortion unless the woman has been informed, “orally and in person by the attending physician who is to perform the abortion, or by the referring physician”: (1) that she is pregnant; (2) of the probable gestational age of the fetus; (3) that the fetus may be viable if enough time has elapsed since the woman’s last menstrual cycle or the time of conception; (4) of the “numerous public and private agencies and services [that] are available to assist her during her pregnancy and after the birth of her child, if she chooses not to have the abortion”; and (5) of “[t]he normal and reasonably foreseeable medical benefits, risks, or both of undergoing an abortion or continuing the pregnancy to term.” Tenn. Code Ann. § 39-15-202(b). At the same time, the attending or referring physician must inform the woman of the “particular risks associated with her pregnancy and continuing the pregnancy to term, . . . as well as the risks of undergoing an abortion,” and provide a description of “the method of abortion to be used and the medical instructions to be followed subsequent to the abortion.” Tenn. Code Ann. § 39-15- 202(c). The temporal component of § 39-15-202(a)–(h) establishes a 48-hour waiting period that begins when the woman receives the mandated information described above. Tenn. Code Ann. § 39-15-202(d)(1). The statute provides that the waiting period will be reduced to 24 hours in the event of a court order “temporarily, preliminarily, or permanently” enjoining enforcement of the 48-hour waiting period or declaring it unconstitutional. Tenn. Code Ann. § 39-15-202(d)(2). After the waiting period has ended, but before having the procedure, the woman must sign a “consent form” acknowledging that she has received the requisite information. Tenn. Code Ann. § 39-15-202(d)(1). Because the woman must receive the statutorily mandated information in person, the effect of the waiting period is that a woman seeking an abortion in Tennessee must make at least two visits to the clinic where she will receive abortion care. The statute’s requirements apply to all abortions in Tennessee, except in cases of medical emergency that prevent compliance. See Tenn. Code Ann. § 39-15-202(d). Plaintiffs—suing on their own behalf and on behalf of their patients—brought suit to challenge, on two grounds, the constitutionality of § 39-15-202(a)–(h). First, Plaintiffs alleged that the statute imposes an undue burden on the accessibility of abortion in violation of the Due No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 5 Process Clause of the Fourteenth Amendment. Second, they alleged that the statute discriminates against women on the basis of sex and gender stereotypes in violation of the Equal Protection Clause of the Fourteenth Amendment. To remedy the alleged constitutional violations, Plaintiffs sought declaratory relief and a permanent injunction prohibiting the enforcement of § 39-15-202(a)–(h). On October 14, 2020, following five years of litigation and a four-day bench trial, the district court issued a comprehensive, 136-page opinion setting forth the findings of fact and conclusions of law that supported its ultimate conclusion that § 39-15-202(a)–(h)’s waiting period (whether 48 or 24 hours) violates the Due Process Clause of the Fourteenth Amendment. Adams & Boyle, P.C. v. Slatery, --- F. Supp. 3d ----, No. 3:15-cv-00705, 2020 WL 6063778 (M.D. Tenn. Oct. 14, 2020).4 Accordingly, the district court declared the statute’s waiting period unconstitutional and permanently enjoined its enforcement. Because its conclusion as to Plaintiffs’ due process claim was sufficient to enjoin the waiting period’s enforcement, the district court declined to rule on Plaintiffs’ equal protection claim. Defendants filed their notice of appeal on November 4, 2020. The same day, Defendants filed in the district court a motion to stay the judgment and the permanent injunction pending appeal. After the district court denied a stay, R. 287 (Op. & Order) (Page ID #6718–23), Defendants filed a motion for the same relief in this court. See Fed. R. App. P. 8(a). II. DISCUSSION The issuance of a stay pending appeal is a matter of our discretion, not a matter of right. Nken v. Holder, 556 U.S. 418, 433 (2009). To determine whether to exercise our discretion, we consider four familiar factors: (1) “whether the stay applicant has made a strong showing that he is likely to succeed on the merits”; (2) the likelihood that, absent a stay, “the applicant will be irreparably injured”; (3) “whether issuance of the stay will substantially injure the other parties interested in the proceeding”; and (4) the public interest. Id. at 434 (quoting Hilton v. Braunskill, 481 U.S. 770, 776 (1987). However, in constitutional cases like this one, the likelihood of 4 The district court upheld the statute’s informational components, a determination that Plaintiffs have not appealed. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 6 success on the merits is typically determinative. See Commonwealth v. Beshear, 981 F.3d 505, 508 (6th Cir. 2020) (order) (per curiam). The party seeking a stay bears the burden of showing that we should exercise our discretion in its favor. See Nken, 556 U.S. at 433–34. Defendants have not met that burden here. A. Most importantly, Defendants have not persuaded us that they are likely to succeed on the merits. Various standards of review will govern Defendants’ appeal of the district court’s grant of a permanent injunction: “we review factual findings for clear error, legal conclusions de novo, and the scope of injunctive relief for abuse of discretion.” Gibson Guitar Corp. v. Paul Reed Smith Guitars, LP, 423 F.3d 539, 546 (6th Cir. 2005). Yet at this stage, Defendants have not endeavored to demonstrate any clear error in the district court’s fact finding, and do not challenge the scope of the injunction except to the extent that they believe no injunction should have issued at all. Rather, Defendants attribute their likelihood of success entirely to legal error. Specifically, they contend that: (1) the district court applied the wrong standard for determining whether a state abortion restriction is invalid because it is an “undue burden” on the right under Casey, 505 U.S. at 877 (plurality opinion)5; (2) as a matter of law, the district court’s factual findings do not establish that the waiting period law unduly burdens the right to abortion; and (3) not enough women are unduly burdened by the waiting period for it to be facially invalid. None of these arguments are likely to succeed. 1. Defendants’ first argument—that they are likely to succeed on appeal because the district court applied the wrong undue burden standard—rests on an overly narrow view of our role as appellate jurists. Our dissenting colleague here stumbles into the same mistake as Defendants. To understand Defendants’ first argument—and our reasons for rejecting it—a truncated history is in order. All agree that the “undue burden” standard announced in 1992 by a plurality of the Supreme Court in Casey governs the constitutionality of state abortion restrictions under 5 All further references to Casey are to the plurality opinion. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 7 the Due Process Clause. As that plurality held, a state abortion restriction violates the Due Process Clause if it places an “undue burden” on a woman’s right to choose to terminate her pregnancy. 505 U.S. at 878. Under Casey, “[a]n undue burden exists, and therefore a provision of law is invalid, if its purpose or effect is to place a substantial obstacle in the path of a woman seeking an abortion before the fetus attains viability.” Id. This standard preserves the state’s “interest in potential life” and its interest in protecting the health of the mother, but only to the extent that the restrictions expressing those interests do not run afoul of the undue burden standard. Id. Fast forward some twenty years and disagreement had arisen as to the proper application of Casey’s undue burden standard (as the Casey plurality itself predicted that it would, see id.). In Whole Woman’s Health v. Hellerstedt, a majority of the Court endeavored to remedy that disagreement by clarifying that the undue burden standard “requires that courts consider the burdens a law imposes on abortion access together with the benefits those laws confer.” 136 S. Ct. 2292, 2309 (2016) (citing Casey, 505 U.S. at 887–901)). Applying this “balancing” test, id., to Texas laws requiring that an abortion provider have admitting privileges at a hospital within thirty miles of its facility and that its facility meet the minimum standards for surgical centers under state law, the Court held that the laws were invalid. Id. at 2300. As for the admitting- privileges law, the Court concluded that it unduly burdened abortion access because its burdens—the law would have closed about half of the state’s existing abortion facilities, increasing wait time, crowding, and the distance women needed to travel in order to receive an abortion—outweighed its benefits, which were essentially nonexistent given the evidence demonstrating that abortion complications requiring hospitalization are exceedingly rare. Id. at 2313–14. As for the surgical-center requirement, the Court concluded that it unduly burdened abortion access because it would further reduce the number of abortion providers in the state to the point where the remaining facilities would be unable to meet the existing demand for abortion services, and again offered no meaningful health benefits. Id. at 2314–18. Despite Whole Woman’s Health’s offer of clarity, the waters muddied again with the Court’s splintered decision in June Medical Services L.L.C. v. Russo, 140 S. Ct. 2103 (2020). Addressing a Louisiana admitting-privileges law that was nearly identical to the Texas law No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 8 invalidated in Whole Woman’s Health, a four-justice plurality balanced the burdens and benefits of the Louisiana law and concluded that it unduly burdened abortion access in violation of the Due Process Clause. Id. at 2112–13. The Chief Justice concurred, but on stare decisis grounds, noting the need to “treat like cases alike” and the similarity between the Louisiana law and its burdens and those at issue in Whole Woman’s Health. Id. at 2133–34. He went on, however, to criticize the balancing test employed by the plurality (and the majority in Whole Woman’s Health) as untethered from Casey. Id. at 2135–39. According to the Chief Justice, the benefits of a law are irrelevant to a determination of whether a state abortion restriction “has the purpose or effect of placing a substantial obstacle in the path of a woman seeking an abortion of a nonviable fetus” under Casey. Id. at 2135–36 (quoting Casey, 505 U.S. at 877). To the Chief Justice, “[l]aws that do not pose a substantial obstacle to abortion access are permissible, so long as they are ‘reasonably related’ to a legitimate state interest.” Id. at 2135 (quoting Casey, 505 U.S. at 878). Four dissenting justices would have held the Louisiana admitting-privileges law constitutional. A few months after the Supreme Court issued June Medical—and before this court had addressed that decision—the district court issued its findings of fact and conclusions of law, declaring Tennessee’s waiting period unconstitutional under Casey and permanently enjoining its enforcement. Adams & Boyle, 2020 WL 6063778, at *67. In doing so, the district court followed the approach taken by the plurality in June Medical and the majority in Whole Woman’s Health, balancing the burdens of the waiting period against its benefits. Id. at *63–64. The district court did not address the Chief Justice’s concurrence, apparently believing that it did not displace a balancing approach as the appropriate means for determining whether a state abortion restriction constitutes an undue burden under Casey. Finding that the waiting period risked women’s health and substantially limited access to abortion without any “countervailing benefit,” the district court concluded that the waiting period amounted to an undue burden under Casey and thus was invalid. Id. at *63–64. Two short days later, this court issued a decision in EMW, which involved a Kentucky law requiring that abortion providers obtain transfer and transport agreements with local hospitals and ambulance services. Unlike the district court here, a two-judge majority conducted No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 9 a lengthy analysis to determine which of the June Medical opinions was controlling given that none had attracted the support of a majority of the Court. 978 F.3d at 430–31. Bemoaning the “‘vexing task’ of deciding which opinion controls” when the Supreme Court issues a decision without a majority opinion, the EMW majority selected the Chief Justice’s concurrence (and its rejection of a balancing approach to the undue burden standard), reasoning that it was the narrower opinion and thus controlling under the doctrine set forth in Marks v. United States, 430 U.S. 188 (1977). EMW, 978 F.3d at 431, 433. That discussion may have been much ado about nothing, however, when the majority concluded that the district court clearly erred in finding that the transfer and transport laws would cause the closure of both of Kentucky’s two abortion providers, which had been the plaintiffs’ only asserted “basis for concluding that the challenged provisions impose a substantial obstacle” and thus were an undue burden under Casey. Id. at 440, 446. As Judge Clay pointed out in dissent, the majority’s conclusion rendered “apparently unnecessary” its prolonged analysis of the competing opinions in June Medical because without a burden supported by the record, the Kentucky transfer and transport law would have been valid under either approach. Id. at 454 n.2. For his part, Judge Clay would have followed the Whole Woman’s Health majority, concluding that even if the Chief Justice’s concurrence in June Medical were the controlling opinion from that case—a proposition that Judge Clay disputed but deemed unnecessary to resolve, id. at 454 n.3—the Chief Justice’s critique of the balancing approach was dicta because his opinion was based on stare decisis. Id. at 455. Applying this understanding of Casey and concluding that the district court’s findings were free of clear error, Judge Clay would have held the Kentucky laws unconstitutional because their burdens outweighed their benefits. Id. at 470. Our discussion of EMW leads us back to Defendants’ argument that they are likely to succeed on the merits because the district court applied the wrong undue burden standard to hold the state’s waiting period unconstitutional. According to Defendants, EMW—including its endorsement of the Chief Justice’s full June Medical concurrence—is binding precedent. Because the district court balanced burdens and benefits when it should have considered only the burdens under EMW, Defendants argue that we are required (at the very least) to vacate the injunction and remand for reconsideration under the correct undue burden standard. Defendants No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 10 are wrong. Even assuming that the district court erred in balancing burdens and benefits to hold Tennessee’s waiting period unconstitutional, the error would not warrant reversal here. As a general matter, Defendants are correct that a lower court’s application of an incorrect legal standard to grant injunctive relief may be grounds to vacate the injunction and remand for reconsideration. See Hamad v. Woodcrest Condo. Ass’n, 328 F.3d 224, 234 (6th Cir. 2003). However, there is no hard and fast rule that appellate courts must remand a case in such circumstances. See, e.g., Kos Pharm., Inc. v. Andrx Corp., 369 F.3d 700, 712 (3d Cir. 2004). Of course, a remand may be the best approach in many cases, for example, where the district court’s application of an incorrect legal standard leads to a further failure to make pertinent findings of fact. But in constitutional cases where the issuance of injunctive relief turns largely on the merits and the district court has already made the relevant findings of fact, a remand may well be unnecessary. That is the case here. After all, “whether particular facts amount to an undue burden is a question of law we must review de novo.” Dissent at 29–30. That is true whichever articulation of the undue burden standard applies, and no one contends that the district court’s factual findings would (or could) change upon remand.6 In any event, Defendants may well have overstated the precedential value of EMW, drawing into question whether the district court committed any error at all in applying the undue burden standard as the Court explained it in Whole Woman’s Health. Where a panel of this court is faced with two different standards for addressing a particular issue, but choosing between them would not change the outcome of the case due to the nature of the underlying facts, the panel’s choice between the two standards is dicta because it is not “necessary to the determination of the issue on appeal.” United States v. Hardin, 539 F.3d 404, 410–11, 413 (6th Cir. 2008) (holding an earlier opinion to be dicta insofar as it chose between a probable cause 6 Indeed, a remand would be particularly inappropriate here, where the district court has already stated that it would have resolved the case in the very same way had it applied the undue burden standard as Defendants understand it. When Defendants sought a stay of the district court’s declaratory judgment and permanent injunction, they raised essentially the same argument that they do here with respect to the district court’s purportedly incorrect legal standard. See R. 281 (Mem. at 3–4) (Page ID #6661–62). But the district court rejected that argument in denying a stay, reasoning that “with or without . . . a balancing test, the Tennessee [waiting period] statute constitutes a ‘substantial obstacle to a woman seeking an abortion’ and, thus, an undue burden.” R. 287 (Op. & Order at 4) (Page ID #6721). Even if we assume that the district court applied the wrong standard when it first ruled Tennessee’s waiting period unconstitutional, it would make little sense and waste considerable resources to remand for reconsideration when the district court has already said that it would not change anything. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 11 and a “lesser reasonable belief standard” for the quantum of proof needed for police officers to enter a residence and execute a search warrant where the facts of the earlier case were such that either standard would have been satisfied). That rule would apply here if Judge Clay were correct in his EMW dissent that it was unnecessary for the EMW majority to choose between the competing June Medical opinions and their distinct takes on Casey’s undue burden standard. See EMW, 978 F.3d at 454 n.2 (Clay, J., dissenting). At this stage, however, we need not resolve the issue of EMW’s precedential value, let alone the questions that would follow as to which understanding of Casey’s undue burden standard controls. As explained above, even if EMW were to compel this panel to apply the Chief Justice’s June Medical concurrence, it would not warrant reversal because the district court’s findings of fact enable us to conduct our review. As we will explain in the next section, we think it likely that, based on the district court’s unchallenged findings of fact, Tennessee’s waiting period amounts to an undue burden under Casey, and is thus invalid, with or without balancing. 2. Defendants’ next argument is that they are likely to succeed on the merits because the district court’s factual findings are insufficient to support its conclusion that Tennessee’s waiting period law unduly burdens access to abortion in the state. Defendants are wrong. As we have just suggested, we are likely to uphold the district court’s legal conclusion under either of the competing articulations of Casey’s undue burden standard. First, we set forth a summary of the district court’s findings of fact, which Defendants have not challenged in seeking a stay. The district court found “that the statutory waiting period burdens the majority of abortion patients with significant, and often insurmountable, logistical and financial hurdles.” Adams & Boyle, 2020 WL 6063778, at *62. In order to attend the two in-person visits required by the statute, “patients must take time off from work, arrange childcare, and find transportation on two different occasions.” Id. These costs are particularly significant in Tennessee because the state had just eight abortion providers in only four cities, forcing many women to travel significant distances to attend an appointment. See id. (noting that 63% of Tennessee women live in a county without an abortion provider). For low-income women especially, who make up the majority of women seeking an abortion in Tennessee, these No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 12 costs make obtaining an abortion almost prohibitively expensive, putting them to an impossible choice between the abortion and its attendant costs and providing for their own and their household’s “basic needs.” Id. at *62–63.7 Crediting the testimony of a plaintiffs’ witness, the district court found that under Tennessee’s abortion regime the significant and unexpected costs of an abortion put low-income women and their families at “grave risk” of having to sacrifice necessities like food, housing, and healthcare. Id. at *34, *62. These logistical and financial hurdles, the district court found, caused delays between appointments that were “significantly longer” than the 48 hours required by § 39-15-202(d), sometimes as much as four weeks, presenting further obstacles for women seeking an abortion in Tennessee. Adams & Boyle, 2020 WL 6063778, at *61. Medically, as gestational age increases, “an abortion becomes lengthier, more invasive, more painful, and riskier for the patient.” Id. at *62. The district court found that the delays caused by Tennessee’s waiting period “can and do cause” women to miss the deadline for an abortion entirely (19 weeks and 6 days from the first day of the woman’s last menstrual period), or at least the deadline for the strongly preferred, safer, and far less invasive medication abortion (10 weeks from the first day of the woman’s last menstrual period). Id. at *61. Indeed, the district court found the waiting period to be responsible for the increase in medically riskier second-trimester abortions in Tennessee after the statute’s passage. Id. at *62. Missing the deadline for a medication abortion further complicates patients’ logistical challenges because only five of Tennessee’s eight abortion providers offer surgical abortions. Id. at *61–62. Financially, abortion becomes more expensive the later into a pregnancy it takes place, and the district court found that the cost of an abortion in Tennessee has increased significantly since 2013. Id. at *18 & n.20, *62.8 This increased cost compounds the 7 An example puts the financial difficulty in perspective. The district court found that the “large majority of plaintiffs’ patients live in poverty” and that “the overwhelming majority of women seeking an abortion [in Tennessee] . . . are already mothers and are either poor or near low-income.” Adams & Boyle, 2020 WL 6063778, at *62 (quoting the trial testimony of Sheila Katz, Ph.D.) (second alteration original). For a single mother of one, the poverty line sits at approximately $17,000 per year. See id. at *30 n.30. The cost of a surgical abortion is approximately $900, meaning that this single mother would have to spend almost her entire monthly income to receive the procedure, and that does not even account for the travel, childcare, and other expenses she would incur. See id. at *18 & n.20. The additional burdens and expenses caused by the statutory waiting period could well place the abortion process out of reach. 8 For example, the district court credited the testimony of Rebecca Terrell, executive director of Choices Memphis (a Tennessee abortion provider), who testified that due to the statute’s two-visit requirement, her clinic No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 13 already significant financial burdens on women who must forgo wages and pay for travel and childcare on multiple occasions in order to receive an abortion under Tennessee law. Id. at *61– 62. Ultimately, the district court considered these logistical, financial, and medical hurdles caused by Tennessee’s waiting period and concluded that the waiting period is a substantial obstacle to abortion access in Tennessee. Adams & Boyle, 2020 WL 6063778, at *64. Finding that these hurdles stemmed largely from the need to attend two in-person appointments rather than the waiting period’s express duration, the district court concluded that even a 24-hour waiting period would constitute a substantial obstacle. Id. at *34, *61, *67. We think that the facts found by the district court likely satisfy either of the available articulations of Casey’s undue burden standard. First, we consider the Chief Justice’s approach, which would have us inquire into whether the waiting period has the “purpose or effect of placing a substantial obstacle in the path of a woman seeking an abortion of a nonviable fetus” without balancing the law’s benefits against its burdens in order to determine whether it unduly burdens access to abortion. June Medical, 140 S. Ct. at 2135 (quoting Casey, 505 U.S. at 877).9 That standard is satisfied here. The burdens imposed by Tennessee’s waiting period mirror (if not surpass) those in Whole Woman’s Health, where an admitting-privileges law led to “fewer doctors, longer waiting times, and increased crowding,” while also significantly increasing the amount of travel time needed to get to an abortion facility. 136 S. Ct. at 2313. Whatever else it may have done, the Chief Justice’s concurrence in June Medical did not upset Whole Woman’s Health’s conclusion that those kinds of burdens, where substantial enough, can render a state abortion law unconstitutional. Indeed, the Chief Justice emphasized the very same sorts of burdens in ruling Louisiana’s admitting-privileges law unconstitutional in June Medical, at least under stare decisis principles. See 140 S. Ct. at 2140–41. increased the costs charged for a surgical abortion from between $425 and $525 in 2013 to between $800 and $1,000 for the same procedure in 2019. Adams & Boyle, 2020 WL 6063778, at *18 & n.20. 9 Because we conclude that the waiting period is likely unduly burdensome, it is of no consequence whether it is “‘reasonably related’ to a legitimate state interest.” June Medical, 140 S. Ct. at 2135 (Roberts, C.J., concurring) (quoting Casey, 505 U.S. at 878). No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 14 Turning to the balancing test that the district court actually employed, Defendants—who bear the burden of persuasion here—have not argued that the district court’s conclusion was unwarranted if that test in fact applies. Under Whole Woman’s Health, Casey “requires that courts consider the burdens a law imposes on abortion access together with the benefits those laws confer,” and balance those considerations to determine whether the law unduly burdens the right to an abortion. 136 S. Ct. at 2309; see June Medical, 140 S. Ct. at 2120 (plurality op.). Compounding the severe burdens imposed by the waiting period, the district court found that there was no evidence that the waiting period furthered the state’s interest in potential life or its interest in the health of the mother. Adams & Boyle, 2020 WL 6063778, at *58–61. As for the state’s interest in potential life, the district court found that “there is no evidence that patients who do not return to an abortion provider for the second appointment (i.e., for the procedure) fail to do so because the challenged statute causes them to change their minds about having an abortion.” Id. at *61. Likewise, “women’s mental and emotional health is not benefited because the mandatory waiting period does nothing to increase the decisional certainty among women contemplating having an abortion.” Id. To the contrary, the district court found that “at least 95% of women are certain of their decisions [to have an abortion when seeking care], post- abortion regret is uncommon, and abortion does not increase women’s risk of negative mental health outcomes.” Id. In short, whatever limited benefits Tennessee’s waiting period may provide, they are “vastly outweighed” by its burdens. See EMW Women’s Surgical Ctr., P.S.C. v. Friedlander, 960 F.3d 785, 808 (6th Cir.), reh’g en banc dismissed, 831 F. App’x 748, 753 (6th Cir. 2020), petition for cert. filed (Oct. 30, 2020). To summarize, we think it likely that the district court’s factual findings compel the conclusion that Tennessee’s waiting period unduly burdens women’s abortion rights, whether analyzed under the Chief Justice’s understanding of Casey, or the understanding articulated by the majority in Whole Woman’s Health and the plurality in June Medical. Defendants—who bear the burden of persuading us that a stay is warranted—have not so much as attempted to challenge those factual findings for clear error. Thus, we reject Defendants’ argument that they are likely to succeed on the merits because the waiting period is not an undue burden on abortion rights. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 15 Our dissenting colleague disagrees. Primarily, he points to Casey, and accuses us of turning a blind eye to Supreme Court precedent. Not so—if anything, Casey supports our conclusion. In Casey, one of the Pennsylvania abortion restrictions that the Supreme Court addressed was a 24-hour waiting period that operated similarly to Tennessee’s here. Acknowledging that such a waiting period “[i]n theory . . . does not amount to an undue burden,” the Court went on to examine whether the waiting period amounted to a substantial obstacle “in practice” based on the district court’s findings of fact. Casey, 505 U.S. at 885 (emphasis added). Calling it a “closer question,” the Court concluded that the delay and increased costs associated with the waiting period did not amount to a substantial obstacle, even for those—like poor women or those who lived furthest from clinics—for whom the burden was the most significant. Id. at 886. The Court was careful, however, to note that its conclusion was based on “the record before [it],” id. at 887, a record that this court has since characterized as “sparse,” Cincinnati Women’s Servs., Inc. v. Taft, 468 F.3d 361, 372 (6th Cir. 2006). As explained above, the record here is substantial—with specific and comprehensive findings as to the logistical, financial, and medical obstacles created by § 39-15-202(a)–(h) that substantially limit the accessibility of abortion in Tennessee—far more so than the record in Casey. Indeed, the district court specifically found that Tennessee’s waiting period “significantly delays this time-sensitive medical procedure, and also makes it so time-consuming, costly, and inconvenient to obtain that the predominantly low-income population seeking the service must struggle to access it, if they can access it at all.” Adams & Boyle, 2020 WL 6063778, at *64; cf. Casey, 505 U.S. at 886 (noting the district court’s failure to “conclude that the increased costs and potential delays amount to substantial obstacles”). Furthermore, as the district court here noted, Pennsylvania had 81 abortion providers at the time that Casey was decided, compared to just eight in comparably sized Tennessee, which illustrates the waiting period’s disproportionate impact on women seeking an abortion in Tennessee. Adams & Boyle, 2020 WL 6063778, at *24 n.24, *64.10 Thus, if anything, Casey supports our conclusion that the waiting period of § 39-15- 202(a)–(h)—whether 48 or 24 hours—is likely unconstitutional. 10 Our dissenting colleague challenges this finding regarding the comparative availability of abortion facilities. Yet Defendants have stopped short of challenging this finding for clear error. Instead, in a footnote, they mention that the availability of Pennsylvania abortion providers was not mentioned in Casey and assert that it was No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 16 Nor does Casey categorically preclude logistical and financial obstacles from establishing undue burden, as our dissenting colleague contends. To the contrary, the Court’s failure to disregard those sort of burdens out of hand and its repeated references to the district court’s failure to find that they amounted to a substantial obstacle suggest that it had no intention of making such a sweeping rule. See 505 U.S. at 885–87. Whole Woman’s Health, where the Court found such kinds of burdens sufficient in portions of its opinion unaffected by the Chief Justice’s June Medical concurrence, only underscores the point. See Whole Woman’s Health, 136 S. Ct. at 2313.11 In sum, our dissenting colleague dramatically overstates Casey in criticizing our purported failure to “treat like cases alike,” Dissent at 28 (quoting June Medical¸ 140 S. Ct. at 2133–34 (Roberts, C.J., concurring)). The fact is that these cases are not alike: although both involve superficially similar waiting period laws, the burdens imposed here are substantially more severe, even if generally the same in kind. In Casey, the Court took pains to avoid a categorical ruling on the constitutionality of waiting period laws—it is not for an intermediate court such as ours to write such a holding into that case over a quarter-century later. See Wright, improper for the district court to consider it. Mot. Stay at 16 n.4. We disagree. The district court relied on record evidence to determine the relative availability of abortion services in Pennsylvania when Casey was decided. See Adams & Boyle, 2020 WL 6063778, at *24 n.24. The district court acknowledged testimony that the Pennsylvania figure likely included types of facilities not counted in Tennessee, but nevertheless found that the evidence established that abortion services were far less available in Tennessee than in Pennsylvania when the Court decided Casey. See id. This evidence helps to explain why the burdens in this case are more severe than those established in Casey, and Defendants’ cursory footnote has not given us reason to think otherwise. 11 Cincinnati Women’s Services, which the dissent also relies on, does not compel a different result. In that case, a panel of this court interpreted Casey to require that courts “determine whether a large fraction of the women ‘for whom the law is a restriction’ will be ‘deterred from procuring an abortion as surely as if the [government] has outlawed abortion in all cases.’” Cincinnati Women’s Services, 468 F.3d at 370 (emphasis added) (quoting Casey, 505 U.S. at 894). Latching onto the emphasized portion of the quoted language, the dissent would uphold any state abortion restriction so long as it remains at all possible for women to receive abortion services, no matter the difficulty they would face in doing so. Dissent at 29. Yet, under Whole Woman’s Health and June Medical, that cannot be the test. In Whole Woman’s Health, the Court invalidated a Texas admitting-privileges law that—like the waiting period law at issue here—imposed severe financial and logistical burdens upon women seeking an abortion. 136 S. Ct. at 2310–14. These burdens were enough to render the law invalid, despite the fact that women could still (though with great difficulty) receive an abortion with the law in place. See id. June Medical only underscores the import of Whole Woman’s Health in this regard. Indeed, in June Medical, the Court specifically rejected a similar formulation of the undue burden test. See 140 S. Ct. at 2133 (“[T]he State argues that Act 620 would not make it ‘nearly impossible’ for a woman to obtain an abortion. But, again, the words ‘nearly impossible’ do not describe the legal standard that governs here.”) (plurality op.); id. at 2139 (Roberts, C.J., concurring). No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 17 939 F.3d at 702. If Casey stood for the proposition that waiting period laws are always constitutional, then it would have said so. 3. Defendants’ final argument is that they are likely to succeed on the merits because the district court misapplied the “large fraction” test that determines whether an abortion restriction is facially invalid. Under Casey, a state abortion restriction is facially invalid if it is an undue burden “in a large fraction of the cases in which [the restriction] is relevant.” 505 U.S. at 895. In Cincinnati Women’s Services, we explained that this test “is more conceptual than mathematical,” but requires that the restriction at issue impose a substantial obstacle for “something more” than 12.5% of women for whom it is relevant. Cincinnati Women’s Services, 468 F.3d at 374. We think that the district court is likely correct that the test is met here. The district court concluded that, at the very least, Tennessee’s waiting period imposes a substantial obstacle on the 60% to 80% of women seeking an abortion who qualify as “low- income,” and concluded that this satisfied the large fraction test. Adams & Boyle, 2020 WL 6063778, at *65. As explained above, we agree that Tennessee’s waiting period is an undue burden for at least those women who qualify as low-income and risk going without basic necessities in order to cover the increased costs that the district court attributed to the waiting period. Thus, even assuming that the denominator for the fraction would be “all women seeking an abortion,” the test would be well satisfied here. See Cincinnati Women’s Services, 468 F.3d at 374 (noting that the “challenged restriction need not operate as a de facto ban for all or even most” of the relevant population to be facially invalid). In any case, the denominator is smaller than that. Casey directs us to compare the number of women unduly burdened by the law at issue (the fraction’s numerator) against the number of women for whom the law “is an actual rather than an irrelevant restriction.” 505 U.S. at 895. This relevance requirement means that the denominator is “a class narrower than ‘all women,’ ‘pregnant women,’ or even ‘the class of women seeking abortions identified by the State.’” Whole Woman’s Health, 136 S. Ct. at 2320 (quoting Casey, 505 U.S. at 894–95). Here, we think that the class of women for whom the waiting period is relevant would be those women No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 18 who were certain that they wanted an abortion when they sought out the service—about 95% of all women seeking an abortion in Tennessee, according to the district court. Adams & Boyle, 2020 WL 6063778, at *65. For those women, the waiting period is relevant because it forces them to delay their abortion when they otherwise would not have; for women who were uncertain whether they wanted to have an abortion, it is possible that the waiting period is irrelevant because they may have chosen to take time to consider their decision before returning to have the procedure (or not). Thus, the pertinent “large fraction” is modestly larger than the 60% to 80% of women seeking an abortion in Tennessee who qualify as low income—either way the fraction is large enough.12 On this point, our dissenting colleague essentially reissues his opinion that Tennessee’s waiting period law does not amount to an undue burden for any woman and thus fails the large fraction test. We have already rejected that argument and will not repeat ourselves here. B. Turning to the remaining stay factors, we conclude that this is the ordinary constitutional case where likelihood of success on the merits is determinative on the issue of whether to grant a stay pending appeal. See Beshear, 981 F.3d at 508, 511. Insofar as § 39-15-202(a)–(h) is likely unconstitutional, neither the state nor anyone else will be harmed by the state’s inability to enforce it. See Planned Parenthood Ass’n of Cincinnati, Inc. v. City of Cincinnati, 822 F.2d 1390, 1400 (6th Cir. 1987). In contrast, staying the injunction and leaving a likely unconstitutional waiting period in place would harm women seeking an abortion while we consider the parties’ full arguments on appeal. As for the final factor—the public interest—it should go without saying that the public has no interest in the enforcement of an unconstitutional law. Deja Vu of Nashville, Inc. v. Metro. Gov’t of Nashville & Davidson Cnty., 274 F.3d 377, 400 (6th Cir. 2001) (“[I]t is always in the public interest to prevent violation of a party’s 12 The district court also concluded that the waiting period is an undue burden for all of the 95% of women seeking an abortion in Tennessee who are certain of their decision to have an abortion, satisfying the large fraction test. Adams & Boyle, 2020 WL 6063778, at *65. However, we need not address that conclusion here because even using the smaller of the two fractions identified by the district court—60% to 80%, which qualifies as a large fraction—Defendants are unlikely to succeed on the merits. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 19 constitutional rights.”) (quoting G & V Lounge, Inc. v. Mich. Liquor Control Comm’n, 23 F.3d 1071, 1079 (6th Cir. 1994)). III. CONCLUSION For the foregoing reasons, we DENY the motion for a stay pending appeal. The appeal will proceed on the previously determined briefing schedule. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 20 _________________ DISSENT _________________ THAPAR, Circuit Judge, dissenting. Since Casey, no federal appellate court has successfully struck down an abortion waiting period. Why? Because the Supreme Court says that waiting periods are constitutional. Planned Parenthood of Se. Pa. v. Casey, 505 U.S. 833, 881–87 (1992) (plurality opinion). Indeed, both the Supreme Court and our court have upheld such laws. But here the majority, like the district court before it, decides to chart its own course. In doing so, the majority ignores Supreme Court and Sixth Circuit precedent, as well as the correct legal standard. Given the weighty interests involved in this case, the majority’s failure to issue a stay merits immediate correction either by our court or a higher one. After all, “it is not our place to ignore precedent and push our own agenda.” United States v. Navarro, 986 F.3d 668, 675–76 (6th Cir. 2021) (Moore, J., dissenting).1 1 The majority insists that Tennessee cannot seek en banc review of today’s decision because it is preliminary. I disagree. See Fed. R. App. P. 35(b) (noting that any “party may petition for a hearing or rehearing en banc”). Our court has often granted petitions to review non-merits rulings. See, e.g., Preterm-Cleveland v. Acton, No. 18-3329 (6th Cir. argued March 11, 2020) (en banc); In re Ohio Execution Protocol, 860 F.3d 881 (6th Cir. 2017) (en banc); City of Pontiac Retired Emps. Ass’n v. Schimmel, 751 F.3d 427 (6th Cir. 2014) (en banc); Warshak v. United States, 532 F.3d 521 (6th Cir. 2008) (en banc); Deja Vu of Cincinnati, L.L.C. v. Union Twp. Bd. of Trs., 411 F.3d 777 (6th Cir. 2005) (en banc). Of course, these are preliminary-injunction cases. But there is no material difference between a preliminary-injunction case and a stay case: Courts apply the same test in both. At best, the rules are confusing on whether a party may seek en banc review of a stay order. What is not confusing, however, is that there are at least two other options in this setting: (1) any active judge may seek sua sponte en banc review, and (2) a party may seek initial hearing en banc on the merits. First, Rule 35 says a majority of a circuit’s judges “may order that an appeal or other proceeding be heard or reheard by the court of appeals en banc.” Fed. R. App. P. 35(a) (emphasis added). And the United States Code provides that a majority of active circuit judges may hear any “case or controversy” en banc. 28 U.S.C. § 46(c). Finally, our operating procedures say that any active judge may request a poll sua sponte. 6th Cir. I.O.P. 35(e). When that occurs, “the clerk will immediately circulate voting forms to the en banc court.” Id. Second, if Tennessee chose to file a petition for initial hearing en banc on the merits, our court could grant it. Both opinions herein comprehensively discuss the merits, and nothing more will be added by an additional panel decision. The majority has already concluded that the law appears to be “unnecessary, unjustified, and unduly burdensome.” Maj. Op. 2. And if the en banc court grants review, it would have jurisdiction over the entire case and could consider any motion for reconsideration Tennessee might file. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 21 I. Tennessee’s legislature enacted a waiting-period law in 2015. The law requires doctors to give women certain information at least 48 hours before performing an abortion (except in cases of medical emergency). Tenn. Code Ann. § 39-15-202(a)–(h). That information includes the age of the unborn child, the alternatives to abortion, and the medical risks and benefits of abortion and pregnancy. Id. The law also provides that a 24-hour waiting period will take effect if a court enjoins the 48-hour waiting period. Id. In June 2015, a group of abortion providers sued. They alleged that Tennessee’s waiting- period law burdens abortion and is facially unconstitutional. They did not seek a preliminary injunction, and the law went into effect in July 2015. For five years, the law remained in force. And for five years, women continued to obtain abortions in Tennessee: Abortion rates remained above 10,000 per year both before and after passage of the law. Four years later, the district court held a bench trial. But it did so using the wrong legal standard: balancing the law’s benefits against its alleged burdens. Adams & Boyle, P.C. v. Slatery, --- F. Supp. 3d ---, No. 3:15-CV-00705, 2020 WL 6063778, at *63 (M.D. Tenn. Oct. 14, 2020). This approach directly conflicts with our published precedent and the standard set forth in the Chief Justice’s controlling opinion in June Medical. See EMW Women’s Surgical Ctr., P.S.C. v. Friedlander, 978 F.3d 418 (6th Cir. 2020). Applying the wrong standard, the district court then said that the law “provides no appreciable benefit.” Adams & Boyle, 2020 WL 6063778, at *63. It also found that the law “causes increased wait times, imposes logistical and financial burdens, subjects patients to increased medical risks, and stigmatizes and demeans women.” Id. at *61. Thus, the district court held that the law unduly burdens abortion. Id. at *63–64. Although the Casey Court held that nearly identical burdens did not render Pennsylvania’s waiting period unconstitutional, the district court distinguished the case in just two paragraphs. It claimed that the “fully developed record” here enabled it to reach a conclusion contrary to Casey: that Tennessee’s waiting period “severely burdens the majority of No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 22 women seeking an abortion.” Id. at *64. The district court permanently enjoined enforcement of the law. Tennessee then sought a stay pending appeal from the district court. It pointed out that the district court’s decision conflicted with Supreme Court and Sixth Circuit precedent governing waiting periods. See Casey, 505 U.S. at 871–76; Cincinnati Women’s Servs., Inc. v. Taft, 468 F.3d 361, 366, 372–74 (6th Cir. 2006). It also flagged that the Sixth Circuit had recently held that a benefits-burden test does not apply in the abortion context. EMW, 978 F.3d 418. Even so, the district court refused to issue a stay pending appeal. Tennessee now seeks a stay in this court. II. Courts consider four factors when determining whether to grant a stay pending appeal: “(1) whether the stay applicant has made a strong showing that he is likely to succeed on the merits; (2) whether the applicant will be irreparably injured absent a stay; (3) whether issuance of the stay will substantially injure the other parties interested in the proceeding; and (4) where the public interest lies.” Nken v. Holder, 556 U.S. 418, 434 (2009). In constitutional cases, the propriety of a stay largely hinges on the likelihood of success on the merits. Coal. to Defend Affirmative Action v. Granholm, 473 F.3d 237, 252 (6th Cir. 2006). Tennessee is likely to prevail on the merits for three reasons. First, the district court impermissibly balanced the benefits and burdens of the law in violation of controlling precedent. Second, the district court failed to apply rational basis review to the State’s basis for enacting the law. And third, the district court improperly held that the law substantially burdens abortion in Tennessee. Each of these errors will require us to reverse the district court’s judgment and injunction, so Tennessee is entitled to a stay pending appeal. A. Start with a foundational problem: The district court applied the wrong legal standard. It weighed the benefits and burdens of the waiting-period law to find the law unconstitutional. Adams & Boyle, 2020 WL 6063778, at *58–64. But our circuit has rejected this approach. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 23 EMW, 978 F.3d at 430–34. So too has the Supreme Court. See June Medical Servs. L.L.C. v. Russo, 140 S. Ct. 2103, 2136–39 (2020) (Roberts, C.J., concurring in judgment) (controlling opinion). And “a district court by definition abuses its discretion when it makes an error of law.” Koon v. United States, 518 U.S. 81, 100 (1996). To be fair, in balancing the benefits and burdens, the district court relied on the four- justice plurality opinion in June Medical. But as our court recently held, the Chief Justice’s separate opinion in June Medical—not the plurality opinion—provides the controlling legal rule in the case. EMW, 978 F.3d at 433; accord Hopkins v. Jegley, 968 F.3d 912 (8th Cir. 2020). That is because when the Supreme Court decides a case and no rationale garners majority support, “the holding of the Court” is the position taken by the justice or justices “who concurred in the judgment[] on the narrowest grounds.” Marks v. United States, 430 U.S. 188, 193 (1977). In June Medical, that means the Chief Justice’s separate opinion controls. Under the Chief Justice’s controlling opinion, the undue burden test does not permit judges to balance the benefits and burdens of an abortion regulation. June Medical, 140 S. Ct. at 2136–39 (Roberts, C.J., concurring in judgment). Instead, an abortion regulation is valid as long as it: (1) is “reasonably related to a legitimate state interest,” and (2) doesn’t place “a substantial obstacle in the path of a woman seeking an abortion.” EMW, 978 F.3d at 433–34. “Because the controlling opinion in June Medical Services clarified that the undue burden standard is not a balancing test, the district court erred in attempting to weigh the benefits of [the law] against [its] burdens.” Id. at 437. What is the majority’s response? It says that our court’s holding in EMW might be “dicta” and that Tennessee “may well have overstated the precedential value of EMW.” Maj. Op. 9–11. It claims (incorrectly) that the panel may have had an alternative basis for its decision, so its holding about the correct standard of review is not binding. (Never mind that alternative, independent holdings are themselves binding. Wright v. Spaulding, 939 F.3d 695, 701 (6th Cir. 2019); Chambers v. United States, 763 F. App’x 514, 522 n.2 (6th Cir. 2019) (Moore, J., concurring) (“I am not prepared to question the practice of alternative holdings, and an alternative holding is still a holding.”).) The majority’s suggestion is premised on a fundamental misunderstanding of EMW. And its invitation to defy precedent runs counter to the settled rule No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 24 of this and every other circuit: that “the holding of a published panel opinion binds all later panels unless overruled or abrogated en banc or by the Supreme Court.” Wright, 939 F.3d at 700. A bit of background is necessary to understand the scope of EMW’s holdings. EMW involved a Kentucky law that requires abortion providers to obtain transfer-and-transport agreements with local hospitals. 978 F.3d at 438. The district court did not have the benefit of June Medical and thus applied the balancing test found in Whole Woman’s Health. Id. After a three-day bench trial, the district court first found that the transfer agreements did not confer “a substantial medical benefit” and so did not promote “the health of women seeking abortions in Kentucky.” Id. Second, the district court found that the law imposed a burden on women seeking abortions because it would result in the closure of clinics. Id. at 426. Balancing the benefits against the burdens, the district court held that the law was unconstitutional. Kentucky appealed. The EMW court began with a thorough and careful analysis of the relevant case law, and of June Medical in particular. That analysis was not easy. Indeed, it spans nearly ten pages in the Federal Reporter. Id. at 429–37. Over the course of this thoughtful analysis, the majority explained why Chief Justice Roberts’ separate opinion in June Medical supplied the relevant legal test. And once the majority determined the relevant test, it applied that test to Kentucky’s transfer-and-transport law. To understand why that determination mattered, one must look at the arguments plaintiffs made in EMW. The plaintiffs advanced two independent arguments. First, the plaintiffs argued that the law was unconstitutional because the district court found that the law offered no benefits—even if it did not pose a substantial obstacle to abortion. Id. at 438. Our court disagreed. Id. Importantly for present purposes, we rejected the notion that a district court gets to independently determine whether a law has benefits. Id. Why? Because under June Medical it is not “the role of courts to objectively assign weight to the State’s interests in passing regulations on abortion.” Id. (cleaned up). Instead, we “are only to ascertain” whether the legislature had a rational basis to act. Id. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 25 Second, the plaintiffs argued that Kentucky’s law posed a substantial obstacle to abortion and so was unconstitutional for that reason too. Id. at 440–41. The district court agreed and found that the law would cause all of Kentucky’s abortion clinics to close. This time, we reversed based on clear error. We concluded that the district court lacked a sound evidentiary basis for finding that clinics would be forced to close. Id. at 441–46. Thus, we held that the law did not pose a substantial obstacle to abortion. My colleagues in the majority suggest they may not be bound by EMW’s analysis of June Medical because it was not “necessary to the determination of the appeal.” Maj. Op. 9–10 (quoting United States v. Hardin, 539 F.3d 404, 410–11 (6th Cir. 2008)). Because the EMW court held that the district court clearly erred in its substantial-obstacle finding, the majority says that EMW’s benefits-burden holding was potentially “unnecessary” and “much ado about nothing.” Id. at 9–11. Not so. In EMW, the plaintiffs argued that—even in the absence of a substantial obstacle—an abortion regulation could be unconstitutional based on its lack of benefits. 978 F.3d at 438 (“The plaintiffs contend that, even after June Medical Services, the district court’s findings establish that the challenged provisions are ‘not reasonably related to a legitimate purpose’ and are therefore unconstitutional even if they do not impose a substantial obstacle.”). If this argument were correct, it would have been an independent basis for affirming. But the EMW court applied the Chief Justice’s opinion in June Medical to reject the argument. Id. at 438–40. Under the Chief Justice’s opinion, courts must defer to legislative judgments about the benefits conferred by a law regulating abortion (so long as the law survives rational basis review); district court findings to the contrary are irrelevant. As this discussion shows, EMW’s application of June Medical plainly “contribute[d] to the judgment.” Wright, 939 F.3d at 701. So it constitutes a binding holding of this court. And since the district court abused its discretion by applying the wrong legal standard, Tennessee is likely to succeed on the merits of its appeal.2 2 The majority notes that defendants “have not argued that the district court’s conclusion was unwarranted if the [Whole Woman’s Health] test in fact applies.” Maj. Op. 14. The reason no one presses this argument is because No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 26 B. But the district court’s errors didn’t stop there. It also concluded that Tennessee’s waiting-period law serves “no legitimate purpose.” Adams & Boyle, 2020 WL 6063778, at *58– 64. This conclusion is deeply flawed. In determining whether abortion regulations “reasonably relate[] to a legitimate state interest,” courts must apply rational basis review. EMW, 978 F.3d at 433, 438 (citation omitted); June Medical, 140 S. Ct. at 2135 (Roberts, C.J., concurring in judgment). A rational basis to regulate abortion exists when there “is a problem at hand for correction and it might be thought that the particular legislative measure was a rational way to correct it.” EMW, 978 F.3d at 438 (cleaned up). This standard is “highly deferential.” Id. Under it, “legislative choice is not subject to courtroom factfinding and may be based on rational speculation unsupported by evidence or empirical data.” FCC v. Beach Commc’ns, Inc., 508 U.S. 307, 315 (1993); see also Vance v. Bradley, 440 U.S. 93, 110–11 (1979) (holding the same). Thus, a law remains constitutional even if it seems to offer “little if any benefit.” EMW, 978 F.3d at 438 (citation omitted). All that matters is whether the legislature had a rational basis to act. Here, the Tennessee legislature plainly did. In Casey, the Supreme Court held that waiting-period laws are rationally related to two legitimate state interests. First, they are “a reasonable measure to implement the State’s interest in protecting the life of the unborn.” 505 U.S. at 885. And second, they are a reasonable means of ensuring that women’s consent is “informed and deliberate.” Id. Just as these rationales justified Pennsylvania’s 24-hour waiting period, so too do they justify Tennessee’s 48-hour waiting period. Thus, Tennessee had (as a matter of law) a rational basis for enacting its waiting-period law. But the district court apparently believed the reasoning of the Tennessee legislature (and the Supreme Court) was not good enough. So it decided to engage in freewheeling factfinding of the Whole Woman’s Health test does not apply after June Medical and EMW. But if that test did apply, the law would almost certainly remain constitutional. As the Supreme Court recognized in Casey, abortion waiting periods play an important role in furthering informed consent and protecting the lives of unborn children. 505 U.S. at 885– 86. And any resulting burden does not amount to a substantial obstacle to abortion for the reasons explained elsewhere in this dissent. Thus, Tennessee’s waiting period would almost certainly survive benefits-burden balancing under the (since overruled) Whole Woman’s Health standard. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 27 its own. And based on that factfinding, it held that Tennessee’s waiting period does not “actually further[]” a legitimate state interest. Adams & Boyle, 2020 WL 6063778, at *58, *63. Yet as noted above, a district court cannot supplant a legislature’s judgment with “courtroom factfinding.” Beach Commc’ns, 508 U.S. at 315; see also EMW, 978 F.3d at 438–39. Nor can it “resolve conflicts in the evidence against the legislature’s conclusion,” reject the legislature’s findings based on “statistics in the record,” or dismiss the legislature’s judgment because it is based on “speculation.” Vance, 440 U.S. at 111 (cleaned up). The district court’s opinion contravenes these key principles of rational basis review. This alone requires vacatur. But that’s not all. In its frolic and detour, the district court also ignored a key part of the Supreme Court’s holding in Casey: that abortion waiting periods are reasonably related to legitimate state interests as a matter of law. 505 U.S. at 885. Because the district court “failed to perform traditional rational-basis review,” and because it failed to apply controlling precedent, Tennessee is likely to succeed on the merits of its appeal. EMW, 978 F.3d at 439–40. C. All of this is reason enough to stay the district court’s decision. But the district court committed another fundamental error: It held that Tennessee’s waiting-period law poses a substantial obstacle to abortion for almost all women in Tennessee. Once again, the district court’s holding defies Supreme Court and Sixth Circuit precedent. It also badly misapplies the large-fraction test, which is used to determine whether abortion regulations are facially valid. 1. A law is unconstitutional if it places “a substantial obstacle in the path of a woman seeking an abortion of a nonviable fetus.” June Medical, 140 S. Ct. at 2138 (Roberts, C.J., concurring in judgment) (cleaned up). “Under the law of our circuit, a woman faces a substantial obstacle when she is ‘deterred from procuring an abortion as surely as if the [government] has outlawed abortion in all cases.’” EMW, 978 F.3d at 434 (quoting Cincinnati Women’s Servs., 468 F.3d at 370). Here, the plaintiffs failed to show that Tennessee’s law imposes such a burden. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 28 Begin with precedent. In Casey, the Supreme Court held that Pennsylvania’s 24-hour waiting period was not a substantial obstacle to abortion. It acknowledged that the law would often cause delays of “much more than a day” given the need to schedule two appointments. 505 U.S. at 886. It recognized that the effects of the law would be “particularly burdensome” for “those women who have the fewest financial resources, those who must travel long distances, and those who have difficulty explaining their whereabouts to husbands, employers, or others.” Id. It noted that the delays might result in women experiencing additional “harassment” or “hostility.” Id. And it “did not doubt” that the law had “the effect of increasing the cost and risk of delay of abortions.” Id. (cleaned up). Still, the Supreme Court held that these findings were not enough to “demonstrate that the waiting period constitutes an undue burden.” Id. This was true even for the women most affected by the law, such as those with low incomes. Id. at 886– 87. Likewise, in Cincinnati Women’s Services, our court held that Ohio’s 24-hour waiting period did not pose a substantial obstacle to abortion for most women it affected. 468 F.3d at 373–74. We recognized that the waiting period might have the “effect of delaying abortions up to two weeks.” Id. at 366. But even still, we held that the law was constitutional. So how did the district court get around this precedent? It claimed that a more “fully developed record” allowed it to part from those decisions and find that Tennessee’s waiting period poses a substantial obstacle to abortion. Adams & Boyle, 2020 WL 6063778, at *64. But the burdens the district court identified are generally the same as those rejected as insufficient in Casey: “logistical and financial hurdles” caused “by having to attend two in-person appointments.” Id. at *62; see also Karlin v. Foust, 188 F.3d 446, 486 (7th Cir. 1999). Courts have an obligation to “treat like cases alike.” See June Medical, 140 S. Ct. at 2133–34. And they cannot circumvent that duty by dressing up departures from precedent in the garb of judicial factfinding. Since this case is nearly identical to Casey, the law compels the same conclusion. Beyond these financial and logistical issues, the district court pointed to a few more potential burdens. But all fall well short of being a substantial obstacle to abortion. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 29 First, the district court argued that Tennessee’s law is both “demeaning to women who have decided to have an abortion” and “undermines patient autonomy and self-determination.” Adams & Boyle, 2020 WL 6063778, at *63. But there is no “constitutional right to abortion on demand.” Casey, 505 U.S. at 887. What’s more, these findings say nothing about whether women are deterred from obtaining an abortion “as surely as if the [government] has outlawed abortion in all cases.” EMW, 978 F.3d at 434 (citation omitted). That is the test. And the district court didn’t even pretend to satisfy it. Next, the district court claimed that “as gestational age increases, an abortion becomes . . . riskier for the patient.” Adams & Boyle, 2020 WL 6063778, at *62. For a health risk to potentially pose an undue burden, that risk must be at least “appreciable.” Casey, 505 U.S. at 885. Yet at trial, the plaintiffs’ own expert testified that “surgical abortion at all times remains very safe.” Adams & Boyle, 2020 WL 6063778, at *8; see also id. at *11 (finding this testimony “fully credible” and giving it “great weight”). To be sure, the district court cites testimony suggesting that medical risk increases slightly as gestational age increases. But minimal increases in risk cannot create appreciable medical risk when, as plaintiffs contend, the procedure isn’t risky to begin with. See Tucson Women’s Ctr. v. Ariz. Medical Bd., 666 F. Supp. 2d 1091, 1103 (D. Ariz. 2009) (“Plaintiffs cite a 38% increase in risk [from an abortion procedure], but 38% of a very small number is still a very small number.”). Thus, any slight increase in risk does not amount to a substantial obstacle to abortion. Finally, the district court found that Tennessee’s waiting-period law “places significant burdens on the clinics themselves.” Adams & Boyle, 2020 WL 6063778, at *63. But burdens on abortion providers are irrelevant unless they unduly burden women’s access to abortion. Tennessee’s law does not require clinics to close or doctors to stop performing abortions. Cf. June Medical, 140 S. Ct. at 2133–34 (Roberts, C.J., concurring in judgment); EMW, 978 F.3d at 446. And for the reasons discussed elsewhere in this opinion, the waiting period does not unduly burden women’s ability to obtain abortions. Thus, this rationale offers no basis for finding that the law poses a substantial obstacle to abortion. The majority sees things differently. It suggests that the district court’s factfinding insulates its undue-burden determination. But whether particular facts amount to an undue No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 30 burden is a question of law we must review de novo. Memphis Planned Parenthood, Inc. v. Sundquist, 175 F.3d 456, 460 n.1 (6th Cir. 1999); see also A Woman’s Choice–East Side Women’s Clinic v. Newman, 305 F.3d 684, 689 (7th Cir. 2002). Otherwise, there would be “arbitrary discretion in the courts.” June Medical, 140 S. Ct. at 2134 (Roberts, C.J., concurring in judgment) (quoting Federalist No. 78); see also Am. Jewish Cong. v. City of Chicago, 827 F.2d 120, 129 (7th Cir. 1987) (Easterbrook, J., dissenting) (“When everything matters, when nothing is dispositive, when we must juggle incommensurable factors, a judge can do little but announce his gestalt.”). And reviewed de novo, there is no question that the district court’s undue-burden determination veered badly off course. Why? Because Casey is dispositive: In the absence of a medical emergency, requiring a woman to attend two appointments before obtaining an abortion is not a substantial obstacle.3 2. An abortion regulation that is unconstitutional in some applications “remains facially valid so long as it does not impose an undue burden ‘in a large fraction of the cases in which [the regulation] is relevant.’” EMW, 978 F.3d at 434 (quoting Casey, 505 U.S. at 895). The district court held that Tennessee’s waiting-period law imposes an undue burden in a large fraction of cases. It took two shots at explaining why this was the case. Both miss the mark. 3 The district court and majority argue that because Pennsylvania had more abortion clinics when Casey was decided than Tennessee does today, Casey does not control. This argument has multiple problems. For starters, Casey did not rely on the number of clinics in Pennsylvania. In fact, no opinion even mentions this statistic. Courts cannot narrow Supreme Court holdings based on facts the Court did not consider. See Seila Law LLC v. CFPB, 140 S. Ct. 2183, 2200 n.4 (2020) (When interpreting precedent, “what matters is the [evidence] the Court considered as the basis for its decision, not any latent [evidence] not alluded to by the Court.”). And even on its own terms, this attempt to distinguish Casey fails. The population of Tennessee in 2020 is about half of Pennsylvania’s population in 1992. Tennessee has abortion clinics in Nashville, Memphis, Knoxville, Bristol, and Mount Juliet. Adams & Boyle, 2020 WL 6063778, at *62 n.52. And the vast majority of Tennesseans live in the metro areas, suburbs, and counties immediately surrounding these cities. Indeed, given the shape of the State and the distribution of clinics, almost every Tennessean lives within a two-hour drive of an abortion clinic. Finally, the district court admits that “the calculation of Pennsylvania providers also included smaller OB/GYNs or hospitals that performed abortions, which were not included in the calculation of Tennessee providers.” Id. at *23 n.24. Thus, it is far from clear that Tennessee has fewer abortion providers than Pennsylvania did when Casey was decided, especially on a per capita basis. The majority criticizes the dissent for discussing this point. It claims that Tennessee “stopped short of challenging this finding for clear error.” Maj. Op. 15–16 n.10. But Tennessee did challenge the district court’s flawed use of these figures on page 16 of its stay motion (in footnote 4). And even if it hadn’t, the scope of a Supreme Court decision is a question of law, not fact, so it is a proper inquiry for appellate judges. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 31 Numerators and Denominators. Large-fraction analysis relies on two key variables. The first is the numerator. This is the number of women unduly burdened by an abortion regulation. Casey, 505 U.S. at 894–95. The second variable is the denominator. This is the group of women for whom an abortion regulation is “an actual rather than an irrelevant restriction.” Id. The district court erred twice in identifying the relevant numerator. Women Who Are Certain. The district court first held that Tennessee’s waiting-period law “unduly burdens all women who are certain of their decisions” to obtain an abortion. Adams & Boyle, 2020 WL 6063778, at *65. Why? Because these women experience delays that carry “no benefit.” Id. The district court estimated that 95% of women are certain of their decision to obtain an abortion and held that Tennessee’s waiting period imposes an impermissible burden on this entire group. But the district court did not find that these women “would effectively be barred from exercising their constitutional right to obtain an abortion.” Cincinnati Women’s Servs., 468 F.3d at 373 (citing Women’s Medical Prof’l Corp. v. Voinovich, 130 F.3d 187, 201 (6th Cir. 1997)). And that is the controlling legal question. So, are 95% of women seeking an abortion “deterred from procuring an abortion as surely as if the [government] has outlawed [it] in all cases?” EMW, 978 F.3d at 434 (citation omitted). All evidence is to the contrary: Tennessee performed 12,373 abortions the year before the law went into effect, and 11,235 abortions the year after. Mot. for Stay, Exs. C, E. The district court thus erred by holding that 95% of women are unduly burdened by Tennessee’s waiting-period law (and using this figure as the numerator in its large- fraction analysis). Low-Income Women. In the alternative, the district court also concluded that “all low- income women who seek an abortion are unduly burdened by the mandatory waiting period because it requires them to make a second trip to a provider.” Adams & Boyle, 2020 WL 6063778, at *65. The district court estimated that this group includes 60% to 80% of all abortion seekers in the State. Id. How are these women burdened? By “the costs and inconvenience of the second trip,” which may prove “either insurmountable or surmounted with great difficulty.” Id. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 32 But “inconvenience, even severe inconvenience, is not an undue burden.” Karlin, 188 F.3d at 481. As the Supreme Court explained in Casey, the fact that a law might be “particularly burdensome” for women with the “fewest financial resources” does not mean the law poses a substantial obstacle to abortion. 505 U.S. at 886. And given the rate at which abortions continue to occur in Tennessee, it is implausible that 60% to 80% of women are encountering a substantial obstacle to abortion in Tennessee, such that they are functionally unable to obtain the procedure. For these reasons, the district court erred in using the 60% to 80% number as the relevant numerator in its large-fraction analysis. Again, this error merits a stay. D. The remaining factors also favor a stay pending appeal. For one, Tennessee faces irreparable injury. “Any time a State is enjoined by a court from effectuating statutes enacted by representatives of its people, it suffers a form of irreparable injury.” Maryland v. King, 567 U.S. 1301, 1303 (2012) (Roberts, C.J., in chambers) (cleaned up). Although the majority contends that Tennessee’s law is “unnecessary” and “unjustified,” that is a question for the people of Tennessee, not unelected appellate judges. Maj. Op. 2. By contrast, a stay will not substantially injure the plaintiffs. The law remained in force for more than five years while the parties litigated in the district court. During that time, plaintiffs continued to operate their clinics, and women in Tennessee continued to obtain abortions. Finally, the public interest supports granting the stay. For as we have explained, “giving effect to the will of the people by enforcing the laws they and their representatives enact serves the public interest.” Thompson v. DeWine, 959 F.3d 804, 812 (6th Cir. 2020). E. One final point. The effects of today’s ruling will extend far beyond this case. The majority functionally overrules Casey. In doing so, it calls into question waiting-period laws in fourteen states.4 It also suggests that district courts (and appellate panels) have free rein to 4 Ariz. Rev. Stat. Ann. § 36-2153(A); Ark. Code Ann. § 20-16-1703(b); Fla. Stat. Ann. § 390.0111(3)(a)(1); Ind. Code Ann. § 16-34-2-1.1(a); Iowa Code Ann. § 146A.1(1); Ky. Rev. Stat. Ann. § 311.725(1); La. Rev. Stat. No. 20-6267 Bristol Reg’l Women’s Center et al. v. Slatery et al. Page 33 disregard controlling precedent and to substitute their preferences for the judgment of the Supreme Court. * * * Abortion may be controversial. Following Supreme Court precedent shouldn’t be. I dissent. Ann. § 40:1061.16(B)–(C); Miss. Code Ann. § 41-41-33(1); Mo. Ann. Stat. § 188.027; Ohio Rev. Code Ann. § 2317.56(B)(1); S.D. Codified Laws § 34-23A-56; Tex. Health & Safety Code Ann. § 171.012(a)(4), (b); Utah Code Ann. § 76-7-305(2); Wis. Stat. Ann. § 253.10(3)(c)(1).