RECOMMENDED FOR PUBLICATION Pursuant to Sixth Circuit I.O.P. 32.1(b) File Name: 23a0088p.06
UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT
┐ THOMAS A. FOX, and all those similarly situated, │ Plaintiff-Appellee, │ > Nos. 22-1265/1272 │ v. │ │ SAGINAW COUNTY, MICHIGAN, et al. (22-1265); ALCONA │ COUNTY, MICHIGAN, et al. (22-1272), │ Defendants-Appellants. │ ┘
Appeal from the United States District Court for the Eastern District of Michigan at Bay City. No. 1:19-cv-11887—Thomas L. Ludington, District Judge.
Argued: January 11, 2023
Decided and Filed: April 28, 2023
Before: KETHLEDGE, READLER, and MURPHY, Circuit Judges. _________________
COUNSEL
ARGUED: Douglas J. Curlew, CUMMINGS, MCCLOREY, DAVIS & ACHO, P.L.C., Livonia, Michigan, for Appellants in 22-1265. Matthew T. Nelson, WARNER NORCROSS + JUDD LLP, Grand Rapids, Michigan, for Appellants in 22-1272. E. Powell Miller, THE MILLER LAW FIRM, P.C., Rochester, Michigan, for Appellee. ON BRIEF: Douglas J. Curlew, CUMMINGS, MCCLOREY, DAVIS & ACHO, P.L.C., Livonia, Michigan, for Appellants in 22-1265. Matthew T. Nelson, WARNER NORCROSS + JUDD LLP, Grand Rapids, Michigan, Theodore W. Seitz, Kyle M. Asher, DYKEMA GOSSETT PLLC, Lansing, Michigan, Gregory M. Meihn, Matthew Wise, GORDON REES SCULLY MANUKHANI, Bloomfield Hills, Michigan, Charles A. Lawler, CLARK HILL, Lansing, Michigan, Frank Krycia, MACOMB COUNTY CORPORATION COUNSEL, Mt. Clemens, Michigan, for Appellants in 22-1272. E. Powell Miller, Sharon S. Almonrode, Christopher D. Kaye, THE MILLER LAW FIRM, P.C., Rochester, Michigan, Philip L. Ellison, OUTSIDE LEGAL COUNSEL PLC, Hemlock, Michigan, Matthew E. Gronda, St. Charles, Michigan, for Appellee. Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 2
_________________
OPINION _________________
MURPHY, Circuit Judge. In this appeal, we must address how Article III’s “standing” requirements apply to class-action lawsuits. In individual litigation, a plaintiff lacks standing to sue a defendant if the plaintiff’s injuries are not “fairly traceable” to that defendant. DaimlerChrysler Corp. v. Cuno, 547 U.S. 332, 342 (2006) (citation omitted). Yet some courts have relied on the “juridical link” doctrine to jettison this standing element in the class setting. See, e.g., Payton v. County of Kane, 308 F.3d 673, 678–79 (7th Cir. 2002). According to these courts, that doctrine allows a named plaintiff in a putative class action to sue defendants who have not injured the plaintiff if these defendants have injured absent class members.
This case shows how the doctrine is supposed to work. When a Michigan county forecloses on a property because its owner has failed to pay property taxes, Michigan law permits the county to obtain ownership of the property outright—even if its value exceeds the taxes owed. That fate befell Thomas Fox. After he racked up some $3,000 in unpaid taxes, Gratiot County took his land. He valued the property at over $50,000, and the county treasurer sold it for over $25,000. But Fox did not see any of the surplus. We have held that similar conduct amounted to an unconstitutional “taking.” Hall v. Meisner, 51 F.4th 185, 187–88 (6th Cir. 2022). Fox thus filed a class action against Gratiot County on behalf of himself and similar landowners. But Fox did not stop there. He also sued 26 other counties that did not injure him, arguing that they engaged in the same conduct against other delinquent taxpayers. The district court certified a class, holding that Fox had standing to sue these other counties under the juridical link doctrine.
We cannot agree. This oddly named doctrine conflicts with the Supreme Court’s precedent holding that a class-action request “adds nothing to the question of standing[.]” Simon v. E. Ky. Welfare Rts. Org., 426 U.S. 26, 40 n.20 (1976). And Fox lacks standing to sue the 26 other counties under normal rules. Nor has he justified the doctrine with historical evidence. Although the English Court of Chancery permitted parties to file class-like “bills of peace,” Fox cites no case in which a plaintiff pursued litigation on behalf of absent parties against a defendant Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 3
who had not harmed the plaintiff. The juridical link doctrine instead originates from dicta in a 1973 case that expressed a desire for the “expeditious” resolution of disputes. La Mar v. H & B Novelty & Loan Co., 489 F.2d 461, 466 (9th Cir. 1973). But expediency concerns cannot supplant Article III’s separation-of-powers protections. Hollingsworth v. Perry, 570 U.S. 693, 704–05 (2013). We thus vacate the certified class and remand for proceedings consistent with this opinion.
I
A
Since 1893, the Michigan General Property Tax Act has set the process that state actors must follow to collect overdue property taxes. See generally Mich. Comp. Laws §§ 211.1–.155; Act of June 1, 1893, No. 206, 1893 Mich. Pub. Acts 354. Historically, the Act permitted government officials to place tax liens on the properties of delinquent taxpayers and to auction off these liens. See Rafaeli, LLC v. Oakland County, 952 N.W.2d 434, 443 n.10 (Mich. 2020). But this process took years and left the titles of affected properties in disarray. See id. In 1999, the Michigan legislature overhauled the Act to promote the speedy “return” of delinquent properties to “productive” uses. Mich. Comp. Laws § 211.78(1); see Act of July 22, 1999, No. 123, 1999 Mich. Pub. Acts 732. Today, the Act permits a “foreclosing governmental unit” to obtain ownership of these properties. Mich. Comp. Laws §§ 211.78(8)(a), 211.78k(6). Because the county treasurer is the “foreclosing governmental unit” in most counties, we will refer to this unit as the “county treasurer.” See Rafaeli, 952 N.W.2d at 443 n.11.
A county treasurer must treat property taxes as “delinquent” if a property owner has not paid them by the March after a tax year. Mich. Comp. Laws § 211.78a(2). The treasurer must send a series of notices to the taxpayer. See id. §§ 211.78b–211.78c, 211.78f. If, by the next March, the taxpayer still has not paid the taxes (along with any interest, penalties, and fees), the treasurer may treat the property as “forfeited.” Id. § 211.78g(1). Within three months of the forfeiture, the treasurer must petition a state court for a judgment that provides the treasurer with “absolute title” to the property. Id. § 211.78h(1); see Rafaeli, 952 N.W.2d at 444. The treasurer must give notice of this suit to all parties with a legal interest in the property. Mich. Comp. Laws Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 4
§ 211.78i(2), (5). The court then must hold a hearing and grant the treasurer “fee simple title” if the taxpayer has not justified the failure to pay the taxes. Id. §§ 211.78h(5), 211.78k(5)–(6).
After obtaining a property, a county treasurer must give the state and the relevant city or township an option to buy it. Id. § 211.78m(1). If these entities decline, the treasurer may sell the property at a foreclosure auction. Id. § 211.78m(1)–(2). The Act requires the treasurer to deposit the proceeds from all of a year’s sales into one account and to disburse these proceeds “in a specific order of priority.” Rafaeli, 952 N.W.2d at 445. The treasurer must first reimburse the relevant tax fund for the unpaid taxes. Mich. Comp. Laws § 211.78m(8)(a). Sometimes, though, a property sale will produce more money than the amount that a property owner owes. In that circumstance, the 1999 amendments to the Act did not originally “provide for any disbursement of the surplus proceeds to the former property owner[.]” Rafaeli, 952 N.W.2d at 446. Instead, the amendments directed the treasurer to use the proceeds to pay for the costs of all foreclosure suits and to subsidize the county’s general fund. Id.
This scheme—which allows a treasurer to seize more of a landowner’s property than the amount that the landowner owed—has generated substantial litigation in state and federal court. Some landowners sought relief in state court under the Michigan Constitution’s takings clause. See id. at 441–42. The Michigan Supreme Court held that a treasurer’s failure to give a landowner the surplus proceeds from a foreclosure sale qualified as a taking under the state constitution. Id. at 449–64. Yet the court added a caveat to this holding. Id. at 465–66. It rejected the notion that the state constitution entitled a landowner to the fair market value of the property (minus the amount owed) if this value exceeded the price at the foreclosure sale. Id. It instead stated that a landowner could seek “no more” and “no less” than the surplus from that sale. Id. at 466.
In December 2020, the Michigan legislature responded to Rafaeli. It amended the Act to create a process by which landowners may seek these surplus proceeds. See Mich. Comp. Laws § 211.78t. The amendment establishes different procedures for landowners to follow depending on whether a treasurer sold their properties before or after the Michigan Supreme Court decided Rafaeli. Compare id. § 211.78t(1)(a), (2)–(5), with id. § 211.78t(1)(b), (6). Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 5
Apart from state-court litigation, landowners have sued in federal court under the Fifth Amendment’s Takings Clause. Their efforts have taken a winding path. Before Knick v. Township of Scott, 139 S. Ct. 2162 (2019), the Supreme Court held that takings plaintiffs must exhaust their claims in state court. Id. at 2167. We dismissed some landowners on this ground. See Wayside Church v. Van Buren County, 847 F.3d 812, 817–22 (6th Cir. 2017). In Knick, however, the Court overruled this exhaustion mandate. 139 S. Ct. at 2167. Since then, we have held that neither exhaustion nor the Tax Injunction Act nor any other comity principle stands in the way of federal courts entertaining these claims. Freed v. Thomas, 976 F.3d 729, 734–41 (6th Cir. 2020).
We thus recently reached the merits in Hall. There, a county treasurer foreclosed on the properties of three landowners but did not auction them off as in Rafaeli. 51 F.4th at 189. Instead, the treasurer exchanged the properties to a city for amounts that the landowners owed. Id. The city then transferred the three properties to a private entity. Id. That entity kept one of the properties and sold the other two for prices well above the landowners’ liabilities. Id. We held that the treasurer violated the Takings Clause by confiscating properties whose fair market value exceeded the landowners’ tax obligations. Id. at 189–96. In the process, we disagreed with the “dictum” in Rafaeli that landowners had a property interest only in the surplus proceeds that a treasurer received at a foreclosure sale. Id. at 189. The sale in Hall had not generated any surplus because the treasurer had obtained from the city only the amounts that the landowners owed. Id. We found this fact irrelevant. Summarizing centuries-old principles, we concluded that landowners have a property interest in their “equitable title” to real estate and that the government may not transform private equity into public property merely by passing a law. Id. at 190–96.
Since Hall, the Supreme Court has decided to review a similar takings question involving Minnesota’s foreclosure process. See Tyler v. Hennepin County, 143 S. Ct. 644 (2023). But the Court’s holding on this takings question will not affect the class-action issues that we confront here. So we see no need to wait for its decision to resolve this appeal. Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 6
B
In June 2019, as this state and federal litigation progressed, Fox brought this putative class action. He failed to pay taxes on property that he owned in Gratiot County. By February 2017, the amount that Fox owed had reached $3,091.23. That month, the Gratiot County treasurer took Fox’s property using the Act’s processes. Months later, the treasurer sold it at an auction for $25,500. Yet Fox claims that his property had a fair market value of at least $50,400. He thus suggests that his equitable title in the property reached at least $47,308.77—the difference between his tax delinquency and the fair market value. The county treasurer did not pay him that surplus. Nor did the treasurer even pay him $22,408.77—the difference between his tax delinquency and the sale price at the auction.
Although Fox claimed that only Gratiot County and its treasurer had harmed him, he named 26 other counties and their treasurers as defendants. (We will generically refer to the defendants as the “Counties.”) Fox alleged that the Counties’ common practice of keeping all surplus equity in the properties of delinquent taxpayers amounted to an uncompensated taking in violation of the U.S. and Michigan Constitutions. He also alleged that this practice qualified as an excessive fine, that it violated procedural and substantive due process, and that the Counties had been unjustly enriched.
The district court stayed Fox’s case while we considered in Freed whether these types of claims could proceed in federal court. After the Michigan Supreme Court’s decision in Rafaeli, however, Fox moved to certify a class action under Federal Rule of Civil Procedure 23.
The district court granted Fox’s motion and certified a class. Fox v. County of Saginaw, 2020 WL 6118487, at *1 (E.D. Mich. Oct. 16, 2020). The court recognized that Fox’s efforts to certify a class faced standing and discovery obstacles separate from Rule 23’s normal requirements. To begin with, Fox sought to certify a class of landowners from dozens of Counties that had not injured him, so he lacked standing to sue them under traditional principles. See id. at *4–5. The court overcame this problem using the “juridical link doctrine.” See id. It reasoned that Fox could sue all 27 Counties because they had all kept surplus proceeds pursuant to the Act and because the class members had all suffered the same type of injury. Id. at *6. Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 7
In addition, Fox sought to certify a class at an early stage before the Counties had taken class-certification discovery. The court found no need for discovery. It held that the existing record was “adequate” to show that it should certify a class action under Rule 23. Id.
Turning to Rule 23’s standards, the court held that Fox established Rule 23(a)’s four requirements: numerosity, commonality, typicality, and adequacy. As for numerosity, the court cited Fox’s public-records evidence identifying potentially thousands of foreclosed properties with fair market values exceeding the taxes owed. Id. at *7. As for commonality and typicality, the court reasoned that all class members presented the same liability question: Did the Constitution permit the Counties to keep the surplus proceeds? Id. at *8–9. As for adequacy, the court saw no conflict among Fox, his counsel, and the absent class members. Id. at *10.
The court next held that Fox had proved the predominance and superiority requirements for certifying a damages class under Rule 23(b)(3). It reasoned that Fox could obtain a common answer to the liability question using “generalized proof” for all class members. Id. It added that class litigation was superior to individual litigation because the small value of some of the landowners’ claims might prevent them from seeking relief absent a class action. Id. at *10–11.
Ultimately, the court defined the class as follows: “All persons and entities that owned real property” in the 27 Counties “whose real property, during the relevant time period, was seized through a real property tax foreclosure, which was worth and/or which was sold at tax auction for more than the total tax delinquency and were not refunded the value of the property in excess of the delinquent taxes owed[.]” Id. at *4. It did not identify the “relevant time period.” But Fox had sought to certify a class between 2008 and the end of this suit. See Mot., R.93, PageID 1286.
The district court later granted in part and denied in part the Counties’ motion to dismiss Fox’s (and now the class’s) complaint. See Fox v. County of Saginaw, 2021 WL 120855, at *16 (E.D. Mich. Jan. 13, 2021). It rejected the Counties’ sovereign-immunity defense. Id. at *6–7. But it accepted the individual treasurers’ argument that qualified immunity protected them from liability, and it dismissed them from this suit. Id. at *7–8. On the merits, the court denied the Counties’ motion to dismiss the class’s federal and state takings claims, its procedural-due- Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 8
process claim, and its unjust-enrichment claim. Id. at *10–13, *14–16. The court dismissed the excessive-fine and substantive-due-process claims. Id. at *13–14, *15.
The Counties sought our review on two grounds. They appealed the order denying them sovereign immunity under the collateral-order doctrine. And they asked for permission to appeal the class-certification order under Federal Rule of Civil Procedure 23(f). We affirmed the denial of sovereign immunity but granted the Counties permission to appeal the class-certification order under Rule 23(f). See Fox v. Saginaw County, 2022 WL 523023, at *3–6 (6th Cir. Feb. 22, 2022); In re Alpena County, 2022 U.S. App. LEXIS 8459, at *8 (6th Cir. Mar. 30, 2022) (order).
The Counties now raise two arguments in this Rule 23(f) appeal. They assert that Fox lacks standing to sue most of them. And they argue that he did not satisfy Rule 23’s requirements.
II. Standing
The Counties first assert that the district court could not certify the class because Fox lacks standing to sue all Counties other than Gratiot County. We review this standing question de novo. See Kanuszewski v. Mich. Dep’t of Health & Hum. Servs., 927 F.3d 396, 405 (6th Cir. 2019).
At the outset, Fox claims that we lack appellate jurisdiction to consider this standing argument because Rule 23(f) permits us to review only the propriety of a class certification, not other legal issues. Both text and precedent prove him wrong. As a matter of text, the rule says that we “may permit an appeal from an order granting or denying class-action certification[.]” Fed. R. Civ. P. 23(f) (emphasis added). Rule 23(f) thus gives us jurisdiction to review the entire certification “order,” not just the parts analyzing Rule 23’s requirements. See BP P.L.C. v. Mayor & City Council of Baltimore, 141 S. Ct. 1532, 1537–38 (2021); Yamaha Motor Corp., U.S.A. v. Calhoun, 516 U.S. 199, 205 (1996). And here, the district court’s order certifying the class concluded that Fox had standing to sue all Counties. See Fox, 2020 WL 6118487, at *4–6. Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 9
As a matter of precedent, courts treat a standing question as “fair game” whether or not a class-certification order addresses it. In re E.I. DuPont de Nemours & Co. C-8 Pers. Inj. Litig., 2022 WL 4149090, at *4 (6th Cir. Sept. 9, 2022); 16 Charles A. Wright et al., Federal Practice and Procedure § 3931.1, at 551–52 & n.31 (3d ed. 2012) (citing cases). That is because standing goes to a court’s subject-matter jurisdiction. See Steel Co. v. Citizens for a Better Env’t, 523 U.S. 83, 102–04 (1998). And a court has no power to certify a class if it lacks jurisdiction. See Olden v. LaFarge Corp., 383 F.3d 495, 498 (6th Cir. 2004). Since a potential standing defect implicates the propriety of a class certification, we may—indeed must—consider that defect in any Rule 23(f) appeal. See Bertulli v. Indep. Ass’n of Cont’l Pilots, 242 F.3d 290, 294 (5th Cir. 2001).
Our appellate jurisdiction secure, we turn to Fox’s standing. The Constitution permits federal courts to hear only “Cases” or “Controversies.” U.S. Const. art. III, § 2, cl. 1. To bring a case or controversy within the meaning of Article III, a plaintiff must have “standing” to sue. See Ass’n of Am. Physicians & Surgeons v. U.S. Food & Drug Admin., 13 F.4th 531, 536–37 (6th Cir. 2021). The Supreme Court’s standing test has three familiar requirements. See Lujan v. Defs. of Wildlife, 504 U.S. 555, 560–61 (1992). Plaintiffs must have suffered an injury. They must trace this injury to the defendant. And they must show that a court can redress it. See id.
If Fox had brought an individual suit, he would lack standing to sue most of the Counties under this test. Undoubtedly, when Fox’s property was foreclosed on and his equitable title taken, he suffered “tangible harms” that satisfy Article III’s injury element. TransUnion LLC v. Ramirez, 141 S. Ct. 2190, 2204 (2021). But a plaintiff who meets the “actual-injury requirement” does not thereby obtain a license to sue anyone over anything. See Lewis v. Casey, 518 U.S. 343, 357 (1996). The plaintiff must show that the injury is “fairly traceable” to the defendant. Cuno, 547 U.S. at 342 (citation omitted). And Fox can trace his injury only to Gratiot County; he does not attempt to connect it to the other Counties. That fact poses a problem for him because standing is not dispensed “in gross.” Davis v. Colerain Township, 51 F.4th 164, 171 (6th Cir. 2022) (quoting Cuno, 547 U.S. at 353). Just as a plaintiff’s standing to seek damages does not give the plaintiff standing to seek an injunction, see id., so too a Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 10
plaintiff’s standing to sue one defendant does not give the plaintiff standing to sue every other defendant. The plaintiff must tie the injury to each defendant. See Universal Life Church Monastery Storehouse v. Nabors, 35 F.4th 1021, 1031 (6th Cir. 2022); Disability Rts. S.C. v. McMaster, 24 F.4th 893, 900 (4th Cir. 2022); Daves v. Dallas County, 22 F.4th 522, 542 (5th Cir. 2022) (en banc); Calzone v. Hawley, 866 F.3d 866, 869 (8th Cir. 2017); Mahon v. Ticor Title Ins. Co., 683 F.3d 59, 62–63, 65–66 (2d Cir. 2012). Under conventional principles, Fox has not done so for any County other than Gratiot County.
How does Fox seek to overcome this problem? He turns to his class-action allegations. Under the juridical link doctrine, Fox says, he has standing to sue the 26 other Counties as a representative of the class. If 26 landowners with property in each of the 26 other Counties had joined Fox’s suit, each landowner would likely have standing to sue his or her respective County in the same way that Fox has standing to sue Gratiot County. See Daves, 22 F.4th at 542. So whether these 27 plaintiffs could litigate as a group would raise a joinder question under Federal Rule of Civil Procedure 20(a), not a standing question under Article III. In the same way, Fox argues, whether he may sue the other Counties turns on Rule 23, not Article III.
A circuit split exists over this doctrine. The Seventh Circuit has held that named plaintiffs may bring a class action against some defendants who did not injure them if the class members would have standing and if the named plaintiff can meet Rule 23’s requirements. Payton, 308 F.3d at 678–82. The Second Circuit, by contrast, has held that the named plaintiffs must have standing to sue every defendant at the outset, even when they seek to certify a class. See Mahon, 683 F.3d at 63–66; see also Wong v. Wells Fargo Bank, 789 F.3d 889, 896 (8th Cir. 2015).
For our part, we have narrowed the juridical-link concept but left open the possibility that it might apply in Fox’s situation. See Thompson v. Bd. of Educ. of Romeo Cmty. Schs., 709 F.2d 1200, 1205 (6th Cir. 1983). Thompson reserved the question whether plaintiffs could sue many state defendants (including those who did not harm them) if the defendants acted within their local jurisdictions pursuant to “a state statute or uniform policy” that applied across the state. See id. The defendants in Thompson had not followed a uniform policy, so that case did not Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 11
implicate this question. See id.; see also Perry v. Allstate Indem. Co., 953 F.3d 417, 420 n.2 (6th Cir. 2020).
Fox’s suit does implicate it. The Counties allegedly took Fox’s (and the class members’) land pursuant to the same Act. We thus must wade into this circuit conflict. And we side with the Second Circuit in this debate. The juridical link doctrine does not comport with the Supreme Court’s standing cases. And Fox does not even attempt to offer any historical analogy for it.
1. Precedent
a. The juridical link doctrine conflicts with Supreme Court precedent in at least three ways. First, and most obviously, the Court has held that its usual three-part test for standing applies with full force in the class-action context. As the Court has said, a decision to assert class allegations in a complaint “adds nothing to the question of standing[.]” Simon, 426 U.S. at 40 n.20. Like all plaintiffs, class representatives must prove their own “case or controversy with the defendants” in order to seek relief for “any other member of the class.” O’Shea v. Littleton, 414 U.S. 488, 494 (1974); Frank v. Gaos, 139 S. Ct. 1041, 1046 (2019) (per curiam). The representatives thus must allege an individual injury; they cannot piggyback off the injuries “suffered by other, unidentified members of the class[.]” Warth v. Seldin, 422 U.S. 490, 502 (1975). And even when they have incurred an injury, they lack standing to seek class-wide relief that would go beyond remedying that injury. See Lewis, 518 U.S. at 357–58. That is, they may not seek relief against different conduct that has harmed other class members even when it is “similar” to the conduct that harmed the representatives. Blum v. Yaretsky, 457 U.S. 991, 999– 1001 & 1001 n.13 (1982).
The juridical link doctrine departs from these rules. Fox himself has not alleged “a case or controversy” with the 26 non-Gratiot Counties. O’Shea, 414 U.S. at 494. And even if these Counties committed actions “similar” to Gratiot County’s “injurious conduct” against Fox, he still lacks standing to challenge their conduct because it did not affect him. Blum, 457 U.S. at 999. Fox also cannot rely on the claim that the other Counties harmed the “members of the class” that he seeks to represent. Id. at 1001 n.13. He must show that they injured him Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 12
“personally” to have standing to sue them. Id. (quoting Warth, 422 U.S. at 502). He cannot do so.
Second, aside from Fox’s mistaken reliance on the class members’ injuries, his theory conflicts with Supreme Court precedent identifying the critical time for standing. Plaintiffs must have standing “at the outset of the litigation.” Friends of the Earth, Inc. v. Laidlaw Env’t Servs. (TOC), Inc., 528 U.S. 167, 180 (2000); see Davis v. FEC, 554 U.S. 724, 734–35 (2008). Why? Because a failure of the plaintiff’s standing is a failure of the court’s subject-matter jurisdiction. See Steel Co., 523 U.S. at 102–04. And a court’s jurisdiction has long turned on the facts as they are when a plaintiff sues, not on later-in-time facts. See Grupo Dataflux v. Atlas Glob. Grp., 541 U.S. 567, 570–71 (2004); Mollan v. Torrance, 22 U.S. 537, 539 (1824). So a plaintiff who lacks standing from the start cannot rely on factual changes during the suit to establish it. See Lujan, 504 U.S. at 569 n.4. Conversely, if a plaintiff possesses standing from the start, later factual changes cannot deprive the plaintiff of standing. See Laidlaw, 528 U.S. at 189–90. Those changes instead will create “mootness” issues and trigger that doctrine’s more forgiving rules. See id. at 190–92; see also County of Riverside v. McLaughlin, 500 U.S. 44, 51–52 (1991).
The juridical link doctrine departs from this approach because it changes the time when standing matters. Here, Fox undisputedly lacked standing as to the 26 non-Gratiot Counties when he sued. At that time, the court had yet to certify a class. The court thus could not rely on the class members’ injuries because they were not parties. See Smith v. Bayer Corp., 564 U.S. 299, 313 (2011). To decide whether a named plaintiff like Fox has standing under the juridical link doctrine, then, a court must look to the time of class certification. The doctrine allows a class certification decision to create standing against defendants that did not injure the named plaintiff. And under the doctrine, a decision that rejects class certification would foreclose that standing. Either way, the time of the certification decision (not the time of the suit) becomes the critical time to decide whether the plaintiff has standing—in conflict with the normal rule. See Laidlaw, 528 U.S. at 180.
Third, Fox’s theory conflicts with more general principles. The Court’s three-part standing test represents the “irreducible constitutional minimum” to assert a case or controversy in federal court. Lujan, 504 U.S. at 560. It is a key component of the Constitution’s separation Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 13
of powers designed to protect defendants from that portion of the federal government’s coercive power vested in the judiciary. See TransUnion, 141 S. Ct. at 2203; Clapper v. Amnesty Int’l USA, 568 U.S. 398, 408 (2013). Given the test’s constitutional roots, the Court has required “strict compliance” from plaintiffs and has rebuffed all efforts to create exceptions. Raines v. Byrd, 521 U.S. 811, 819–20 (1997). For example, the Court has refused to create a standing exception even if no party could satisfy its three-part test and the lack of a proper “plaintiff” might force the Court to leave an important legal issue unaddressed. Valley Forge Christian Coll. v. Ams. United for Separation of Church & State, 454 U.S. 464, 489–90 (1982). Similarly, it has refused to create an “exemption” from its standing test even when plaintiffs claim that they need one to prevent a “chilling effect” on important rights. Whole Woman’s Health v. Jackson, 142 S. Ct. 522, 538 (2021); United Presbyterian Church in the U.S.A. v. Reagan, 738 F.2d 1375, 1380 (D.C. Cir. 1984) (Scalia, J.). And the Court has refused to water down standing’s strict requirements with mootness’s more flexible exceptions. See Laidlaw, 528 U.S. at 190–91.
In conflict with these principles, the juridical link doctrine creates an “efficiency” exception to standing. The doctrine grew out of a stray sentence about judicial efficiency in a 1973 Ninth Circuit case—a time at which the Supreme Court had yet to clearly identify standing’s constitutional floor. See La Mar, 489 F.2d at 466; see also Ass’n of Am. Physicians, 13 F.4th at 538. In La Mar, the named plaintiff sued not just his own pawn broker but also all other pawn brokers in Oregon, alleging that they had committed the same statutory violation. 489 F.2d at 462. The court merely “assume[d] the presence of standing” because it went on to hold that Rule 23 prevented the plaintiff from litigating against the other pawn brokers. Id. at 464–66. Nevertheless, the court opined that a plaintiff might be able to pursue this sort of class action if the defendants who did not harm the plaintiff were “juridically related in a manner that suggests a single resolution of the dispute would be expeditious.” Id. at 466. The court thus justified the juridical link doctrine on the ground that it could streamline litigation. When later defending this doctrine, the Seventh Circuit similarly suggested that courts could make more “efficient uses of the class action device” if class representatives could sue defendants who did not harm them. Payton, 308 F.3d at 681. Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 14
But efficiency cannot trump the separation of powers. See INS v. Chadha, 462 U.S. 919, 959 (1983). Congress could certainly enact laws in a more “efficient” manner if it could disregard Article I’s requirement to, say, present bills to the President. Id. at 944. But that fact has never justified departures from Article I’s lawmaking process. See id. at 946–59; Clinton v. City of New York, 524 U.S. 417, 438–49 (1998). Likewise, even if courts could resolve issues with more “convenience and efficiency” by disregarding Article III’s case-or-controversy requirement, the Constitution bars them from doing so. Hollingsworth, 570 U.S. at 704–05 (quoting Raines, 521 U.S. at 820). By making each branch’s exercise of its powers more difficult, see Chadha, 462 U.S. at 959, these separation-of-powers rules promote individual and political liberty, see Bond v. United States, 564 U.S. 211, 221–24 (2011). Just as Article I’s requirements protect the general public from coercive legislation, Article III’s requirements protect private and public defendants from coercive judgments. See Clapper, 568 U.S. at 408; cf. Clinton, 524 U.S. at 449–52 (Kennedy, J., concurring). By relying on “convenience” to depart from the Constitution’s design, the juridical link doctrine dilutes its protections. Hollingsworth, 570 U.S. at 705 (citation omitted).
b. Fox responds with several precedent-based counterpoints. He initially cites two cases stating that class-certification issues are “logically antecedent” to standing concerns. Ortiz v. Fibreboard Corp., 527 U.S. 815, 831 (1999) (quoting Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 612 (1997)). Fox suggests that this statement permits a court to delay all standing questions until it certifies a class. See Payton, 308 F.3d at 680. He overlooks the context of the statement. See Mahon, 683 F.3d at 64–65. Unlike Fox’s case, Amchem and Ortiz did not concern standing challenges to named plaintiffs; they concerned standing challenges to class members. See Ortiz, 527 U.S. at 831; Amchem, 521 U.S. at 612–13. It makes sense to consider certification ahead of standing for that type of challenge. Because class members are not parties before class certification, Smith, 564 U.S. at 313, a court need not worry about their standing until it certifies the class, see Amchem, 521 U.S. at 612–13. So a class-certification denial will make a class member’s standing problem irrelevant. But the opposite is not true. Named plaintiffs are parties from the start. A court thus must immediately concern itself with their standing because jurisdictional issues precede the merits. See Steel Co., 523 U.S. at 93–101. Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 15
So Fox switches to the Supreme Court’s mootness cases in this class-action setting. The Court has held that a named plaintiff who had standing when the court certified a class may continue to seek relief for class members even if later events moot the plaintiff’s individual claim. Sosna v. Iowa, 419 U.S. 393, 399 (1975). It reasoned that the class members obtain a “legal status” of their own after the court’s class certification, and their claims can keep the controversy alive. Id.; see United States v. Sanchez-Gomez, 138 S. Ct. 1532, 1538 (2018). Here, however, Fox sued the non-Gratiot Counties before the court certified the class. At that stage, the class members did not have the separate “status” that could create a live controversy against these Counties. Sosna, 419 U.S. at 399.
That may be so, Fox responds, but the Court has also held that a named plaintiff who had standing at the outset can—in “limited” circumstances—represent a class even if the plaintiff’s claim becomes moot before a class-certification decision. Sanchez-Gomez, 138 S. Ct. at 1539; see Gerstein v. Pugh, 420 U.S. 103, 110 n.11 (1975). This narrow mootness exception applies when the “inherently transitory” nature of the plaintiff’s claim makes it impossible for the court to rule on a certification motion before the claim expires. McLaughlin, 500 U.S. at 51–52 (citation omitted). Because a plaintiff with a mootness problem may litigate on behalf of a class before class certification, Fox argues, a plaintiff with a standing problem should likewise be able to litigate on behalf of a class before that certification.
Fox wrongly conflates mootness and standing. See Laidlaw, 528 U.S. at 189. Outside the class-action context, the Court has already rejected efforts to incorporate the flexible mootness rules into the “strict” standing framework. Raines, 521 U.S. at 819. For example, a litigant may continue to pursue a now-moot claim by showing that the claim will arise in the future but evade review again. See Laidlaw, 528 U.S. at 190–91. Standing has no similar “capable of repetition yet evading review” exception. Id. If a plaintiff’s short-fused claim expires before the plaintiff can get into court, no standing exception will save it. See Steel Co., 523 U.S. at 109; Renne v. Geary, 501 U.S. 312, 320 (1991). The same logic dooms Fox’s argument here. The class-action mootness exception for inherently transitory claims resembles the individual mootness exception for claims that are capable of repetition yet evading review. Both rest on the premise that some claims expire too quickly. See Sanchez-Gomez, 138 S. Ct. at Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 16
1539–40. But if that premise does not allow for a standing exception in individual litigation, it should not allow for a standing exception in class litigation. After all, a class-action request “adds nothing” to a plaintiff’s standing. Simon, 426 U.S. at 40 n.20.
Fox thus falls back on Fallick v. Nationwide Mutual Insurance Co., 162 F.3d 410 (6th Cir. 1998). There, a plaintiff sued two Nationwide Insurance entities for their conduct in administering a health-care plan. Id. at 411. The district court held that the plaintiff lacked standing to represent class members who participated in Nationwide’s other health-care plans. Id. at 421–22. We disagreed, reasoning that the plaintiff had standing because the defendants harmed him and his requested relief would redress the injury. See id. at 421, 423. We added that whether he could represent those who participated in other plans turned on Rule 23, not standing. See id. at 423–24. Yet we made clear that “individual standing is a prerequisite” to “class actions.” Id. at 423. And we noted that a plaintiff must meet standing’s test vis-à-vis “the defendant”—foreshadowing the defendant-by-defendant approach that we take here. Id. Fallick thus confirms our result.
2. History
Although the Supreme Court’s precedent is best read to reject the juridical link doctrine, the Court does not have a decision directly addressing the doctrine. In the cases in which it has demanded that a named plaintiff meet standing’s usual test, the plaintiff lacked standing to sue all of the defendants, not just some of them. See, e.g., Warth, 422 U.S. at 518. We thus find it worthwhile to consider how the Court might approach this question from first principles.
In recent decades, the Court’s standing caselaw has placed great emphasis on history. As a general matter, the Court has read the words “Case” and “Controversy” in Article III to refer to those disagreements that were “traditionally amenable to, and resolved by, the judicial process” before the founding. Steel Co., 523 U.S. at 102; see Vt. Agency of Nat. Res. v. United States ex rel. Stevens, 529 U.S. 765, 774 (2000). As a specific matter, the Court has relied on tradition when refining its standing rules. So to evaluate whether an alleged harm meets Article III’s “concrete injury” element, the Court asks whether parties have traditionally sued over that type of injury. See TransUnion, 141 S. Ct. at 2204. Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 17
The Court has also regularly relied on history to decide whether one party who would otherwise lack standing may sue as a “representative” of another party who has standing. Ass’n of Am. Physicians, 13 F.4th at 538; see Hollingsworth, 570 U.S. at 711. Suppose, for example, that a party suffers from an impediment that undermines the party’s ability to sue (say, because the party is a child, struggles with mental-health problems, or sits behind prison walls). The Court has held that a “next friend” has standing to sue on this party’s behalf because that type of representative litigation predates the founding. Whitmore v. Arkansas, 495 U.S. 149, 161–66 (1990). Likewise, the Court has held that private “relators” have standing to bring qui tam suits on the government’s behalf because of a similarly lengthy historical record. See Stevens, 529 U.S. at 773–78; see also Sprint Commc’ns Co. v. APCC Servs., Inc., 554 U.S. 269, 274–85 (2008).
These decisions offer insights here because class actions are, at bottom, “a form of representative litigation.” 1 William B. Rubenstein, Newberg and Rubenstein on Class Actions § 1:1, at 2 (6th ed. 2022); see Ortiz, 527 U.S. at 832–33. The named plaintiffs litigate not just their own claims but also the claims of class members who become bound by the judgment. See Taylor v. Sturgell, 553 U.S. 880, 894–95 (2008). Like the qui tam relator in Stevens, then, perhaps one could attempt to justify the juridical link doctrine by showing a “long tradition” of named plaintiffs in class-like litigation suing not just the defendant who harmed them but also other defendants who inflicted similar harms on others. 529 U.S. at 774.
Yet Fox has not done so. Indeed, he has not even tried to do so. Although the Supreme Court has clarified the importance of tradition in the standing inquiry, see TransUnion, 141 S. Ct. at 2204, he offers no historical support for the juridical link doctrine. Nor have we found any “long tradition” of courts resolving suits like Fox’s. Stevens, 529 U.S. at 774.
That is not for a lack of source material. The modern class action has a lengthy pedigree in English equity. See Ortiz, 527 U.S. at 832; Geoffrey C. Hazard, Jr. et al., An Historical Analysis of the Binding Effect of Class Suits, 146 U. Pa. L. Rev. 1849, 1858–78 (1998). Equity courts traditionally required a plaintiff to add all necessary parties to a suit. See Joseph Story, Commentaries on Equity Pleadings § 72, at 74 (1838); Hazard et al., supra, at 1859. But they developed an exception if the necessary parties were so “numerous” as to make joinder Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 18
impracticable. Story, supra, at §§ 94–97, at 93–98. In that circumstance, the English Court of Chancery used the “bill of peace” to allow a named plaintiff or defendant to represent a broader group. See 7A Charles A. Wright et al., Federal Practice and Procedure § 1751, at 9–10 (4th ed. 2021); 1 John Norton Pomeroy, Treatise on Equity Jurisprudence § 246, at 257 (1881).
This type of suit arose in a variety of circumstances. The named tenants of a manor might represent all of the tenants in a dispute against the manor’s lord over hunting rights on manorial lands. How v. Tenants of Bromsgrove, 23 Eng. Rep. 277, 277 (Ch. 1681); see Stephen C. Yeazell, Group Litigation and Social Context: Toward a History of the Class Action, 77 Colum. L. Rev. 866, 868–69 (1977). Or the named parishioners of a parish might represent all of the parishioners in a dispute against the vicar over the proper tithes. Brown v. Vermuden, 22 Eng. Rep. 802, 802 (Ch. 1676); see Yeazell, supra, at 869–70. Similarly, the named partners in a partnership might represent all of the partners in a dispute against managers who embezzled funds. Chancey v. May, 24 Eng. Rep. 265, 265 (Ch. 1722); see Hazard et al., supra, at 1874. Or the named creditors of a debtor might represent all of the creditors in a dispute over the debts. See Leigh v. Thomas, 28 Eng. Rep. 201, 201 (Ch. 1751); Hazard et al., supra, at 1866–74.
Early American authorities viewed these “bills of peace” as consistent with a cognizable “case.” See Hazard et al., supra, at 1878–85. Joseph Story, for example, relied on the English precedent to suggest that heirs could sue an estate’s executors on “behalf of themselves and all other” heirs before he found the specific bill inadequate on other grounds. West v. Randall, 29 F. Cas. 718, 723 (C.C.D.R.I. 1820) (citing Brown v. Ricketts, 3 Johns. Ch. 553, 555 (N.Y. Ch. 1818)). And, in perhaps its first case on this topic, the Supreme Court allowed preachers to sue on behalf of themselves and similarly situated clergy to obtain a share of a fund held by their former church. See Smith v. Swormstedt, 57 U.S. 288, 298–303 (1853). Although “great controversy” later arose over whether a “multiplicity” of parties alone was “a sufficient reason for getting into [the] equity” courts, nobody appears to have doubted that bills of peace created justiciable cases. Zechariah Chafee, Jr., Bills of Peace with Multiple Parties, 45 Harv. L. Rev. 1297, 1312 (1932); compare Pomeroy, supra, § 269, at 293, with Tribbette v. Illinois Cent. R.R., 12 So. 32, 32–34 (Miss. 1892). Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 19
Still, this history does not support Fox’s broader standing theory. As should be evident from our description of bills of peace, a live dispute existed between the named plaintiffs (say, tenants or creditors) and the defendant or defendants they sued (say, a lord or debtor). Or, as Professor Chafee summarized things, bills of peace involved “several persons (conveniently called the multitude) on one side of a controversy, and one person (whom we may call the adversary) on the other side.” Chafee, supra, at 1297; see Pomeroy, supra, § 245, at 256–57. Despite Fox’s effort to sue 26 other Counties, he identifies no bills of peace in which, say, named tenants sued not just the lord of their own manor on behalf of their cotenants but also the lords of all other manors on behalf of unrelated tenants. The bills instead are analogous to a suit against only Gratiot County on behalf of Fox and other landowners that this County harmed. Cf. Pomeroy, supra, § 260, at 276–80; Hazard et al., supra, at 1893–97. Fox has not identified any historical analog that shows that his broader class action was “traditionally amenable to” judicial resolution. Steel Co., 523 U.S. at 102. And our reading of the history does not support the juridical link doctrine.
All told, Fox’s suit conflicts with the Supreme Court’s standing cases in the class-action context. And he offers no historical support for it. He thus may not pursue this class action.
III. Rule 23’s Requirements
Fox’s lack of standing to sue most Counties requires us to vacate the certified class. So we need not address the Counties’ alternative argument that the district court improperly analyzed Rule 23’s requirements. On remand, however, Fox might attempt to pursue a class action by joining named plaintiffs for the 26 other Counties. We will leave whether he may do so to the district court. At the least, he may attempt to certify a smaller class of Gratiot County landowners. Either choice could resurrect Rule 23 questions. Because we have our doubts about the district court’s prior order, we offer a few observations to help guide any renewed certification proceedings.
A court may not certify a class unless it engages in a “rigorous analysis” showing that the named plaintiff meets all of Rule 23’s requirements. Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 350–51 (2011) (quoting Gen. Tele. Co. of Sw. v. Falcon, 457 U.S. 147, 161 (1982)). The Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 20
plaintiff must prove, among other things, that “questions of law or fact common to class members predominate over any questions affecting only individual members[.]” Fed. R. Civ. P. 23(b)(3); Comcast Corp. v. Behrend, 569 U.S. 27, 33 (2013). This text requires a court to add up all the suit’s common issues (those that the court can resolve in a yes-or-no fashion for the class) and all of its individual issues (those that the court must resolve on an individual-by-individual basis). See Martin v. Behr Dayton Thermal Prods. LLC, 896 F.3d 405, 413 (6th Cir. 2018). The court must then qualitatively evaluate which side “predominates” over the other. See id.; see also Tyson Foods, Inc. v. Bouaphakeo, 577 U.S. 442, 453–54 (2016).
The district court here found this predominance element met because the question whether the Counties “unlawfully retained surplus proceeds is a specific legal question that is subject to generalized proof by all members of the putative class.” Fox, 2020 WL 6118487, at *10. This reasoning falls short of the “rigorous analysis” that the Supreme Court requires. Wal- Mart, 569 U.S. at 351 (citation omitted). The fact that a common issue exists does not prove that it predominates. It proves only that “there are questions of law or fact common to the class”— Rule 23(a)’s threshold requirement. Fed. R. Civ. P. 23(a)(2). And while a common issue may well exist here, the court’s high-level identification of it (did the Counties behave unlawfully?) may not even satisfy Rule 23(a). Although Fox has raised a variety of claims, the court did not describe any of their elements, let alone explain which could be proved across the board for the entire class. See Doster v. Kendall, 54 F.4th 398, 430–32 (6th Cir. 2022). In any event, the predominance requirement demands more than the commonality requirement. See Comcast, 569 U.S. at 34.
In that respect, several individual issues may pose obstacles to showing that common issues predominate. For instance, how will each landowner prove that owner’s damages? If a court may identify the amount of damages using a “formulaic calculation” for each class member, common issues may well predominate. 2 Rubenstein, supra, § 4:54, at 284; Zehentbauer Fam. Land, LP v. Chesapeake Expl., L.L.C., 935 F.3d 496, 506, 510 (6th Cir. 2019). But if fact-specific damage trials “will inevitably overwhelm” common liability questions, individual issues may predominate. Comcast, 569 U.S. at 34. We made this precise point when upholding the denial of class certification for an Ohio class of landowners who Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 21
alleged similar claims to Fox’s. See Tarrify Props, LLC v. Cuyahoga County, 37 F.4th 1101, 1106–08 (6th Cir. 2022).
In any further certification proceedings, therefore, the district court must consider how Fox will prove damages. If landowners seek only the surplus proceeds from a treasurer’s foreclosure sale (Rafaeli’s remedy), perhaps these mechanical calculations could allow the maintenance of the class action. See 952 N.W.2d at 465–66; 2 Rubenstein, supra, § 4:50, at 258. Yet, after the district court certified the class, we clarified that landowners may instead seek the difference between the fair market value of their homes and the taxes they owe—whether or not the treasurer obtained surplus proceeds at a foreclosure sale. See Hall, 51 F.4th at 195. While potentially more lucrative for class members, this damages method necessitates fact-specific evidence about the worth of each class member’s property on a property-by-property basis. It thus could “overwhelm” any common liability issues. Comcast, 569 U.S. at 34; see Tarrify, 37 F.4th at 1106.
Apart from damages, the district court should consider to what extent the Counties have unique defenses against class members. 2 Rubenstein, supra, § 4:55, at 296–98; see Sandusky Wellness Ctr., LLC v. ASD Specialty Healthcare, Inc., 863 F.3d 460, 468–69 (6th Cir. 2017); see also, e.g., Gorss Motels, Inc. v. Brigadoon Fitness, Inc., 29 F.4th 839, 845 (7th Cir. 2022). When we decided Hall, for instance, we affirmed the dismissal of several landowners’ claims on grounds of res judicata and standing. See Hall v. Meisner, 2022 WL 7478163, at *1–2 (6th Cir. Oct. 13, 2022). Because the district court certified the class without taking discovery, we have no idea how many class members might face these or similar defenses.
After the court’s certification order, moreover, the Michigan legislature adopted a process for landowners to obtain the surplus proceeds from a foreclosure sale. Mich. Comp. Laws § 211.78t(1)–(6). Yet the procedures differ depending on when a treasurer took the property. See id. § 211.78t(1)(a)–(b). This law might also supply different defenses against different class members. Cf. id. § 211.78t(11). The court should likewise consider it in the predominance inquiry. Nos. 22-1265/1272 Fox v. Saginaw County, Mich., et al. Page 22
Lastly, many landowners might have liens on their property from sources other than the government, such as the bank from which they took out a mortgage. If a landowner with $30,000 in surplus proceeds has a $20,000 mortgage, should the landowner get the full amount? Or only $10,000? Because lienholders are not class members, should an entity like the bank be able to sue separately? Or would that permit double recovery? The district court suggested in conclusory fashion that this lienholder problem did not suffice to bar class certification. See Fox, 2020 WL 6118487, at *8, *10. But it did not explain how it planned to handle the problem.
At day’s end, we do not seek to prejudge the propriety of any future class. To certify a class action, however, a district court must forecast how the parties will conduct the litigation from the certification stage through the trial to the final judgment. See Sandusky, 863 F.3d at 468. As of now, we have little understanding of how this suit would progress as a class action.
* * *
We vacate the district court’s order certifying this class action and remand for proceedings consistent with this opinion.