Anthony Daunt v. Jocelyn Benson

999 F.3d 299

• Court: Court of Appeals for the Sixth Circuit
• Decided: May 27, 2021
• Docket: 20-1734
• Status: Published

Opinion

RECOMMENDED FOR PUBLICATION Pursuant to Sixth Circuit I.O.P. 32.1(b) File Name: 21a0119p.06

UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT

┐ ANTHONY DAUNT, TOM BARRETT, AARON BEAUCHINE, │ KATHY BERDEN, STEPHEN DAUNT, GERRY HILDENBRAND, │ GARY KOUTSOUBOS, LINDA LEE TARVER, PATRICK MEYERS, │ MARIAN SHERIDAN, MARY SHINKLE, NORM SHINKLE, PAUL │ SHERIDAN, BRIDGET BEARD, CLINT TARVER (19-cv-00614), > No. 20-1734 Plaintiffs-Appellants, │ │ MICHIGAN REPUBLICAN PARTY (19-cv-00669), │ Plaintiff, │ │ v. │ │ JOCELYN BENSON, in her official capacity as Michigan │ Secretary of State; COUNT MI VOTE, doing business as │ Voters Not Politicians, │ │ Defendants-Appellees. ┘

Appeal from the United States District Court for the Western District of Michigan at Grand Rapids. Nos. 1:19-cv-00614; 1:19-cv-00669—Janet T. Neff, District Judge.

Argued: March 17, 2021

Decided and Filed: May 27, 2021

Before: MOORE, GILMAN, and READLER, Circuit Judges. _________________

COUNSEL

ARGUED: John J. Bursch, BURSCH LAW, Caledonia, Michigan, for Appellants. Erik A. Grill, OFFICE OF THE MICHIGAN ATTORNEY GENERAL, Lansing, Michigan, for Appellee Jocelyn Benson. Mark P. Gaber, CAMPAIGN LEGAL CENTER, Washington, D.C., for Appellee COUNT MI VOTE. ON BRIEF: John J. Bursch, BURSCH LAW, Caledonia, Michigan, Jason Torchinsky, HOLTZMAN VOGEL JOSEFIAK TORCHINSKY PLLC, Warrenton, Virginia, Eric E. Doster, DOSTER LAW OFFICES, PLLC, Okemos, Michigan, for Appellants. Erik A. Grill, Heather S. Meingast, OFFICE OF THE MICHIGAN ATTORNEY No. 20-1734 Daunt et al. v. Benson et al. Page 2

GENERAL, Lansing, Michigan, for Appellee Jocelyn Benson. Mark P. Gaber, Paul M. Smith, CAMPAIGN LEGAL CENTER, Washington, D.C., Graham K. Crabtree, FRASER TREBILCOCK DAVIS & DUNLAP, P.C., Lansing, Michigan, Annabelle E. Harless, CAMPAIGN LEGAL CENTER, Chicago, Illinois, for Appellee COUNT MI VOTE.

MOORE, J., delivered the opinion of the court in which GILMAN, J., joined. READLER, J. (pp. 28–43), delivered a separate opinion concurring in the judgment. _________________

OPINION _________________

KAREN NELSON MOORE, Circuit Judge. Even before the Supreme Court declared the issue of partisan gerrymandering a nonjusticiable political question in Rucho v. Common Cause, 139 S. Ct. 2484, 2491 (2019), some states had begun to address the issue head on. This case involves one such endeavor: Michigan’s Independent Citizens Redistricting Commission (the “Commission”), which was established by ballot initiative in the state’s 2018 general election. The Commission is composed of thirteen registered voters—eight who affiliate with the state’s two major political parties (four per party) and five who are unaffiliated with those parties—who must satisfy various eligibility criteria designed to ensure that they lack certain political ties.

Plaintiffs here are Michigan citizens who allege that they are unconstitutionally excluded from serving on the Commission by its eligibility criteria, in violation of the First and Fourteenth Amendments to the U.S. Constitution. They appeal the district court’s dismissal of their Complaint for failure to state a claim. See Fed. R. Civ. P. 12(b)(6). We rejected similar—if not identical—arguments to those that Plaintiffs raise here when we affirmed the district court’s earlier denial of Plaintiffs’ motion for a preliminary injunction in Daunt v. Benson, 956 F.3d 396 (6th Cir. 2020) (“Daunt I”). Plaintiffs’ arguments are no more persuasive this time around. For the reasons that follow, we hold that the Commission’s eligibility criteria do not offend the First or Fourteenth Amendments, and therefore we AFFIRM the judgment of the district court.

I. BACKGROUND

On November 6, 2018, Michiganders went to the polls and voted in favor of Proposal 18-2 on their general-election ballots. That proposal—filed by defendant-intervenor No. 20-1734 Daunt et al. v. Benson et al. Page 3

Count MI Vote d/b/a Voters Not Politicians (“VNP”)—called for a state constitutional amendment “to establish a commission of citizens with exclusive authority to adopt district boundaries for the Michigan Senate, Michigan House of Representatives and U.S. Congress.” Michigan Board of State Canvassers, Official Ballot Wording approved by the Board of State Canvassers, August 30, 2018, Voters Not Politicians, https://www.michigan.gov/documents/sos/Official_Ballot_Wording_Prop_18-2_632052_7.pdf. The commission would be composed of thirteen registered voters, randomly selected by the Secretary of State, of whom four each would be affiliated with Michigan’s two “major political parties” and five would be unaffiliated with those two parties. Id. Pursuant to the Proposal, “partisan officeholders and candidates, their employees, certain relatives, and lobbyists” would be prohibited from serving on the Commission. Id. The Proposal also called for the amendment to “[e]stablish new redistricting criteria including geographically compact and contiguous districts of equal population, reflecting Michigan’s diverse population and communities of interest. Districts shall not provide disproportionate advantage to political parties or candidates.” Id.

Consistent with Proposal 18-2, Michigan’s constitution was amended effective December 22, 2018, to establish the Commission and new redistricting criteria. See Mich. Const. art. IV, § 6 (the “Amendment”). The Commission is made up of the aforementioned thirteen registered voters in the requisite four-four-five division. See id. § 6(1). The Amendment sets forth eligibility criteria for commissioners, who must not be currently, or have been in the past six years:

(i) A declared candidate for partisan federal, state, or local office; (ii) An elected official to partisan federal, state, or local office; (iii) An officer or member of the governing body of a national, state, or local political party; (iv) A paid consultant or employee of a federal, state, or local elected official or political candidate, of a federal, state, or local political candidate’s campaign, or of a political action committee; (v) An employee of the legislature; (vi) Any person who is registered as a lobbyist agent with the Michigan bureau of elections, or any employee of such person; or No. 20-1734 Daunt et al. v. Benson et al. Page 4

(vii) An unclassified state employee who is exempt from classification in state civil service pursuant to article XI, section 5, except for employees of courts of record, employees of the state institutions of higher education, and persons in the armed forces of the state[.]

Id. at § 6(1)(b).1 In addition, the Amendment bars from the Commission “a parent, stepparent, child, stepchild, or spouse” of an individual in any of the above categories. Id. at § 6(1)(c). Commissioners are ineligible for partisan elected offices in Michigan for five years from the day of their appointment. Id. at § 6(1)(e).

The Amendment provides a detailed procedure for the selection of commissioners. The Secretary of State is to make commissioner applications available to the public no later than January 1 of the year that the decennial census is to take place. Id. at § 6(2)(a)(i). As part of the application, prospective commissioners must “attest under oath” that they satisfy the Amendment’s eligibility criteria and “either that they affiliate with one of the two political parties with the largest representation in the legislature (hereinafter, the ‘major parties’), and if so, identify the party with which they affiliate, or that they do not affiliate with either of the major parties.” Id. at § 6(2)(a)(iii). After eliminating incomplete applications and those applicants who do not satisfy the eligibility criteria, the Secretary of State divides the applications into pools for those affiliated with each of the state’s two major political parties and a third pool for unaffiliated candidates. Id. at § 6(2)(d). The Secretary of State then randomly selects sixty applications from each of the affiliated pools and eighty applications from the unaffiliated pool, using a statistical weighting method to ensure that each of the three remaining sets “as closely as possible, mirror the geographic and demographic makeup of the state.” Id. Once the Secretary of State has completed this process, she submits the applications to the majority and minority leaders of the state senate, the speaker of the state house of representatives, and the minority leader of the state house of representatives, each of whom may strike up to five applicants (up to twenty total strikes across the remaining applications). Id. at

1 Neither defendant has offered a clear explanation for why a six-year look-back period was selected for the eligibility criteria, but we note that it corresponds with the term for United States Senate, U.S. Const. art. I, § 3, cl. 1, and the maximum number of years that a regularly elected representative can serve in Michigan’s House of Representatives, see Mich. Const. art. IV, §§ 3, 54 (establishing two-year terms for representatives, limited to three total terms). No. 20-1734 Daunt et al. v. Benson et al. Page 5

§ 6(2)(e). The remaining applications are returned to the Secretary of State, who “shall randomly draw the names of four commissioners from each of the two pools of remaining applicants affiliating with a major party, and five commissioners from the pool of remaining non-affiliating applicants.” Id. at § 6(2)(f). These thirteen commissioners hold office until the “Commission has completed its obligations for the census cycle,” id. at § 6(18), and receive “compensation equal to at least [twenty-five] percent of the governor’s salary,” id. at § 6(5).

With respect to the Commission’s operations, “[a] final decision of the [C]ommission to adopt a redistricting plan requires a majority vote of the [C]ommission, including at least two commissioners who affiliate with each major party, and at least two commissioners who do not affiliate with either major party.” Id. at § 6(14)(c). Commissioners are required to “abide by the following criteria in proposing and adopting” these plans:

(a) Districts shall be of equal population as mandated by the United States constitution, and shall comply with the voting rights act and other federal laws. (b) Districts shall be geographically contiguous. Island areas are considered to be contiguous by land to the county of which they are a part. (c) Districts shall reflect the state’s diverse population and communities of interest. Communities of interest may include, but shall not be limited to, populations that share cultural or historical characteristics or economic interests. Communities of interest do not include relationships with political parties, incumbents, or political candidates. (d) Districts shall not provide a disproportionate advantage to any political party. A disproportionate advantage to a political party shall be determined using accepted measures of partisan fairness. (e) Districts shall not favor or disfavor an incumbent elected official or a candidate. (f) Districts shall reflect consideration of county, city, and township boundaries. (g) Districts shall be reasonably compact.

Id. § 6(13). Before voting on a plan, “the commission shall ensure that the plan is tested, using appropriate technology, for compliance with the criteria described above.” Id. at § 6(14)(a). Each commissioner is to “perform his or her duties in a manner that is impartial and reinforces public confidence in the integrity of the redistricting process.” Id. at § 6(10). No. 20-1734 Daunt et al. v. Benson et al. Page 6

On July 30, 2019, Plaintiffs initiated this lawsuit against Michigan Secretary of State Jocelyn Benson shortly after she released a draft of the commissioner application. See R. 1 (Compl. at ¶ 33) (Page ID #13–14). Each Plaintiff alleges that they “desire to serve on the Commission” but are barred because of their past or present partisan ties. See id. at ¶¶ 2, 39 (Page ID #3, 17). For example, Anthony Daunt alleges that he has served as a registered lobbyist in Michigan since 2013, a member and officer of the Clinton County Republican Party since 2017, and a member of the Michigan Republican Party governing body since April 18, 2017. Id. at ¶ 7 (Page ID #5); see Mich. Const. art. IV, § 6(1)(b)(iii), (vi) (excluding registered lobbyists and officers or members of local, state, or national political party governing bodies). Tom Barrett alleges that he declared his candidacy as a State Representative in September 2017 and has served in that office since January 1, 2019 having been elected in November 2018. R. 1 (Compl. at ¶ 8) (Page ID #5); see Mich. Const. art. IV, § 6(1)(b)(i)–(ii) (excluding declared candidates and elected officials). And Bridget Beard alleges that her parent, Marian Sheridan (also a plaintiff here), “has been the Grassroots Vice Chair of the Michigan Republican Party since February of 2019, and therefore a member of a governing body of a state political party.” Id. at ¶ 20 (Page ID #8); see Mich. Const. art. IV, § 6(1)(b)(iii), (1)(c) (excluding the children of members of the governing body of state political parties). Plaintiffs allege that their exclusion from the Commission based on these partisan ties violates the First and Fourteenth Amendments. See R. 1 (Compl. at ¶ 2) (Page ID #3).

On the same day that they filed their complaint, Plaintiffs filed a motion for a preliminary injunction, seeking to prevent Secretary Benson from implementing the Amendment to set up the Commission during the pendency of the case. See generally R. 4 (Mot. P.I.) (Page ID #53–92). Arguing that “there are only legal questions at issue,” Plaintiffs requested that the district court “consolidate the preliminary injunction hearing with the trial on the merits and rule on the merits in accordance with [Federal Rule of Civil Procedure] 65(a)(2).” R. 4 (Mot. P.I. at 1–2) (Page ID #53–54). In support of the motion, each Plaintiff filed a brief declaration that confirmed the partisan ties alleged in the complaint and added that “[b]ut for [the eligibility criteria and the partisan ties] I would seek a position on the Independent Redistricting Commission.” See, e.g., R. 4-3 (Daunt Decl. at ¶ 7) (Page ID #111). The district court denied Plaintiffs’ motion for a preliminary injunction, concluding that they were unlikely to succeed on the merits of their No. 20-1734 Daunt et al. v. Benson et al. Page 7

constitutional claims. See Daunt v. Benson, 425 F. Supp. 3d 856, 864 (W.D. Mich. 2019), aff’d, 956 F.3d 396 (6th Cir. 2020).

The case came before us for the first time on Plaintiffs’ appeal from the district court’s denial of their motion for a preliminary injunction. We affirmed.2 See Daunt I, 956 F.3d at 401. Observing that no federal court had passed upon the constitutional standard applicable to challenges to the eligibility criteria for an independent redistricting commission, let alone the constitutionality of eligibility criteria for a redistricting commission resembling the Commission’s, we identified and applied the two possibly applicable analytical frameworks. Id. at 406. First, we considered the standard from Anderson v. Celebrezze, 460 U.S. 780 (1983), and Burdick v. Takushi, 504 U.S. 428 (1992), which generally applies to “[c]onstitutional challenges to specific provisions of a State’s election laws,” Anderson, 460 U.S. at 789. Daunt I, 956 F.3d at 406. Applying Anderson-Burdick, we concluded that Michigan’s interests in limiting conflicts of interest in the redistricting process and in structuring its own government outweighed the limited burdens imposed by the Amendment’s eligibility criteria. Id. at 408–09. Second, we considered the unconstitutional-conditions doctrine, applicable where the government is alleged to have denied a benefit on the basis of a would-be recipient’s protected speech or association. Id. at 409. Drawing upon a long line of Supreme Court and circuit precedent upholding laws designed to cleanse federal and state employees and officials of partisan conflicts of interest, we concluded that the Amendment’s eligibility criteria survived under the unconstitutional- conditions doctrine. Id. at 410–13. Because the Amendment’s eligibility criteria would be constitutional under both standards, we declined to “choose between the two,” leaving for another day the question of the definitive standard to apply in these sorts of cases. Id. at 406

2 In Daunt I, we also addressed a companion lawsuit brought by the Michigan Republican Party (and other individuals), which the district court consolidated with Plaintiffs’ case. See 956 F.3d at 405. In addition to claims mirroring Plaintiffs’, the Michigan Republican Party alleged that the Amendment’s provision allowing applicants to self-identify as Republicans violated the Michigan Republican Party’s freedom of association, that the Commission’s composition was viewpoint-discriminatory, and that a provision restricting commissioners’ ability to speak publicly about redistricting matters violated the First Amendment. No. 1:19-cv-669, R. 1 (Compl. at ¶¶ 65–129) (Page ID #15–24); id., R. 2 (Mot. for Prelim. Inj. at 2) (Page ID #36). The district court also denied the Michigan Republican Party’s motion for a preliminary injunction, and we affirmed. See Daunt I, 956 F.3d at 401. On remand, the district court dismissed the Michigan Republican Party’s complaint, adopting the reasoning that we gave in Daunt I for concluding that its claims were unlikely to succeed. See generally R. 75 (Op. & Order) (Page ID #1036–69). The Michigan Republican Party has not appealed that decision, and thus we consider only Plaintiffs’ claims. No. 20-1734 Daunt et al. v. Benson et al. Page 8

(quoting Citizens for Legislative Choice v. Miller, 144 F.3d 916, 920 (6th Cir. 1998)). Judge Readler concurred, opining that he did not believe Anderson-Burdick to be the appropriate analytical framework, and would instead have afforded Michigan substantial deference in how it chose to organize its own government, consistent with principles of state sovereignty and federalism. Id. at 431.

Following our resolution of Daunt I, the district court considered Secretary Benson’s and VNP’s motions to dismiss Plaintiffs’ complaint for failure to state a claim. See Fed. R. Civ. P. 12(b)(6). Concluding that our reasoning in Daunt I “should be given effect” under the law-of- the-case doctrine, the district court granted both motions and entered judgment against Plaintiffs. R. 75 (Op. & Order at 20, 34) (Page ID #1055, 1069); R. 76 (J. at 1) (Page ID #1070). The district court further concluded that the same result would have followed even if the court treated Daunt I as persuasive authority, given that Plaintiffs’ legal arguments regarding the constitutionality of the Amendment’s eligibility criteria were duplicative of those presented earlier in the case. R. 75 (Op. & Order at 20) (Page ID #1055). The individual Plaintiffs timely appealed. R. 77 (Notice of Appeal at 1–2) (Page ID #1071–72).

II. DISCUSSION

A. Standard of Review

“We review de novo the district court’s grant of a motion to dismiss.” DaVita, Inc. v. Marietta Mem’l Hosp. Emp. Health Benefit Plan, 978 F.3d 326, 333 (6th Cir. 2020). Thus, like the district court below, we must determine whether the complaint “fail[s] to state a claim upon which relief can be granted,” in which case dismissal is warranted. Fed. R. Civ. P. 12(b)(6). To make this determination, we “construe the complaint in the light most favorable to the plaintiff and accept all [factual] allegations as true.” Keys v. Humana, Inc., 684 F.3d 605, 608 (6th Cir. 2012).

B. Daunt I: Law of the Case

At the outset, we must consider the impact that the law-of-the-case doctrine and Daunt I have on Plaintiffs’ appeal. “The law of the case doctrine provides that ‘when a court decides No. 20-1734 Daunt et al. v. Benson et al. Page 9

upon a rule of law, that decision should continue to govern the same issues in subsequent stages in the same case.’” Westside Mothers v. Olszewski, 454 F.3d 532, 538 (6th Cir. 2006) (quoting Scott v. Churchill, 377 F.3d 565, 569–70 (6th Cir. 2004)). Put another way, “[t]he law-of-the- case doctrine precludes reconsideration of issues decided at an earlier stage of the case.” Moody v. Mich. Gaming Control Bd., 871 F.3d 420, 425 (6th Cir. 2017) (quoting Caldwell v. City of Louisville, 200 F. App’x 430, 433 (6th Cir. 2006)). “The doctrine applies only to issues that were actually decided, whether explicitly or by necessary implication.” Id.

Although the law-of-the-case doctrine applies to higher and lower courts alike, it is primarily “intended to enforce a district court’s adherence to an appellate court’s judgment, and so is applied only loosely when we reconsider our own decisions.” Id. (quoting Miller v. Maddox, 866 F.3d 386, 390 (6th Cir. 2017)). “Therefore, while we generally will not, for prudential reasons, consider issues addressed by a prior panel, the doctrine does not limit our power of review, and we may, in exceptional circumstances, deem it necessary to depart from a prior ruling.” Id. at 425–26. There are three circumstances that we have identified as “exceptional,” such that they warrant reconsideration of a previously decided issue: “(1) where substantially different evidence is raised on subsequent trial; (2) where a subsequent contrary view of the law is decided by the controlling authority; or (3) where a decision is clearly erroneous and would work a manifest injustice.” Id. at 426 (quoting United States v. Rayborn, 495 F.3d 328, 337 (6th Cir. 2007)).

Plaintiffs contend that the law-of-the-case doctrine is inapplicable here—and thus that they need not provide us with extraordinary reasons to depart from Daunt I—because Daunt I came before us for consideration of a ruling on a motion for a preliminary injunction. “Whether a panel should treat a prior panel’s ruling on a preliminary injunction as the law of the case is tricky.” Howe v. City of Akron, 801 F.3d 718, 740 (6th Cir. 2015). On the one hand, “when a court reviewing the propriety of a preliminary injunction issues a fully considered ruling on an issue of law with the benefit of a fully developed record, then the conclusions with respect to the likelihood of success on the merits are the law of the case in any subsequent appeal.” Id. On the other hand, the law-of-the-case doctrine may be inapplicable when the legal conclusions in a preliminary-injunction decision were based on an underdeveloped record, issued under time No. 20-1734 Daunt et al. v. Benson et al. Page 10

pressures related to the circumstances of the preliminary injunction at issue, or were otherwise not conclusively decided. See id. Ultimately, we must ask whether “the appellate panel considering the preliminary injunction has issued ‘[a] fully considered appellate ruling on an issue of law.’” Id. (quoting 18B Charles Alan Wright, Arthur R. Miller, and Edward H. Cooper, Federal Practice and Procedure: Jurisdiction and Related Matters § 4478 (4th ed. 2015)). If so, “then that opinion becomes the law of the case.” Id.

Daunt I has the hallmarks of a preliminary-injunction ruling that should be afforded law- of-the-case status: we issued a “fully considered legal decision” as to the constitutionality of the Commission’s eligibility criteria based on “a decisionmaking process that included full briefing and argument without unusual time constraints.” Sherley v. Sebelius, 689 F.3d 776, 782 (D.C. Cir. 2012); see Daunt I, 956 F.3d at 413. And according to Plaintiffs themselves, who asked the district court to consolidate their motion for a preliminary injunction with a full trial on the merits, we did so with the benefit of a fully developed record. See R. 4 (Mot. P.I. at 1–2) (Page ID #53–54). Indeed, Plaintiffs’ complaint—which we construe liberally in considering a motion to dismiss for failure to state a claim—establishes the same facts that were before us in Daunt I. We accept that Plaintiffs have the partisan ties that they allege, see e.g., R. 1 (Compl. at ¶ 7) (Page ID #5), accept that they in fact “desire” to serve on the Commission, id. at ¶ 39 (Page ID #17), and accept that the eligibility criteria force Plaintiffs to choose between certain partisan activities (for a time) and the possibility of serving on the Commission. But all of this was established in Daunt I, and Plaintiffs have not directed us to any allegations in their complaint that differ from or go beyond the preliminary-injunction record. Thus, the law-of-the case doctrine applies here.

It is worth remembering that Plaintiffs argued in their motion for a preliminary injunction before the district court that “there are only legal questions at issue” in this case, effectively conceding that there was nothing that further factual development would contribute to the resolution of the case. See R. 4 (Mot. P.I. at 1–2) (Page ID #53–54). We agree. Yet, with the law-of-the case doctrine now looming over their present appeal, Plaintiffs have changed their tune somewhat, suggesting that further factual development could lead to a different outcome. Not only is this new tact inconsistent with Plaintiffs’ earlier position, but also Plaintiffs have No. 20-1734 Daunt et al. v. Benson et al. Page 11

failed to provide anything other than that conclusory assertion for why that would be the case. When pressed at oral argument, Plaintiffs’ counsel expressed that discovery would allow him to depose his own clients, but failed to explain (at oral argument, in the briefing, or anywhere else) what his clients would say beyond what they attested to in their declarations filed in support of their motion for a preliminary injunction. Oral Arg. at 12:03–12:50. Again, those declarations go no further than the allegations in Plaintiffs’ complaint, and Plaintiffs insisted that those declarations were all that was needed to decide this case in its entirety when they asked the district court to consolidate their motion for a preliminary injunction with a full trial on the merits. To the extent that Plaintiffs believed there were other or new facts material to the resolution of this case—and they have pointed to none—they could have moved on remand to amend their complaint to add them. Surely those facts would be available to Plaintiffs if the discovery they seek would be depositions of themselves. Yet Plaintiffs, represented by experienced counsel who doubtless knew that they stood little chance of succeeding in this appeal if they presented allegations mirroring the factual record before us in Daunt I, declined even to seek permission to amend their complaint to add new allegations.

At this stage, it is the factual allegations in Plaintiffs’ complaint that we must consider— it is not for us to conjure up facts that Plaintiffs declined to plead and then to draw inferences therefrom. Those allegations that Plaintiffs did plead mirror the factual record before us in Daunt I, leaving for us the question of whether a particular set of facts admits of a constitutional violation. We decided that legal question against Plaintiffs in Daunt I and do the same here. As we will explain, Plaintiffs have not identified any circumstances—let alone exceptional circumstances—that make us question our prior conclusion that the Amendment’s eligibility criteria pass constitutional muster. Even without the law-of-the case doctrine in place, we would reach the same conclusion.

C. Constitutional Analysis

Turning to the merits, Plaintiffs allege that the Amendment’s eligibility criteria offend the First and Fourteenth Amendments. Plaintiffs do not, however, claim a constitutional right to be considered for the Commission. See Moncier v. Haslam, 570 F. App’x 553, 559 (6th Cir. 2014) (concluding that there is “no federally protected interest in seeking” state office) (collecting No. 20-1734 Daunt et al. v. Benson et al. Page 12

cases). Rather, the gravamen of their constitutional attack is that the Amendment’s eligibility criteria exclude them from the Commission “based on their exercise of one or more of their constitutionally protected interests, i.e., freedom of speech . . . , right of association . . . , and/or the right to petition.” R. 1 (Compl. at ¶ 40) (Page ID #18). This amounts, according to Plaintiffs, to a violation of the First Amendment (made applicable to Michigan by the Fourteenth Amendment) and of the Equal Protection Clause of the Fourteenth Amendment.

In Daunt I, we observed that the federal courts had not considered the appropriate analytical framework for addressing constitutional challenges like Plaintiffs’. 956 F.3d at 406. We identified Anderson-Burdick and the unconstitutional-conditions doctrine as the two frameworks that could be applied to Plaintiffs’ First and Fourteenth Amendment claims, but declined to “choose between the two” because we concluded that the Commission’s eligibility criteria are constitutional under both. Id. (quoting Miller, 144 F.3d at 920). Plaintiffs have not given us a reason to reach a different conclusion here. We do not doubt that at least some of the partisan activities enumerated by the eligibility criteria involve the exercise of constitutionally protected interests, see, e.g., Elrod v. Burns, 427 U.S. 347, 370–71 (1976) (plurality op.) (explaining that Supreme Court precedent explicitly regards “political campaigning and management” as “activities . . . protected by the First Amendment”); Autor v. Pritzker, 740 F.3d 176, 182 (D.C. Cir. 2014) (“[R]egistered lobbyists are protected by the First Amendment right to petition.”), but—under either standard—Michigan’s compelling interest in cleansing its redistricting process of partisan influence justifies the limited burden that the Amendment imposes. Accordingly, we hold that the Amendment’s eligibility criteria do not violate either the First or Fourteenth Amendments.

1. The Anderson-Burdick Test

Under Anderson-Burdick, the level of scrutiny applied “depends upon the extent to which a challenged regulation burdens First and Fourteenth Amendment rights.” Burdick, 504 U.S. at 434.

[W]hen those rights are subjected to “severe” restrictions, the regulation must be “narrowly drawn to advance a state interest of compelling importance.” But when a state election law provision imposes only “reasonable, nondiscriminatory No. 20-1734 Daunt et al. v. Benson et al. Page 13

restrictions” upon the First and Fourteenth Amendment rights . . . , “the State’s important regulatory interests are generally sufficient to justify” the restrictions.

Id. “Regulations falling somewhere in between—i.e., regulations that impose a more-than- minimal but less-than-severe burden—require a ‘flexible’ analysis, ‘weighing the burden on the plaintiffs against the state’s asserted interest and chosen means of pursuing it.’” Ohio Democratic Party v. Husted, 834 F.3d 620, 627 (6th Cir. 2016) (quoting Green Party of Tenn. v. Hargett, 767 F.3d 533, 546 (6th Cir. 2014)). “While this standard is flexible, we must ultimately ‘make the “hard judgment” that our adversary system demands.’” Mays v. LaRose, 951 F.3d 775, 784 (6th Cir. 2020) (quoting Crawford v. Marion Cnty. Election Bd., 553 U.S. 181, 190 (2008) (Stevens, J., announcing the judgment of the Court)).

Our first task in applying Anderson-Burdick is thus to ascertain the magnitude of the burden that the Amendment’s eligibility criteria impose on Plaintiffs’ First and Fourteenth Amendment rights. To do this, we consider “content-neutrality and alternate means of access.” Miller, 144 F.3d at 921. A law would not be content-neutral, and would thus impose a severe burden, if it “limit[ed] political participation by an identifiable political group whose members share a particular viewpoint, associational preference, or economic status.” Id. (quoting Anderson, 460 U.S. at 793). And a law would impose a severe burden if it left “few alternate means of access to the ballot,” “restrict[ing] ‘the availability of political opportunity.’” Id. (quoting Anderson, 460 U.S. at 793).

Plaintiffs have given us no reason to reconsider our conclusion in Daunt I that the Amendment’s eligibility criteria impose, at most, a moderate burden on Plaintiffs’ First and Fourteenth Amendment rights. 956 F.3d at 408–09. As we explained before, the eligibility criteria do not raise substantial concerns regarding content-neutrality because they do not burden candidates “based on the content of protected expression, party affiliation, or inherently arbitrary factors such as race, religion, or gender.” Id. at 407 (quoting Miller, 144 F.3d at 922). Indeed, the eligibility criteria apply regardless of the prospective commissioner’s affiliation with the Republican, Democratic, or other political party, and do not inquire as to (let alone discriminate according to) candidates’ “views on any of the substantive ‘issues of the day,’ such as taxes or abortion.” Miller, 144 F.3d at 922. Although the Amendment’s eligibility criteria are not purely No. 20-1734 Daunt et al. v. Benson et al. Page 14

“generally applicable” or “nondiscriminatory,” Obama for Am., 697 F.3d at 433, insofar as they target individuals with current or recent political connections, the Supreme Court has broadly upheld similar restrictions on partisan activities in government. See, e.g., U.S. Civil Serv. Comm’n v. Nat’l Ass’n of Letter Carriers, AFL-CIO, 413 U.S. 548, 550 (1973) (“Letter Carriers”); Broadrick v. Oklahoma, 413 U.S. 601, 602 (1973). Nor do the Amendment’s eligibility criteria present substantial concerns regarding alternate means of access. Daunt I, 956 F.3d at 408. The criteria restrict eligibility for an array of individuals with partisan ties, but those restrictions look back only to the past six years, Mich. Const. art. IV, § 6(1)(b), and the Supreme Court has explained that “[a] ‘waiting period’ is hardly a significant barrier to candidacy,” Clements v. Fashing, 457 U.S. 957, 967 (1982) (plurality op.) (characterizing a two-year waiting period as imposing a “de minimis burden” in a pre-Anderson-Burdick decision); see also Chimento v. Stark, 414 U.S. 802 (1973) (summarily affirming a three-judge district court opinion upholding the constitutionality of New Hampshire’s seven-year durational residency eligibility requirement for the office of Governor). In the interim, prospective candidates are free to participate in Michigan’s redistricting process through the various opportunities for public hearings and comment that the Amendment establishes. See Mich. Const. art. IV, § 6(9), (14)(b).

Weighed against Michigan’s asserted interests, the burdens imposed by the Amendment’s eligibility criteria do not offend the First or Fourteenth Amendments. First, Michigan has a compelling interest “in limiting the conflict of interest implicit in legislative control over redistricting.” Ariz. State Legislature v. Ariz. Indep. Redistricting Comm’n, 576 U.S. 787, 821 (2015) (quoting Bruce Cain, Redistricting Commissions: A Better Political Buffer?, 121 Yale L. J. 1808, 1808 (2012)) (alteration omitted). Independent commissions like Michigan’s “curb partisan gerrymandering, by which the majority in the legislature draws district lines to their party’s advantage.” Id. at 822. “They thus impede legislators from choosing their voters instead of facilitating the voters’ choice of their representatives.” Id. at 821. Second, and similarly, Michigan has a compelling interest in avoiding the appearance of conflicts of interest in its redistricting. “Confidence in the integrity of our electoral processes is essential to the functioning of our participatory democracy.” Purcell v. Gonzalez, 549 U.S. 1, 4 (2006). The appearance of conflicts of interest related to partisan gerrymandering, no different from voter fraud, “drives honest citizens out of the democratic process and breeds distrust of our No. 20-1734 Daunt et al. v. Benson et al. Page 15

government.” Id. Thus, even if we assume that candidates barred by past partisan activity or by their spouse’s, parent’s, or child’s partisan activity do not suffer from an actual conflict of interest, the appearance of a conflict that is inherent in that relationship is a compelling reason for Michigan to restrict their eligibility for the Commission. See id. Third, and finally, “[a]s a sovereign polity, Michigan has a fundamental interest in structuring its government.” Miller, 144 F.3d at 923. The Amendment’s eligibility criteria directly advance each of these interests, curtailing partisan influence on redistricting and the appearance thereof according to a state constitutional amendment adopted by direct ballot initiative. In light of the relatively limited burden that the eligibility criteria impose, we conclude that they are constitutional under Anderson-Burdick.3

Although Plaintiffs argue that we should disregard Daunt I and our Anderson-Burdick analysis therein because it rested upon an underdeveloped factual record, we fail to see how further factual development could change our prior reasoning. True, Anderson-Burdick can, in many if not most cases, be a fact-intensive inquiry, see Tennessee State Conf. of N.A.A.C.P. v. Hargett, 420 F. Supp. 3d 683, 700–01 (M.D. Tenn. 2019), but we have not shied away from disposing of Anderson-Burdick claims at the motion-to-dismiss stage where a plaintiff’s allegations “failed as a matter of law.” Comm. to Impose Term Limits on Ohio Supreme Ct. & to Preclude Special Legal Status for Members & Emps. of Ohio Gen. Assembly v. Ohio Ballot Bd., 885 F.3d 443, 448 (6th Cir. 2018) (collecting cases); see also Hawkins v. DeWine, 968 F.3d 603, 607 (6th Cir. 2020). Where, as here, the alleged severity of the burdens imposed can be gleaned from the face of the challenged law and they can be weighed against the asserted state interests, dismissal on the pleadings is warranted. See Ohio Ballot Bd., 885 F.3d at 448. Plaintiffs offer

3 Our concurring colleague takes issue with our characterization of the state’s interest here as “compelling.” Concurrence at 30. His criticism is curious, to say the least, given his insistence that courts respect states’ interests in structuring their own governments. See Concurrence at 37–43. Indeed, our concurring colleague himself characterized that interest as “compelling” in Daunt I, attributing that judgment to Miller, 144 F.3d at 923, a case that he renews his reliance on here (without, it must be said, his earlier invocation of the suddenly dirty word “compelling”), compare Daunt I, 956 F.3d at 427 (Readler, J., concurring), with Concurrence at 41. Suffice it to say, we think it uncontroversial to conclude that the state interest at stake here is compelling when compared to the other interests that have earned that moniker. See, e.g., Eu v. San Francisco Cnty. Democratic Cent. Comm., 489 U.S. 214, 231 (1989) (“A State indisputably has a compelling interest in preserving the integrity of its election process.”); Ky. Right to Life, Inc. v. Terry, 108 F.3d 637, 651 (6th Cir. 1997) (recognizing state’s “compelling interest in combatting perceived corruption”). No. 20-1734 Daunt et al. v. Benson et al. Page 16

only the vague contention that further factual development could demonstrate the severity of the burdens imposed by the Amendment’s eligibility criteria and the absence of the conflicts of interest that the Amendment is predicated on. But the burden—the duration for which a prospective commissioner or their relation must abstain from enumerated partisan activities in order to satisfy the eligibility criteria—is plain according to the Amendment’s fixed terms. That burden is, at most, moderate. See Obama for Am., 697 F.3d at 429 (“Most cases fall in between these two extremes.”). And even assuming some individuals excluded from the Commission lack an actual conflict of interest (say, because they are excluded because of a relative’s partisan ties, or their own past partisan ties), Michigan’s interest in avoiding the appearance of such conflicts in order to preserve confidence in its redistricting is ample justification for the burdens imposed.

Perhaps more importantly, as explained above, Plaintiffs’ cries for further factual development ring hollow when compared to their stance at the preliminary injunction stage, when they insisted that “there are only legal questions at issue” in this case. R. 4 (Mot. P.I. at 1) (Page ID #53). In support of their motion for a preliminary injunction, each Plaintiff offered a declaration confirming the bare allegations in their complaint but going no further. Compare, e.g., R. 1 (Compl. at ¶ 7) (Page ID #5) (allegations regarding Anthony Daunt’s political affiliations) with R. 4-3 (Daunt Decl. at ¶¶ 1–7) (Page ID #111). Then, on remand, with every reason to bolster their factual allegations if they could, in order to distinguish their complaint from the preliminary injunction record before us in Daunt I, Plaintiffs did not so much as seek leave to amend their complaint. To put a fine point on it, we had everything we needed in Daunt I to issue a definitive ruling as to the constitutionality of the Amendment’s eligibility criteria, and we did so.4

4 Our concurring colleague is right to point out that the application of Anderson-Burdick in many cases will require development beyond the pleadings—our decision today does not change that. See Concurrence at 35. What our decision does do is recognize that in this case—which comes before us in an uncommon posture—there is nothing that would come from further factual development given the Plaintiffs’ arguments to this point and the clarity afforded by the Amendment. Although more typical Anderson-Burdick cases will turn on factual questions about whether voters’ burdens or the state’s interests are as impacted or as strong as they claim, this case involves a “relatively simple . . . restriction, with few if any facts in dispute” where the Concurrence acknowledges ruling on the pleadings would be warranted. Concurrence at 36. No. 20-1734 Daunt et al. v. Benson et al. Page 17

Plaintiffs’ discussion of Anderson-Burdick does not otherwise take issue with our analysis under the standard, but instead criticizes our choice to apply that standard in the first place. Indeed, Plaintiffs dedicate a significant portion of their briefing to arguments for why Anderson-Burdick is inapplicable to this sort of case. See Appellant’s Br. at 15–24. Our concurring colleague joins them, making the same points that he took up against Anderson-Burdick in Daunt I, 956 F.3d at 422–26. Indeed, he dedicates a full twelve pages of his fourteen-page opinion to railing against the perceived vagaries of Anderson-Burdick. We think that Plaintiffs’ arguments—and the Concurrence’s—on this point are misplaced.

Most obviously, as in Daunt I, we have stopped short from conclusively adopting Anderson-Burdick as the controlling standard for challenges to the independent redistricting commissions’ eligibility criteria. Instead, we have identified the two frameworks that we believe could apply, but declined to “choose between the two,” see Miller, 144 F.3d at 920, because we hold the eligibility criteria are constitutional under each framework. Indeed, we recently took a similar approach in Libertarian Party of Ohio v. Wilhem, 988 F.3d 274, 279 (6th Cir. 2021), acknowledging Anderson-Burdick as a possible standard for assessing the constitutionality of the Ohio Election Commission’s composition, but assuming without deciding that the unconstitutional-conditions doctrine applied because the parties in that case agreed that it did. In doing so, we have appropriately avoided overreach—deciding only those questions “necessary to the outcome of the case, not merely to the outcome of a debate regarding an abstract legal question,” United States v. Hardin, 539 F.3d 404, 415 n.5 (6th Cir. 2008)—while contributing to the incremental development of a relatively novel area of law. In this case, we have identified those standards that could apply to a challenge like Plaintiffs’, and we have narrowed the choices to two: Anderson-Burdick and the unconstitutional-conditions doctrine.

In any event, we do not find Plaintiffs’ arguments—or the Concurrence’s—to be persuasive. As this court has explained, Anderson-Burdick applies to a wide array of claims touching on the election process, including First Amendment and Equal Protection Clause claims like Plaintiffs’. See, e.g., Obama for Am., 697 F.3d at 430. But see Memphis A. Philip Randolph Inst. v. Hargett, 978 F.3d 378, 402 (6th Cir. 2020) (Moore, J., dissenting) (reasoning that the standard procedural due process analysis, not Anderson-Burdick, should apply to a claim that No. 20-1734 Daunt et al. v. Benson et al. Page 18

Tennessee’s signature-verification law violated procedural due process). Although Anderson- Burdick has not been conclusively applied in this context, we think that it is uniquely well suited to this sort of case given that it is carefully calibrated to ensure that “democratic processes” are “fair and honest,” Burdick, 504 U.S. at 434 (quoting Storer v. Brown, 415 U.S. 724, 730 (1974)), and to “maintain the integrity of the democratic system,” id. at 441, concerns that are central to the Amendment here and its eligibility criteria. Ultimately, measures designed to prevent partisan gerrymandering serve the same ends as measures that more directly concern the administration of elections—“partisan gerrymanders . . . are incompatible with democratic principles,” Ariz. Indep. Redistricting Comm’n, 576 U.S. at 791 (internal quotation marks and original brackets omitted), and it is those principles that other elections laws properly protect, see Storer, 415 U.S. at 730.

The Supreme Court has found the Anderson-Burdick standard to be good enough in an array of highly sensitive cases since 1983, despite its need for balancing. To be sure, balancing tests are more nuanced than hard-and-fast rules or blind deference to the states, but we are appointed to exercise our judgment and sometimes the difficult questions we are faced with lack easy answers. That is especially so in cases like these where competing interests meet head on and we are called upon to exercise the judgment that we were appointed to employ. To acknowledge that simple truth—that not every case has a bright-line answer—is not “to enhance judicial power,” Concurrence at 33, but rather it is to exercise judicial power faithfully. Balancing tests are quite common and are used in many contexts. A balancing test, for example, is used to determine whether drug-testing programs are reasonable under the Fourth Amendment. Nat’l Treasury Emps. Union v. Von Raab, 489 U.S. 656, 665–66 (1989) (balancing “the individual’s privacy expectations against the Government’s interests”). In the First Amendment context, the Pickering balancing test, based on the Supreme Court decision in Pickering v. Board of Education, 391 U.S. 563 (1968), is commonly used to determine the limits of public employees’ free-speech rights. The Federal Rules of Evidence incorporate a balancing test in one of its most pivotal rules––Rule 403––which requires that judges balance the probative value of evidence against its potentially prejudicial impact. Anderson-Burdick, like the tests mentioned above, strikes a balance between competing rights and interests. And––also like the aforementioned tests––it does not give judges unlimited discretion. No. 20-1734 Daunt et al. v. Benson et al. Page 19

Turning back to Plaintiffs’ reliance on specific cases, we do not believe that Moncier v. Haslam, 570 F. App’x 553, 558 (6th Cir. 2014), benefits their argument that Anderson-Burdick is inapplicable. In that case, we considered a challenge to a Tennessee law allowing the Governor temporarily to fill judicial vacancies by appointment and requiring appointees then to run in a retention election. Id. at 553–54. In the course of concluding that the plaintiff lacked standing to sue, we remarked that Anderson-Burdick bore “little weight” on the case because the plaintiff had not identified a constitutional interest that the law offended. Id. at 559. Here, by contrast, Plaintiffs’ claims implicate their First and Fourteenth Amendment rights, and do so in the context of redistricting, which is intimately related to the electoral process. Thus, even assuming that Moncier were a published opinion and that its reference to Anderson-Burdick were not dicta, it would not compel us to eschew Anderson-Burdick here.

Nor is McIntyre v. Ohio Elections Commission, 514 U.S. 334 (1995), of any aid to Plaintiffs. There, the Supreme Court addressed the constitutionality of an Ohio statute prohibiting the distribution of anonymous campaign literature. Id. at 336. In holding the law to be unconstitutional, the Court rejected Anderson-Burdick and applied a traditional First Amendment analysis because the Ohio law was “a regulation of pure speech” and thus was distinct from “an ordinary election restriction.” Id. at 345. Although the Ohio law was viewpoint neutral, it directly regulated the content of speech by requiring campaign literature to contain information regarding its issuer. Id. Here, by contrast, the Amendment’s eligibility criteria are not regulations of pure speech, and so McIntyre does little to inform the analysis. We think that the otherwise widespread use of Anderson-Burdick to assess election-related laws counsels in favor of its applicability here.

Most misplaced of all is Plaintiffs’ reliance on the concurrence in Daunt I, which was just that—a concurrence. It reflected our concurring colleague’s own analysis of the issues and his alone. That has not changed this time around as the concurrence in Daunt I is now our concurring colleague’s primary authority in what will be his Daunt II concurrence. In any event, although our concurring colleague rejected the Anderson-Burdick framework as inapplicable to a determination of the constitutionality of the Amendment’s eligibility criteria in Daunt I, he would have taken—and continues to insist upon—an approach that is more deferential to No. 20-1734 Daunt et al. v. Benson et al. Page 20

Michigan’s interest in “structuring its government, including how it seeks to administer elections.” Daunt I, 956 F.3d at 426–27. Thus, we question Plaintiffs’ dedication to the Daunt I concurrence, given that, if anything, their claims fare worse under it than under our majority opinion in Daunt I.

In the end, Plaintiffs have not given us any reasons—exceptional or otherwise—to depart from our reasoning in Daunt I. If Anderson-Burdick does apply, as we believe it does, then the Amendment’s eligibility criteria survive Plaintiffs’ First and Fourteenth Amendment challenges. If not, then the unconstitutional-conditions doctrine applies, and begets the same result, as we shall now explain.

2. Unconstitutional Conditions

The unconstitutional-conditions doctrine restricts the government’s ability to deny a valuable government benefit

to a person on a basis that infringes his constitutionally protected interests— especially, his interest in freedom of speech. For if the government could deny a benefit to a person because of his constitutionally protected speech or associations, his exercise of those freedoms would in effect be penalized and inhibited. This would allow the government to “produce a result which [it] could not command directly.” Such interference with constitutional rights is impermissible.

Perry v. Sindermann, 408 U.S. 593, 597 (1972) (original alteration) (quoting Speiser v. Randall, 357 U.S. 513, 526 (1958)). Plaintiffs contend that appointment to the Commission, which comes with a salary of about $40,000, is a valuable government benefit and that the Amendment’s eligibility criteria are unconstitutional because the criteria penalize their partisan activities, which are protected by the First and Fourteenth Amendments. Supreme Court precedent—which recognizes that “neither the right to associate nor the right to participate in political activities is absolute,” Letter Carriers, 413 U.S. at 567—forecloses Plaintiffs’ argument.

The Supreme Court has applied the unconstitutional-conditions doctrine in an array of contexts. See Perry, 408 U.S. at 597 (“We have applied this general principle to denials of tax exemptions, unemployment benefits, and welfare payments.” (internal citations omitted)). Most relevant here, however, are the litany of cases where the Court has held that laws designed to No. 20-1734 Daunt et al. v. Benson et al. Page 21

avoid partisan conflicts of interest do not run afoul of the First Amendment or the Equal Protection Clause. As we explained in Daunt I, we find three cases particularly compelling in concluding that the Commission’s eligibility criteria are constitutional under the unconstitutional-conditions doctrine.

The first two cases involved the Hatch Act’s mandate that “[n]o officer or employee in the executive branch of the Federal Government . . . shall take any active part in political management or in political campaigns.” United Pub. Workers of Am. (C.I.O.) v. Mitchell, 330 U.S. 75, 82 (1947). The Court upheld that provision in Mitchell, explaining that “Congress and the President are responsible for an efficient public service. If, in their judgment, efficiency may be best obtained by prohibiting active participation by classified employees in politics as party officers or workers, we see no constitutional objection.” 330 U.S. at 99. The Court emphasized that the restriction targeted only “partisan political activity” and that “[e]xpressions, public or private, on public affairs, personalities and matters of public interest, not an objective of party action, are unrestricted by law so long as the Government employee does not direct his activities toward party success.” Id. at 100. The Court dismissed the suggestion that no harm could be done by federal employees engaging in these activities in their “free time” outside work hours. Id. at 95. “The influence of political activity by government employees, if evil in its effects on the service, the employees or people dealing with them, is hardly less so because that activity takes place after hours.” Id. Nor did the Court think it dispositive that the Hatch Act covered employees for whom partisan activities might not have directly impacted the effective performance of their duties, reasoning that “[f]or regulation of employees it is not necessary that the act regulated be anything more than an act reasonably deemed by Congress to interfere with the efficiency of the public service.” Id. at 101.

The Court reaffirmed Mitchell in Letter Carriers, broadly endorsing the constitutionality of laws designed to purge the government workforce of partisan conflicts of interest. In “unhesitatingly affirm[ing] the Mitchell holding,” the Court asserted that that if Congress

forbade activities such as organizing a political party or club; actively participating in fund-raising activities for a partisan candidate or political party; becoming a partisan candidate for, or campaigning for, an elective public office; actively managing the campaign of a partisan candidate for public office; No. 20-1734 Daunt et al. v. Benson et al. Page 22

initiating or circulating a partisan nominating petition or soliciting votes for a partisan candidate for public office; or serving as a delegate, alternate or proxy to a political party convention[,]

such actions would “unquestionably be valid.” 413 U.S. at 556 (“Our judgment is that neither the First Amendment nor any other provision of the Constitution invalidates a law barring this kind of partisan political conduct by federal employees.”). Detailing the long history of restrictions like the Hatch Act, the Court explained that

the judgment of Congress, the Executive, and the country appears to have been that partisan political activities by federal employees must be limited if the Government is to operate effectively and fairly, elections are to play their proper part in representative government, and employees themselves are to be sufficiently free from improper influences.

Id. at 564. Elaborating on the rationale for upholding these sorts of laws, the Court explained that “it is not only important that the Government and its employees in fact avoid practicing political justice, but it is also critical that they appear to the public to be avoiding it, if confidence in the system of representative Government is not to be eroded to a disastrous extent.” Id. at 565 (emphasis added). The Court also remarked upon the fact that each of the fifty states had adopted measures similar to the Hatch Act, restricting their own employees’ partisan activities, one of which it had upheld in a companion case to Letter Carriers. Id. at 563 & n.11 (citing Broadrick, 413 U.S. 601).

The third case, Clements, 457 U.S. 957, involved two sections of the Texas constitution, the first of which prohibited certain officials from holding a seat in the state legislature prior to the expiration of their terms of office, and the second of which required an officeholder to resign before running for any other elected office. The Court relied on Mitchell and Letter Carriers to uphold both provisions, despite acknowledging that their practical effect was to impose a waiting period of up to two years for candidacy. Clements, 457 U.S. at 967 (plurality op.); id. at 972–73. Explaining that “[a] ‘waiting period’ is hardly a significant barrier to candidacy,” id. at 967 (plurality op.), the Court held that, whether under the Equal Protection Clause or the First Amendment, “the burden on appellees’ First Amendment interests in candidacy are so insignificant that the classifications of [the two sections at issue] may be upheld consistent with No. 20-1734 Daunt et al. v. Benson et al. Page 23

traditional equal protection principles.” Id. at 971.5 The Court plurality’s application of rational-basis review under the Equal Protection Clause “dispose[d] of” the challengers’ First Amendment claim. Id.

Together, Mitchell, Letter Carriers, and Clements are an insurmountable barrier for Plaintiffs. Just as the Supreme Court in those cases permitted federal and state governments to restrict the “partisan political activity” of federal employees, Mitchell, 330 U.S. at 100, and state officeholders, Clements, 457 U.S. at 972, we discern no constitutional infirmity in Michigan making the temporary forbearance from such activity a condition of sitting on an independent redistricting commission. Like the broad swath of restrictions that the Court endorsed in Letter Carriers, the Amendment’s eligibility criteria

are not aimed at particular parties, groups, or points of view, but apply equally to all partisan activities of the type described. They discriminate against no racial, ethnic, or religious minorities. Nor do they seek to control political opinions or beliefs, or to interfere with or influence anyone’s vote at the polls.

413 U.S. at 564. In sum, “[o]ur judgment is that neither the First Amendment nor any other provision of the Constitution invalidates,” id. at 556, the Amendment’s eligibility criteria, which are well within the range of constitutionally permitted measures by which Michigan may “reduce the hazards to fair and effective government,” id. at 565, posed by partisanship in its redistricting process. Efforts to purge conflicts of interest from the democratic process “have been commonplace for over 200 years,” Nev. Comm’n on Ethics v. Carrigan, 564 U.S. 117, 122 (2011), and we are loath to disturb this longstanding practice, particularly when “public confidence in the integrity of the redistricting process” is at stake. Mich. Const. art. IV, § 6(10); see Rucho, 139 S. Ct. at 2507 (noting states’ interests in “restricting partisan considerations in districting” and citing the Michigan Amendment as an example).

Plaintiffs’ efforts to distinguish this case from Mitchell, Letter Carriers, and Clements are no more persuasive now than when they raised the same arguments in Daunt I. Once again, they

5 Even Justice Brennan’s dissenting opinion in Clements, which faulted the plurality for focusing its rational-basis review on whether the “class of candidates or voters that was burdened was somehow suspect” (for example, based on their wealth) instead of focusing on “the impact on the First Amendment rights of candidates and voters,” acknowledged that “some greater deference may be due the State because these restrictions affect only public employees.” 457 U.S. at 977–78 n.2 (Brennan, J., dissenting). No. 20-1734 Daunt et al. v. Benson et al. Page 24

argue that Mitchell, Letter Carriers, and Clements should not apply because the restrictions at issue in those cases applied only to “current” federal employees and state officials. Appellant’s Br. at 41. “These cases,” Plaintiffs contend, involve limitations on partisan activity that “are thus a far cry from the extensive six-year, post-service period at issue in this case” and “the fit between the government’s asserted interests on the one hand and [the eligibility criteria’s] categorical restrictions on the other is insufficient to pass constitutional muster.” Id. at 41–42. Yet we rejected this very argument in Daunt I, 956 F.3d at 411, and we reject it again here. Even if we assume that individuals barred from appointment to the Commission by their past partisan activities or by their relationship to individuals who would themselves be barred do not suffer from an actual conflict of interest, Michigan still has a compelling interest in addressing the appearance of a conflict of interest in order to ensure confidence in its redistricting process. See Letter Carriers, 413 U.S. at 565. That interest is sufficient to justify the eligibility criteria’s restriction of those who are not themselves “actively partisan,” Mitchell, 330 U.S. at 98, but whose recent partisan involvement, or whose association with active or recent partisans, could create the appearance that the Commission is staffed by political insiders. See Letter Carriers, 413 U.S. at 565. If the two-year waiting period implicated by the sections of the Texas constitution at issue in Clements was no more than a “de minimis” burden, then we do not think that a six-year waiting period renders the Amendment’s eligibility criteria constitutionally infirm. 457 U.S. at 967 (plurality op.).

Furthermore, we are unpersuaded by Plaintiffs’ argument that “heightened scrutiny”— either strict scrutiny or exacting scrutiny—should apply to this case. This is a twist on their earlier argument in Daunt I, where they advocated for exacting scrutiny, but not strict scrutiny. No matter—our earlier reasoning again forecloses Plaintiffs’ argument. For one thing, Mitchell, Letter Carriers, and Clements—none of which applied the scrutiny that Plaintiffs call for here— counsel strongly against this argument. Moreover, as we explained in Daunt I, our sister circuits have not hesitated to apply a less-than-exacting standard of review to challenges to the eligibility criteria for elected offices—a context that is highly analogous to eligibility criteria for an independent redistricting commission. For example, in Fletcher v. Marino, the Second Circuit applied rational-basis review to a law restricting certain political party officers from being elected to community school boards. 882 F.2d 605, 613 (2d Cir. 1989). In doing so, the court No. 20-1734 Daunt et al. v. Benson et al. Page 25

pointed to Mitchell, Letter Carriers, and Clements for support, and explained that “laws that implicate, in a limited fashion, a person’s rights to participate in politics and to serve as an elected official have survived review under the First Amendment and have not been subjected to strict scrutiny.” Id. at 612–13; see also Grizzle v. Kemp, 634 F.3d 1314, 1326 (11th Cir. 2011). We continue to believe that, if anything, the case for rational-basis review to apply here is even stronger. The most salient difference between this case and Fletcher and Grizzle is that those cases involved elected positions, thereby implicating voters’ access to the ballot by limiting the candidates for whom they can cast a ballot. Fletcher, 882 F.2d at 614; Grizzle, 634, F.3d at 1321; see Bullock v. Carter, 405 U.S. 134, 142–43 (1972). Here, positions on the Commission are not elected, and thus the eligibility criteria do not raise ballot-access concerns, which underscores the applicability of rational-basis review. Cf. Bullock, 405 U.S. at 142–43. And as we have explained, here and in Daunt I, that standard is met. Michigan has a compelling interest in ridding its redistricting process of partisan conflicts of interest and the appearance thereof, and the Amendment’s eligibility criteria are a reasonable mechanism by which to achieve that end. Cf. Mitchell, 330 U.S. at 100 (“Congress may reasonably desire to limit party activity of federal employees so as to avoid a tendency toward a one-party system. It may have considered that parties would be more truly devoted to the public welfare if public servants were not over active politically.”). Under this standard, Plaintiffs’ complaints that the eligibility criteria are under and over inclusive are unavailing—Michigan is free to “regulate ‘one step at a time, addressing itself to the phase of the problem which seems most acute.’” Clements, 457 U.S. at 969 (plurality op.) (quoting Williamson v. Lee Optical Co., 348 U.S. 483, 489 (1955)).

Finally, we remain unpersuaded by Plaintiffs’ continued reliance on the Supreme Court’s political-patronage cases, a subset of the Court’s unconstitutional-conditions doctrine precedent. As in Daunt I, we think that those cases harm rather than help Plaintiffs’ case. The Court’s political-patronage cases, most notably Elrod, 427 U.S. 347, Branti v. Finkel, 445 U.S. 507 (1980), and Rutan v. Republican Party of Illinois, 497 U.S. 62 (1990), condemn “the conditioning of public employment on political faith,” Elrod, 427 U.S. at 357 (plurality op.), except where “the hiring authority can demonstrate that party affiliation is an appropriate requirement for the effective performance of the public office involved,” Branti, 445 U.S. at 518. Thus, in Branti, the Court held that “the continued employment of an assistant public defender No. 20-1734 Daunt et al. v. Benson et al. Page 26

cannot properly be conditioned upon his allegiance to the political party in control of the county government,” 445 U.S. at 519, and held in Rutan that “promotion, transfer, recall, and hiring decisions involving low-level public employees may [not] be constitutionally based on party affiliation and support,” 497 U.S. at 65.

Here, by contrast, the Amendment’s eligibility criteria apply across the board to those associated with professional politics, regardless of the candidate’s association with a particular political party, belief, or ideology. Being fired from one’s job because one is a Republican “unquestionably inhibits protected belief and association,” Elrod, 427 U.S. at 359 (plurality op.), in a way that the Amendment here unquestionably does not. See Wilhem, 988 F.3d at 283 (“There is no comparison to be drawn from laws which afford equality of opportunity to all political parties, and those that expressly prohibit a person from government employment because of a protected characteristic.”). Indeed, the Court recognized as much in Elrod, where it distinguished patronage practices from laws like those in Mitchell, Letter Carriers, and Clements, addressing partisan conflicts of interest. See Elrod, 427 U.S. at 370. Whereas patronage practices “can result in the entrenchment of one or a few parties to the exclusion of others,” and are “a very effective impediment to the associational and speech freedoms which are essential to a meaningful system of democratic government,” laws restricting partisan activity are permissible insofar as they “safeguard the core interests of individual belief and association.” Id. at 369–71. In other words, barring government employees from “taking an active part in political management or political campaigns,” Letter Carriers, 413 U.S. at 554, serves to “safeguard the core interests of individual belief and association” that patronage-based systems undermine. Elrod, 427 U.S. at 371. The Court’s line of patronage cases thus supports the conclusion that the eligibility criteria do not impose an unconstitutional condition on Plaintiffs.

As was the case under Anderson-Burdick, Plaintiffs have failed to present any reasons, let alone exceptional reasons, for why their claims should survive under the unconstitutional-conditions doctrine. Accordingly, they have failed to state a claim under the First or Fourteenth Amendments, and the district court properly dismissed Plaintiffs’ Complaint. No. 20-1734 Daunt et al. v. Benson et al. Page 27

III. CONCLUSION

Ultimately, under the Anderson-Burdick and the unconstitutional-conditions frameworks—the only two analytical frameworks that could apply—the same flaws in Plaintiffs’ claims shine through: Michigan’s interest in cleansing its redistricting process of political conflicts of interest and the appearance thereof is ample justification for the limited burdens that the Amendment’s eligibility criteria impose on those who would like to be considered for the Commission. Although claims of unconstitutional partisan gerrymandering may be nonjusticiable, Rucho, 139 S. Ct. at 2491, Michigan is free to employ its political process to address the issue head on. It did so, adopting the Amendment after Michiganders overwhelmingly voted in favor of Proposal 18-2, and its eligibility criteria for the Commission do not offend the First or Fourteenth Amendments.6

For the foregoing reasons, we AFFIRM the judgment of the district court.

6 The Parties disagree as to whether the eligibility criteria (or portions thereof) are severable, such that the rest of the Amendment would be left in place if we were to hold the eligibility criteria to be unconstitutional. Our conclusion that the Commission’s eligibility criteria do not offend the First or Fourteenth Amendments renders these arguments moot (in the colloquial sense). Were we to reach the question, however, we would agree with Defendants that, to the extent that aspects of the eligibility criteria were to offend the Constitution—though they do not—those portions of the Amendment would be severable, and thus their invalidity would leave the remainder of the Amendment intact. “Severability is of course a matter of state law.” Leavitt v. Jane L., 518 U.S. 137, 139 (1996). The Amendment specifically includes a severability clause, which provides that “[a]ny provision held invalid is severable from the remaining portions of this section.” Mich. Const. art. IV, § 6(20). That clause must be given effect according to its plain meaning. See County of Wayne v. Hathcock, 684 N.W.2d 765, 779 (Mich. 2004) (“The primary objective in interpreting a constitutional provision is to determine the text’s original meaning to the ratifiers, the people, at the time of ratification.”); Bond v. Pub. Schs. of Ann Arbor Sch. Dist., 178 N.W.2d 484, 487 (Mich. 1970) (“The first rule a court should follow in ascertaining the meaning of words in a constitution is to give effect to the plain meaning of such words as understood by the people who adopted it.”); cf. In re Apportionment of State Legislature--1982, 321 N.W.2d 565, 580 (Mich. 1982) (invalidating, in the absence of a severability clause, the entirety of Michigan’s earlier apportionment scheme after the United States Supreme Court held the state’s formula for apportionment was unconstitutional under the Equal Protection Clause). Plaintiffs’ severability argument relies almost exclusively on Michigan caselaw pertaining to statutory severability, but Michigan’s Supreme Court has declined to apply that caselaw to questions of constitutional severability. See In re Apportionment of State Legislature--1982, 321 N.W.2d at 581 (“This Court will not apply case law developed in the resolution of controversies concerning statutory invalidity where the issue presented concerns constitutional invalidity.”). No. 20-1734 Daunt et al. v. Benson et al. Page 28

____________________________________

CONCURRING IN THE JUDGMENT ____________________________________

CHAD A. READLER, Circuit Judge, concurring in the judgment. In last year’s decision affirming the denial of a preliminary injunction, the panel refused to “choose between the two” constitutional doctrines it believed might apply to this challenge to Michigan’s Independent Citizens Redistricting Commission. See Daunt v. Benson (Daunt I), 956 F.3d 396, 406 (6th Cir. 2020) (citation omitted). One was the framework from Anderson-Burdick. As that framework had never been applied to evaluating eligibility requirements for a state office (as opposed to actual ballot-access matters), however, I concurred only in the judgment. Id. at 422–26 (Readler, J., concurring in the judgment). Today’s appeal comes to us in a different posture. But otherwise, little has changed in the last year: the panel remains the same, the parties (and their arguments) are largely identical, and the majority opinion’s approach echoes its earlier conclusion. So too then for my decision again to concur only in the judgment. For application of the Anderson-Burdick framework has nothing more to recommend it today.

The majority opinion in Daunt I entertained the possibility that Anderson-Burdick applied in this setting, but ultimately refused to squarely hold as much. Id. at 406 (majority opinion). Further developments understandably have not lessened that ambivalence. In this second act, the intervenors only faintly praise Anderson-Burdick’s balancing test as a “conceivable framework of analysis” here. See Intervenor’s Br. at 20. And the Secretary of State, when pressed on the issue during oral argument, was decidedly tepid, observing that the Daunt I concurrence was “well-taken,” noting that the State has argued that there are limits to Anderson-Burdick, and then adding only that the eligibility requirements were not “so attenuated from the administration of elections” as to make use of the balancing test inappropriate here. Oral Arg. at 18:17–20:22. It is thus no surprise that the majority opinion once again merely surmises, as it did in Daunt I, that Anderson-Burdick “could” apply here, yet rejects a wholesale adoption of the test as “overreach.” Maj. Op. at 17. That reticence is understandable. The Anderson-Burdick framework, after all, centrally “governs First and Fourteenth Amendment challenges to ballot- access restrictions.” See Kishore v. Whitmer, 972 F.3d 745, 749 (6th Cir. 2020). We have No. 20-1734 Daunt et al. v. Benson et al. Page 29

utilized that framework only in ballot-access cases or, at its most generous application, voting- access matters, such as voter-identification laws or early-voting regulations. Id.; see also Appellants’ Br. at 23 (explaining that “courts have never applied the Anderson-Burdick test in cases directly involving challenges to redistricting plans or redistricting commissions”). Yet those election features have nothing in common with this dispute, where, as the majority opinion acknowledges, “positions on the Commission are not elected,” and the “eligibility criteria do not raise ballot-access concerns.” Maj. Op. at 25.

While the majority opinion’s ambivalence towards applying Anderson-Burdick here coupled with its holding on the unconstitutional conditions doctrine renders the following “dicta about dicta,” United States v. Mathews, 846 F. App’x 362, 364 n.3 (6th Cir. 2021), I write separately to amplify the views I expressed in Daunt I. There, I explained that Anderson- Burdick offers a dubious framework for resolving even cases about elections, let alone cases (like this one) that have nothing to do with an election. Were there any doubt on that score, the majority opinion’s treatment of Anderson-Burdick here confirms the rampant subjectivity inherent in applying that framework. Equally troubling is the fact that, over time, we have applied Anderson-Burdick more frequently—and often less deferentially to a state’s political choices—than any of our sister circuits. Going forward, we should strictly confine (if not abandon altogether) our reliance on Anderson Burdick, to curb the judicial overreach we currently invite.

1. Anderson-Burdick’s hallmark is standardless standards. See Daunt I, 956 F.3d at 424 (Readler, J., concurring in the judgment) (“Anderson-Burdick is a dangerous tool. In sensitive policy-oriented cases, it affords far too much discretion to judges in resolving the dispute before them.”). The Anderson-Burdick framework requires a court to undertake the metaphysical task of weighing a state’s interests in maintaining its election laws against the burden those laws impose on a plaintiff’s rights, such as the right to vote. Burdick v. Takushi, 504 U.S. 428, 434 (1992) (quoting Anderson v. Celebrezze, 460 U.S. 780, 789 (1983)). As I explained in Daunt I, that framework seemingly is little more than a “grand ‘balancing test in which unweighted factors mysteriously are weighed.’” June Med. Servs. LLC v. Russo, 140 S. Ct. 2103, 2135 No. 20-1734 Daunt et al. v. Benson et al. Page 30

(2020) (Roberts, C.J., concurring in the judgment) (quoting Marrs v. Motorola, Inc., 577 F.3d 783, 788 (7th Cir. 2009)).

Any lingering doubts on that front are put to rest by today’s majority opinion. On one side of the scale, the majority opinion weighs Michigan’s purported “compelling interest[s]” in adopting the Commission’s eligibility requirements, which the majority opinion variously describes as interests in “cleansing its redistricting process of partisan influence,” “limiting the conflict of interest implicit in legislative control over redistricting,” and “avoiding the appearance of conflicts of interest in its redistricting.” Maj. Op. at 12, 14. As an initial matter, whether attempting to wholly eliminate partisanship from redistricting is possible, let alone the true motivation behind the Commission, is debatable. Cf. Ariz. State Legislature v. Ariz. Indep. Redistricting Comm’n, 576 U.S. 787, 848–49 (2015) (Roberts, C.J., dissenting) (dismissing the supposedly “high motives” behind Arizona’s redistricting commission and recounting how partisanship infused that “unelected, unaccountable institution” to the disadvantage of one party). Yet even setting those realities aside, the fact remains that a compelling interest is one that attacks “[o]nly the gravest abuses, endangering paramount interest[s].” See Sherbert v. Verner, 374 U.S. 398, 406 (1963) (first alteration in original) (citation omitted). That hardly describes eliminating politics from the redistricting process, when redistricting has been a “familiar” partisan exercise since the early days of our republic. See Rucho v. Common Cause, 139 S. Ct. 2484, 2494–95 (2019). Tellingly, the majority opinion musters not a single citation to a First Amendment or Equal Protection decision (where compelling interests are routinely ascertained) to support its summary declaration that Michigan’s interest here is “compelling.” Instead, the opinion’s lone citation is to an Elections Clause case quoting a political scientist’s assessment of the value of redistricting commissions. See Maj. Op. at 14 (quoting Ariz. State Legislature, 576 U.S. at 821 (quoting Bruce Cain, Redistricting Commissions: A Better Political Buffer?, 121 Yale L.J. 1808, 1808 (2012))).

The majority opinion does no better elsewhere in its assessment of Michigan’s interests in the Commission’s eligibility requirements. In some places, the majority opinion utilizes generalized considerations about elections to bolster the purported importance of credentialing for redistricting commissions, something many steps removed from an election. For instance, No. 20-1734 Daunt et al. v. Benson et al. Page 31

the majority opinion values the considerations at play when choosing who can serve as a redistricting commissioner as “no different” than the interest in stopping “voter fraud.” See Maj. Op. at 14. Really? The latter, when successful, results in illegal ballots being counted in an election, potentially tainting the outcome. The former, by comparison, has nothing to do with ballots or a ballot box. In other places, the majority opinion simply guesses as to the exact rationale that drove the challenged eligibility requirements. See id. at 4 n.1 (hypothesizing that while Michigan did not offer a “clear explanation” for its “six-year look-back” rule, the rule may pertain to the length of a U.S. senator’s term). To be sure, reasonable minds can debate (in the proper forum) how “compelling” the arguments are for establishing an independent redistricting commission like Michigan’s, a debate that likely informed the electorate’s consideration of the underlying ballot initiative. But resting the Commission’s constitutionality on such a subjective and unconstrained inquiry has an anything-goes quality to it, a feature rarely seen when judicial decisionmaking is constrained.

On the other side of the scale is the majority opinion’s take on the burdens the eligibility requirements impose on plaintiffs. Of course, those eligibility requirements, it bears repeating, do not burden any voter with respect to any election, a fundamental reason why Anderson- Burdick is so misplaced in this setting. Nonetheless, the majority opinion soldiers on to evaluate what my two colleagues here have previously described as the “‘character and magnitude of the asserted injury’ to plaintiffs’ constitutional rights.” Graveline v. Benson, 992 F.3d 524, 534 (6th Cir. 2021) (quoting Anderson, 460 U.S. at 789). Needless to say, “character and magnitude” are qualities entirely left to the eye of the beholder. Today’s beholders conclude that the eligibility criteria impose “at most, a moderate burden” on plaintiffs’ First and Fourteenth Amendment rights. Maj. Op. at 13. Yet in making that assessment, the majority opinion cites only cases that either apply Anderson-Burdick in a classic election context or decline to apply that framework at all. And none of those cases, suffice it to say, would fairly be described as apt comparisons. The majority opinion first equates the burden imposed by a two-year restriction on a political candidate from serving in a particular office, see Clements v. Fashing, 457 U.S. 957, 967 (1982) (plurality opinion), with the burden imposed by a restriction three times as long on a nonpolitical officeholder. See Maj. Op. at 14. It then embraces a summary affirmance upholding a law imposing a seven-year residency requirement on gubernatorial candidates—even where the No. 20-1734 Daunt et al. v. Benson et al. Page 32

decision predates Anderson and concerns a law that merely prevented carpetbagging and did not impose any condition prohibiting the exercise of core political speech. Id. at 14 (citing Chimento v. Stark, 414 U.S. 802 (1973)).

That my colleagues would so quickly accept the State’s purported interests in the redistricting law yet so quickly reject the law’s proposed burdens comes as quite a surprise. Our authoring judge, for one, customarily requires much more from a state. See, e.g., Ohio State Conf. of NAACP v. Husted, 768 F.3d 524, 545 (6th Cir.) (explaining that a state must do more than assert abstract interests under Anderson-Burdick, and instead must “explain why the particular restriction imposed is actually necessary, meaning it actually addresses, the interest put forth”), vacated, No. 14-3877, 2014 WL 10384647 (6th Cir. Oct. 1, 2014); see also Mich. State A. Philip Randolph Inst. v. Johnson, 833 F.3d 656, 665–66 (6th Cir. 2016) (rejecting as “vague speculation” that a state, by eliminating straight-party voting, will foster an engaged electorate). And both of my colleagues, in their recent Graveline decision, found another law’s purported burdens in the election context “severe.” 992 F.3d at 539. There, my colleagues (over a dissent by Judge Griffin) held that a law requiring a statewide candidate in Michigan to collect 30,000 signatures (or 0.72% of the votes cast in last relevant election) to appear on the general election ballot resulted in a “severe” burden on the plaintiffs’ First and Fourteenth Amendment rights. See id. at 543. But see id. at 548 (Griffin, J., dissenting) (finding this “miniscule,” temporally limited requirement to not impose a severe burden under Anderson-Burdick). If so, how does a six-year prohibition on a host of core political activities for potential members of Michigan’s Commission impose at most only a modest burden under the same rubric? Like other aspects of its Anderson-Burdick analysis, the majority opinion leaves this tension unexplained.

Anderson-Burdick’s rampant subjectivity does not end there. For once jurists have cooked up these competing interests, they weigh them, without any safeguards or benchmarks to channel the process. See Bendix Autolite Corp. v. Midwesco Enters., Inc., 486 U.S. 888, 897 (1988) (Scalia, J., concurring in the judgment) (analogizing to the hopeless task of assessing “whether a particular line is longer than a particular rock is heavy”). Nor, it seems, without much illumination, if the majority opinion here is any guide. With all of this in mind, it is quite fanciful to believe that Anderson-Burdick is a “carefully calibrated” test “uniquely well suited” No. 20-1734 Daunt et al. v. Benson et al. Page 33

to this context. Maj. Op. at 18. After all, those purported calibrations are almost entirely reliant on the predilections of the jurist who undertakes them. Daunt I, 956 F.3d at 424–25 (Readler, J., concurring in the judgment). Like many a decision in our circuit purporting to apply Anderson- Burdick, the majority opinion’s application of that framework seemingly turns more on judicial whims than bright-line standards. See Am. Jewish Congress 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.”).

2. My colleagues nonetheless have faith that Anderson-Burdick can fairly apply to a “wide array of claims” that merely “touch[] on the election process,” believing that the framework is “uniquely well suited to this sort of case” because the law at issue is geared toward maintaining the “integrity of the democratic system” and promoting “democratic principles.” See Maj. Op. 17–18. But consider the dramatic consequences that would ensue were one to follow that line of thinking. Extending the Anderson-Burdick framework to any law that ostensibly touches on “democratic principles” or the “democratic system” would license unrestrained judicial scrutiny of an endless list of public laws—“campaign finance, the conduct of legislators, or the terms of service of elected judges,” just to name a few. See Daunt I, 956 F.3d at 423 (Readler, J., concurring in the judgment). That slope is not just slippery; it is greased, frozen, and polished, and likely why neither the Supreme Court nor this Court has ever applied Anderson-Burdick to those types of disputes. See, e.g., Moncier v. Haslam, 570 F. App’x 553, 559 (6th Cir. 2014) (holding that a plaintiff lacked a “federally protected interest” under Anderson-Burdick to challenge laws concerning how “states organize their governments in a particular manner”); see also Ohio A. Philip Randolph Inst. v. Householder, 373 F. Supp. 3d 978, 1153 (S.D. Ohio) (per curiam) (opining that “the Anderson-Burdick standard [is] well suited to address . . . the partisan-gerrymandering context”), vacated and remanded, 140 S. Ct. 101 (2019) (vacating the judgment in light of Rucho v. Common Cause, 139 S. Ct. 2484 (2019)).

All of this begs the question: Why would a judge choose to proliferate the use of such a standardless standard and then extend that approach into a frontier several steps removed from a ballot box? The only conceivable answer, it seems, is to enhance judicial power. As deployed No. 20-1734 Daunt et al. v. Benson et al. Page 34

by the majority opinion and our Court on other occasions, Anderson-Burdick affords judges a veneer of neutral calculus to disguise the immense power first to pick several variables, and then to pick how to weigh them, with each step constrained only by a judge’s self-perceived wisdom about the merits of a state’s political choices. See Luft v. Evers, 963 F.3d 665, 671 (7th Cir. 2020) (observing that interpreting Anderson-Burdick to allow judges to decide whether a “given election law is necessary” or imposes an “excessive burden” “allows a political question— whether a rule is beneficial, on balance—to be treated as a constitutional question and resolved by the courts rather than by legislators”). And what better tool than this version of Anderson- Burdick for unelected jurists who routinely know better than a state itself how to run state elections and state government? See, e.g., Graveline, 992 F.3d at 528 (striking down “a constellation of Michigan laws . . . govern[ing] an independent candidate’s ability to be placed on the ballot for statewide office”); Mich. State A. Philip Randolph Inst., 833 F.3d at 659 (denying a stay of a ruling prohibiting the enforcement of a law that eliminates straight-party voting); Ohio State Conf. of NAACP, 768 F.3d at 529 (affirming a ruling enjoining enforcement of Ohio laws regulating the hours for voting).

In a different setting, the use of such malleable judicial standards might not raise an eyebrow. But the cases to which my colleagues increasingly apply Anderson-Burdick are no small matters. Rather, they concern elections and, more broadly, the “democratic process.” How ironic (if not dispiriting) that the permissibility of those democratic choices in essence turns on how three (here two) unelected judges pick and choose the ways in which a state may structure its government or elections. At a time when some have expressed skepticism over the judiciary’s historical nonpartisan nature, one might think judges would do all they could to remove themselves from unduly interfering with a state’s political choices. Cf. Tennessee v. FCC, 832 F.3d 597, 611 (6th Cir. 2016) (recognizing that “the decision-making structure of a state” is at the “core” of a state’s sovereignty). But our approach toward Anderson-Burdick does just the opposite—it puts a judge’s inherent policy preferences front-and-center when deciding critical matters of public and political interest. In nudging our Court even further in that direction, the majority opinion regrettably aggrandizes power for the courts and the courts alone. Cf. Graveline, 992 F.3d at 553 (Griffin, J., dissenting) (explaining “why applying Anderson-Burdick’s grant of discretion to the federal judiciary can lead to tension with the No. 20-1734 Daunt et al. v. Benson et al. Page 35

principles of federalism and separation of powers”); Stewart v. Blackwell, 444 F.3d 843, 881 (6th Cir.) (Gilman, J., dissenting) (lamenting the majority opinion for “subjecting an indeterminate number of state and local election decisions to the strictest level of constitutional scrutiny” and arguing for a “more cautious approach . . . that recognizes the primacy of the executive and legislative branches in the elector process”), vacated (July 21, 2006), superseded, 473 F.3d 692 (6th Cir. 2007) (en banc). All things considered, doggedly employing Anderson-Burdick—especially in cases where it plainly has no application—strikes me as a camouflaged interest in asserting “will” rather than “judgment” beyond the ken of Article III. See The Federalist No. 78, at 465 (Alexander Hamilton) (J. Cooke ed., 1961).

3. The majority opinion’s zeal for reaching today’s outcome becomes even more evident by its willingness to dispatch plaintiffs’ claims at the Rule 12(b)(6) stage. If nothing else, our approach toward Anderson-Burdick, as our lead judge has previously explained, necessitates the development of an evidentiary record, doubly so it seems when the test is applied to a novel law. See Ohio State Conf. of NAACP, 768 F.3d at 545 (noting the need for the state to “present[] actual evidence to support the justifications it provided for the challenged law”) (citation omitted); Ohio A. Philip Randolph Inst. v. Householder, 367 F. Supp. 3d 697, 710 (S.D. Ohio 2019) (per curiam) (noting that only “[a]fter hearing and weighing the evidence at trial” would the court “consider how substantial the burden” of the challenged redistricting law was on “the Plaintiffs’ right to vote”); see also Green Party of Tenn. v. Hargett, 767 F.3d 533, 548–49 (6th Cir. 2014) (reviewing the application of Anderson-Burdick, remanding “for further factual development,” and advising “the district court to re-open the record on remand so that the parties may submit evidence”); see generally Soltysik v. Padilla, 910 F.3d 438, 448–50 (9th Cir. 2018) (addressing the distinctions between evaluating Anderson-Burdick challenges at the motion to dismiss stage versus summary judgment stage). This is unsurprising, given the inherently fact- intensive nature of Anderson-Burdick. See Crawford v. Marion Cnty. Election Bd., 553 U.S. 181, 191 (2008) (plurality opinion) (recognizing that there is no “litmus test for measuring the severity of a burden that a state law imposes on . . . an individual voter”); Gill v. Scholz, 962 F.3d 360, 365 (7th Cir. 2020) (noting the “fact-intensive analysis required” by Anderson-Burdick and the need for the district court to examine “factual differences” in the application of election laws under Anderson-Burdick). To be sure, we have sometimes affirmed the dismissal of No. 20-1734 Daunt et al. v. Benson et al. Page 36

Anderson-Burdick claims early in litigation. But we did so where, unlike here, a plaintiff challenged a relatively simple election restriction, with few if any facts in dispute. See Comm. to Impose Term Limits on Ohio Sup. Ct. v. Ohio Ballot Bd., 885 F.3d 443, 448 (6th Cir. 2018) (evaluating a challenge to Ohio’s single-subject rule for statewide initiatives); Lawrence v. Blackwell, 430 F.3d 368, 369 (6th Cir. 2005) (affirming a motion to dismiss a challenge to an Ohio law requiring congressional candidates to obtain a minimum number of signatures before the primary election); Taxpayers United for Assessment Cuts v. Austin, 994 F.2d 291, 297 (6th Cir. 1993) (affirming dismissal of a challenge to a Michigan law requiring a minimum number of qualified signatures to trigger an initiative).

Plaintiffs’ complaint alleges that the eight eligibility requirements at issue, each with its own complexities, impose “substantial[] burdens” on plaintiffs’ exercise of their constitutional rights. See Compl. ¶ 59. Refusing to accept plaintiffs’ allegations as true, the majority opinion instead summarily concludes that the eligibility requirements could never impose such a burden on these plaintiffs, either facially or as applied. Assuming Anderson-Burdick governs, under the loose contours we employ, discovery would likely help discern the extent of the burden imposed by the eligibility requirements. See Ne. Ohio Coal. for Homeless v. Husted, 696 F.3d 580, 593 (6th Cir. 2012) (evaluating the burden aspect of Anderson-Burdick by looking at the lower court’s factual findings); Ohio State Conf. of NAACP, 768 F.3d at 544 (observing the need to provide evidence to satisfy the burden prong of Anderson-Burdick and citing how “extensive statistical, survey, and anecdotal evidence” can alter a court’s calculus in the review of similar election laws). It could also help elucidate whether any (or all) of the eight conflict-of-interest rules actually serves their intended purpose. Or fact-finding could test the majority opinion’s assumption about whether the burden imposed by a six-year ban on political activity is truly equivalent to a two-year ban. All to say, if we are going to second guess the State’s choices and decide for ourselves the efficacy of the Commission’s eligibility requirements, we ought, at least under the contours of the majority opinion’s “flexible analysis” employ that flexibility to understand fully the import of those requirements.

But the majority opinion takes a different path. Having made up its mind as to the wisdom of the redistricting amendment, the majority opinion declares that discovery is No. 20-1734 Daunt et al. v. Benson et al. Page 37

unnecessary when the “burden” a law imposes is “plain” according to its “fixed terms.” See Maj. Op. at 16. The opinion then speculates that discovery could not possibly inform our Anderson- Burdick analysis, going so far as to chastise plaintiffs’ lack of clairvoyance as to what fruits discovery might bear. Perhaps, as the majority opinion protests, it need not “conjure up facts” to support plaintiffs’ view of the burdens imposed by the eligibility criteria. See id. at 11. But why then is the majority opinion more than willing to conjure up the State’s reasons for enacting its redistricting amendment? See, e.g., id. at 4 n.1 (conceding that the State has not offered a clear explanation for its look-back provision). How this approach squares with the ordinary niceties we afford plaintiffs at the 12(b)(6) stage is left unsaid. See, e.g., Taylor v. City of Saginaw, 922 F.3d 328, 331 (6th Cir. 2019) (“On a motion to dismiss, we must construe the complaint in the light most favorable to the plaintiff and accept all allegations as true. The defendant has the burden of showing that the plaintiff has failed to state a plausible claim for relief.” (cleaned up)). Nor does it make sense effectively to conclude that plaintiffs conceded (or are somehow estopped from contesting) that dismissal is appropriate in view of plaintiffs’ strategic choice not to contest the facts at the preliminary injunction stage, notwithstanding the very different standards governing each motion. Cf. Teledyne Indus., Inc. v. NLRB, 911 F.2d 1214, 1218 n.3 (6th Cir. 1990) (observing that parties ordinarily can argue in the alternative during the course of litigation).

I remain deeply skeptical that Anderson-Burdick is the appropriate framework to apply in this context. But if we are going to apply that framework to a novel and complex law, the motion to dismiss stage is not the end point. It is one thing to choose to employ Anderson- Burdick at all. But it is quite another to deploy that malleable, fact-intensive framework through a superficial glance at the complaint. Just one more corner, it seems, that the majority opinion is willing to cut in its half-hearted embrace of Anderson-Burdick.

4. Today’s majority opinion is in keeping with a broader trend from this Court, one in which we continue to “double down” to aggressively employ Anderson-Burdick to second guess a state’s political choices. See Graveline, 992 F.3d at 548 (Griffin, J., dissenting). By doing so, we stand alone when compared to our sister circuits. Those circuits wisely have resisted deploying Anderson-Burdick in new frontiers. See, e.g., Jacobson v. Fla. Sec’y of State, No. 20-1734 Daunt et al. v. Benson et al. Page 38

974 F.3d 1236, 1260–61 (11th Cir. 2020) (holding that Anderson-Burdick could not be applied with respect to a challenge to Florida’s ballot order statute for general elections); Molinari v. Bloomberg, 564 F.3d 587, 604–05 (2d Cir. 2009) (concluding that Anderson-Burdick was “completely inapposite” with respect to a challenge to New York City’s term-limits law, which did not involve “direct restrictions on speech or access to the ballot”); cf. Tex. League of United Latin Am. Citizens v. Hughs, 978 F.3d 136, 144 n.6 (5th Cir. 2020) (applying Anderson-Burdick but expressing skepticism that the test applies to a law that limits the “claimed right to receive absentee ballots,” as opposed to the right to vote); accord Schmitt v. LaRose, 933 F.3d 628, 644 (6th Cir. 2019) (Bush, J., concurring in part and in the judgment) (explaining that Anderson- Burdick does not address whether “the First Amendment [is] impinged upon by statutes regulating the election mechanics concerning initiative petitions”). And when these circuits do apply Anderson-Burdick’s balancing test, they tend to start from a strong presumption that judicial invalidations of state election laws should be a narrow exception, not the rule. See, e.g., Libertarian Party of Va. v. Alcorn, 826 F.3d 708, 714 (4th Cir. 2016) (“We begin with the uncontroversial proposition that the legislature in each state of our federal system possesses the presumptive authority to regulate elections within that state’s sovereign territory.”); Tully v. Okeson, 977 F.3d 608, 611 (7th Cir. 2020); New Ga. Project v. Raffensperger, 976 F.3d 1278, 1280 (11th Cir. 2020). With that presumption in mind, other circuits customarily view Anderson-Burdick as a fairly deferential test, see, e.g., SAM Party of N.Y. v. Kosinski, 987 F.3d 267, 274 (2d Cir. 2021); Common Cause Ind. v. Lawson, 978 F.3d 1036, 1040 (7th Cir. 2020) (per curiam), and regularly uphold laws that do not severely burden “First and Fourteenth Amendment rights of voters” so long as the state has an important regulatory interest, SAM Party of N.Y., 987 F.3d at 274 (quoting Burdick, 504 U.S. at 434); see also Marcellus v. Va. State Bd. of Elections, 849 F.3d 169, 175 (4th Cir. 2017); Stone v. Bd. of Election Comm’rs for City of Chicago, 750 F.3d 678, 681 (7th Cir. 2014). When considering Anderson-Burdick’s state- interests prong, other circuits do not require “elaborate, empirical” evidence from a state to justify its asserted regulatory interests. See, e.g., Miller v. Thurston, 967 F.3d 727, 740 (8th Cir. 2020) (quoting Timmons v. Twin Cities Area New Party, 520 U.S. 351, 364 (1997)). Likewise, before finding that a state law imposes a severe burden, those courts require a plaintiff to clear the high bar of showing that a state’s restrictions “go beyond the merely inconvenient.” See id. No. 20-1734 Daunt et al. v. Benson et al. Page 39

at 739 (citation omitted); see also Tex. Democratic Party v. Abbott, 961 F.3d 389, 405 (5th Cir. 2020); Utah Republican Party v. Cox, 892 F.3d 1066, 1077 (10th Cir. 2018). And our sister courts have limited Anderson-Burdick’s reach in other ways, for example, by refusing to invalidate generally applicable election laws that may disparately burden a “small number of voters,” see Richardson v. Tex. Sec’y of State, 978 F.3d 220, 236 (5th Cir. 2020), or by applying the framework only to a state’s election laws as a whole, rather than in isolation, see, e.g., Luft, 963 F.3d at 675 (“In isolation, any rule reducing the number of days or hours available for any kind of voting seems like an unjustified burden. But electoral provisions cannot be assessed in isolation.”); LaRouche v. Fowler, 152 F.3d 974, 994 (D.C. Cir. 1998). Collectively, this approach to Anderson-Burdick better respects a state’s sovereign interest in regulating elections.

We, on the other hand, have all too often failed to show similar restraint. Unlike our sister circuits, we interject Anderson-Burdick into contexts unrelated to the ballot-access or voter-access issues. See Daunt I, 956 F.3d at 423 (Readler, J., concurring in the judgment). And seemingly no matter the context, we voice no presumption of deference to a state’s democratic choices. See Libertarian Party of Ohio v. Blackwell, 462 F.3d 579, 605 (6th Cir. 2006) (Griffin, J., dissenting) (noting the majority opinion’s failure to highlight the constitutional presumption against invalidating a state election law). Standing alone from our sister circuits, we deem “most” Anderson-Burdick cases to “fall in between” the extremes of laws that impose severe burdens and no burdens at all, thereby subjecting a wide swath of state laws to the supposed “hard judgment[s]” that the whims of Anderson-Burdick’s “flexible standard” “demand[].” See Obama for Am. v. Husted, 697 F.3d 423, 429 (6th Cir. 2012) (citation omitted); see also Mays v. LaRose, 951 F.3d 775, 783 n.4 (6th Cir. 2020) (observing that our circuit invented a “three-track approach” under Anderson-Burdick, through which we largely dispense with the Supreme Court’s deferential balancing approach); Stewart, 444 F.3d at 884–85 (Gilman, J., dissenting) (noting that “our sister circuits” evaluate “many state and local voting regulations” under a rational basis test). We then deploy (as the majority opinion does today) an unchecked balancing test, where results largely turn on the idiosyncrasies of a given panel. In applying that balancing test, we (unlike other circuits) all too often require a state to present extensive evidence to justify its proffered interests. See, e.g., Obama for Am., 697 F.3d at 434 (deeming the state’s interest in the smooth functioning of local boards of election to be “vague” and requiring specific evidence No. 20-1734 Daunt et al. v. Benson et al. Page 40

to support that interest); Ohio State Conf. of NAACP, 768 F.3d at 545 (maintaining that “abstract state interests” are insufficient under Anderson-Burdick). Our version of Anderson-Burdick, moreover, regularly allows for the invalidation of generally applicable state laws that, at most, disparately burden pockets of voters. See, e.g., Ohio State Conf. of NAACP, 768 F.3d at 543; Mich. State A. Philip Randolph Inst., 833 F.3d at 666; Ne. Ohio Coal. for Homeless, 696 F.3d at 592. And unlike our sister courts, we have, at times, applied Anderson-Burdick to evaluate the effects of discrete voting rules, as opposed to addressing a state’s broader legal framework for voting, which, when read together, may alleviate the burdens of a particular rule. See, e.g., Obama for Am., 697 F.3d at 431.

All told, our aggressive deployment of Anderson-Burdick as a “one-size-fits-all” tool for resolving any matter touching the “democratic process” has us in the record books. By my count, over the last two decades, no circuit has issued more opinions citing to both Anderson and Burdick than ours. But that is no record to brag about. The vagaries of the Anderson-Burdick balancing test coupled with the unusual ways in which we have warped that framework have produced results that are at best inconsistent, and at worst border on the absurd. See, e.g., Stein v. Thomas, 672 F. App’x 565, 570 (6th Cir. 2016) (per curiam) (order) (concluding that a candidate who received roughly one percent of the votes in a statewide election suffered a severe burden from having to wait two business days for a recount to commence). Regrettably, our cases license routine and unwarranted second-guessing of a state’s democratic decisions, just the sort of thing unelected federal judges are least well-equipped to do. See Crawford, 553 U.S. at 208 (Scalia, J., concurring in the judgment) (rejecting “detailed judicial supervision of the election process” because “[i]t is for state legislatures to weigh the costs and benefits of possible changes to their election codes”). One can only hope this trend reverses course before all notions of judicial modesty are lost.

5. Given my strong misgivings about applying Anderson-Burdick here, I return to my concurring opinion in Daunt I, which explained that history, tradition, and precedent should instead be the guide for evaluating the eligibility requirements for service on the Commission. 956 F.3d at 422 (Readler, J., concurring in the judgment). In line with long-held historical practice, the Supreme Court has deferentially reviewed state and federal conflict-of-interest laws No. 20-1734 Daunt et al. v. Benson et al. Page 41

as well as laws limiting the eligibility for service in a public office. See id. at 428–29 (citing Nev. Comm’n on Ethics v. Carrigan, 564 U.S. 117 (2011) (conflict-of-interest recusal rules), and Clements, 457 U.S. at 957 (state limitations on public employment)). Given the important federalism principles at play when a state, acting in its sovereign capacity, chooses to structure its public offices in a way that does not invidiously discriminate on a basis plainly prohibited by the Constitution, see Citizens for Legis. Choice v. Miller, 144 F.3d 916, 925 (6th Cir. 1998), something akin to deferential rational basis review should govern. See Daunt I, 956 F.3d at 429 (Readler, J., concurring in the judgment) (citing Williamson v. Lee Optical of Okla., Inc., 348 U.S. 483, 489 (1955)); see also Bates v. Jones, 131 F.3d 843, 859 (9th Cir. 1997) (en banc) (Rymer & O’Scannlain, JJ., concurring in the result) (rejecting the application of Anderson-Burdick in favor of a rule that a state qualification for office is constitutional “unless the qualification is plainly prohibited by some other provision in the Constitution”); cf. Schmitt, 933 F.3d at 644 (Bush, J., concurring in part and in the judgment) (arguing that “non- discriminatory referendum regulations are, at most, subject to rational-basis review”).

Viewed in this light, the eligibility requirements to serve on the Commission pass constitutional muster, even at the motion to dismiss stage. Under rational basis review, a state law comports with the federal Constitution if “any state of facts reasonably can be conceived that would sustain it.” Allied Stores of Ohio, Inc. v. Bowers, 358 U.S. 522, 528 (1959). In this setting, to survive a motion to dismiss for failure to state a claim, a plaintiff must allege “facts sufficient to overcome the presumption of rationality that applies to government classifications,” including facts that “negat[e] every conceivable basis which might support” the law. Bower v. Vill. of Mount Sterling, 44 F. App’x 670, 677 (6th Cir. 2002) (emphasis and citations omitted); see also Gilmore v. City of Douglas, 406 F.3d 935, 937 (8th Cir. 2005) (“[B]ecause all that must be shown is any reasonably conceivable state of facts that could provide a rational basis for the classification, it is not necessary to wait for further factual development in order to conduct a rational basis review on a motion to dismiss.” (cleaned up)).

According to the complaint, the redistricting amendment was spearheaded by a group seeking to displace Michigan’s longstanding process of drawing legislative district boundaries through the approval of the state legislature and Governor. See Compl. ¶¶ 25–26. Possibly No. 20-1734 Daunt et al. v. Benson et al. Page 42

inspired by Justice Scalia’s musings over judgments made by “nine people picked at random from the Kansas City telephone directory,” see Cruzan v. Dir., Mo. Dep’t of Health, 497 U.S. 261, 293 (1990) (Scalia, J., concurring), voters in Michigan approved a state constitutional amendment that allows for the selection of redistricting commissioners in somewhat random fashion, with restrictions on service imposed based upon party affiliation and relational and professional characteristics perceived to create the risk of a conflict of interest. See Mich. Const. art. IV, § 6. While a more robust factual record would be relevant for evaluating how the challenged constitutional amendment fares under Anderson-Burdick, given that the complaint does not allege why it would be entirely irrational for Michigan to set the qualifications for Commission membership in the manner that it did, dismissal of the complaint was appropriate. After all, when a state chooses to create a redistricting commission with a purportedly nonpartisan mission, it can likewise choose, without offending the First Amendment, to have the commission run by those whose present or recent partisan experiences arguably do not cloud the government’s message. See Garcetti v. Ceballos, 547 U.S. 410, 421–22 (2006) (“Restricting speech that owes its existence to a public employee’s professional responsibilities does not infringe any liberties the employee might have enjoyed as a private citizen. It simply reflects the exercise of employer control over what the employer itself has commissioned or created.”); see also Branti v. Finkel, 445 U.S. 507, 518 (1980) (holding that the First Amendment permits consideration of a person’s partisan affiliation when making personnel decisions for high-level policymaking roles).

Plaintiffs seemingly accept the case law that instructs us to review deferentially state laws governing conflicts of interest (including those imputed to family members) and political participation by government employees (including past political activity). They nevertheless argue that the combination of two aspects of the eligibility requirements—neither of which standing alone would be unconstitutional—together exceed constitutional limits. To that end, plaintiffs emphasize that the eligibility requirements should be measured under a heightened standard of review in that they extend the Commission’s conflict-of-interest rules (1) retrospectively (2) to family members of covered parties. See Appellants’ Reply Br. at 13. Yet plaintiffs cite no case law for the proposition that two constitutional rights nonetheless can make a wrong. Instead, the level of scrutiny by which we judge a given law depends on the No. 20-1734 Daunt et al. v. Benson et al. Page 43

broad nature of the government classification in question. City of Cleburne v. Cleburne Living Ctr., 473 U.S. 432, 440 (1985). With the Commission’s eligibility requirements reflecting Michigan’s determination as to self-government, and as those requirements do not invidiously discriminate or otherwise implicate core constitutional interests warranting heightened scrutiny, a deferential basis for review governs those requirements, no matter their particular contours. See Citizens for Legis. Choice, 144 F.3d at 925. As a result, I continue to adhere to the view that a “state’s prerogative in organizing its government, including its election system,” counsels for deferentially reviewing Michigan’s redistricting amendment. Daunt I, 956 F.3d at 431 (Readler, J., concurring in the judgment). On that basis, I would affirm the district court’s judgment.