Nancy Carola Jacobsen v. Florida Secretary of State

957 F.3d 1193

Opinion

[PUBLISH]

IN THE UNITED STATES COURT OF APPEALS

FOR THE ELEVENTH CIRCUIT

________________________

No. 19-14552

________________________

D.C. Docket No. 4:18-cv-00262-MW-CAS

NANCY CAROLA JACOBSON, TERENCE FLEMING, et al.,

Plaintiffs-Appellees,

versus

FLORIDA SECRETARY OF STATE, NATIONAL REPUBLICAN SENATORIAL COMMITTEE, et al.,

Defendants-Appellants.

________________________

Appeal from the United States District Court for the Northern District of Florida

________________________

(April 29, 2020)

Before WILLIAM PRYOR, JILL PRYOR, and LUCK, Circuit Judges. WILLIAM PRYOR, Circuit Judge:

This appeal requires us to decide whether several voters and organizations have standing to challenge a law that governs the order in which candidates appear on the ballot in Florida’s general elections. The law provides that candidates of the party that won the last gubernatorial election shall appear first for each office on the ballot and that candidates of the second-place party shall appear second. Several Democratic voters and organizations sued the Florida Secretary of State to enjoin enforcement of the law. They alleged that the law violates their rights under the First and Fourteenth Amendments because candidates who appear first on the ballot—in recent years, Republicans—enjoy a “windfall vote” from a small number of voters who select the first candidate on a ballot solely because of that candidate’s position of primacy. After a bench trial, the district court permanently enjoined the Secretary—and the 67 county Supervisors of Elections, none of whom were made parties to this lawsuit—from preparing ballots in accordance with the law.

We hold that the voters and organizations lack standing to sue the Secretary.

None of them proved an injury in fact. And any injury they might suffer is neither fairly traceable to the Secretary nor redressable by a judgment against her because she does not enforce the challenged law. Instead, the Supervisors—county officials independent of the Secretary—are responsible for placing candidates on the ballot in the order the law prescribes. The district court lacked authority to enjoin those

3

officials in this suit, so it was powerless to provide redress. Because the voters and organizations lack standing, we vacate and remand with instructions to dismiss for lack of justiciability.

I. BACKGROUND

As part of a comprehensive revision to the election code, the Florida Legislature enacted a statute in 1951 that governs the order in which candidates appear on the ballot in general elections. 1951 Fla. Laws 871 (originally codified at Fla. Stat. § 101.151(4) (1951)). The statute requires the candidate of the party that won the last gubernatorial election to appear first beneath each office listed on the ballot, with the candidate of the second-place party appearing second. Fla. Stat. § 101.151(3)(a). In the nearly 70 years since its enactment, the statute has placed Democrats first on the ballot in 20 general elections and Republicans first in 14, including the 10 most recent general elections.

In 2018, three voters and six organizations that support the Democratic Party filed a complaint against the Florida Secretary of State to enjoin enforcement of the statute. They alleged that, because of “position bias,” the statute confers “an unfair electoral advantage” on Republicans, who have held the Governorship for the past 20 years and whose candidates have appeared first on the ballot during that time. Position bias, or the “primacy effect,” refers to the phenomenon that a small number of voters select the candidate who is listed first for an office on the ballot

4

solely because of the candidate’s position. In close elections, the complaint alleged, the primacy effect can give Republican candidates the “bump” needed to secure victory. By awarding the benefits of the primacy effect entirely to Republican candidates in recent years, the voters and organizations argued that the statute violates their rights under the First and Fourteenth Amendments as interpreted in Anderson v. Celebrezze, 460 U.S. 780 (1983), and Burdick v. Takushi, 504 U.S. 428 (1992).

Shortly after the voters and organizations filed their complaint, the National Republican Senatorial Committee and Republican Governors Association moved to intervene as defendants. See Fed. R. Civ. P. 24(b). The district court granted the motion. The Republican intervenors joined the Secretary in defending the challenged law as constitutional and opposing the relief the voters and organizations sought.

At a bench trial, the voters and organizations presented the testimony of three expert witnesses. Jon Krosnick, a professor at Stanford University, reviewed the academic literature and testified that the existence of the primacy effect is well- established by academic studies of elections. Based on his regression analyses of past Florida elections, Krosnick testified that candidates listed first on Florida ballots have historically gained an average advantage of about five percentage points. Jonathan Rodden, also a professor at Stanford University, testified about

5

the primacy effect in down-ballot races. Rodden testified that the primacy effect is more pronounced in down-ballot races, where voters often have less information about the candidates, than in top-of-ticket races. And Paul Herrnson, a professor at the University of Connecticut, testified about how ballot order contributes to “proximity error.” Herrnson testified that when voters make proximity errors—that is, accidentally select the candidate listed before or after the one they mean to select—the second-listed candidate is especially disadvantaged in races with more than two candidates. The reason for this disadvantage, Herrnson explained, is that voters who intend to select the first or last candidate in a list can err in only one direction, but voters who intend to select the second candidate can err in either direction.

The Secretary and the Republican intervenors presented the testimony of an expert witness, several election officials, and a corporate representative for one of Florida’s election machine vendors. Michael Barber, a professor at Brigham Young University, critiqued Krosnick’s methods and testified that Krosnick’s estimate of an average five-percent primacy effect was not valid. Maria Matthews, Director of the Florida Division of Elections, and several county Supervisors of Elections testified about the state interests the challenged law serves. They explained that the statute helps prevent voter confusion, allows voters to more quickly find their preferred candidate or party for a particular office, promotes

6

uniformity in administering elections across Florida’s 67 counties and over 6,000 precincts, and helps limit errors in ballot layout. Matthews and the Supervisors also testified about the logistical difficulties of implementing the voters’ and organizations’ requested relief, such as rotating the names of Democratic and Republican candidates between counties or between voting precincts within a county. And a corporate representative for an election machine vendor testified that he did not know whether the election machines could rotate Democratic and Republican candidates between the top two ballot positions and that it could take up to a year for the company to take the steps necessary for rotating candidate names.

After trial, the district court entered a final order. It rejected the Secretary’s and intervenors’ arguments that the lawsuit presented a nonjusticiable political question and that the voters and organizations lacked standing. And on the merits, it ruled that Florida’s method of ordering candidates on the ballot is unconstitutional.

The district court ruled that both the voters and the organizations proved Article III standing. It reasoned that an “impact on the right to vote” is “common to all election laws,” so the voters necessarily had an injury in fact. It also concluded that the organizations were injured because they spent resources to combat the primacy effect and because some unidentified voters who were members of the

7

organizations would be harmed by the primacy effect. The district court did not squarely address whether any injury from ballot order is traceable to the Secretary, but it reasoned that the Secretary is responsible for ballot order because she is Florida’s “chief election officer.” And although Florida law tasks the nonparty Supervisors with placing candidates on the ballot in the correct order, Fla. Stat. § 99.121, the district court ruled that relief against the Secretary could redress the voters’ and organizations’ injuries.

On the merits, the district court ruled that the law is unconstitutional under the approach established in Anderson, which requires courts to weigh the burdens imposed by an election regulation against the state interests justifying the measure. See 460 U.S. at 789. The district court found that “candidates of the major parties in Florida receive an average primacy effect vote of approximately five percent when listed first in their office block on the ballot.” And based on “Florida’s history of election results in which the margin of victory or defeat is less than three to five percentage points,” the district court found that the ballot statute “has impacted Plaintiffs’ First and Fourteenth Amendment rights by systematically allocating that small but statistically significant advantage to Republican candidates” in recent years. It concluded the statute was “politically discriminatory” because it awarded the benefits of the primacy effect to a single political party in any given election. And it found that the state’s asserted

8

justifications for the statute—upholding the legislature’s policy choice, preventing voter confusion, promoting uniformity, and promoting voter confidence in the election administration process—were “weak,” “not particularly persuasive,” and “not particularly strong on the specific facts of this case.”

The district court awarded declaratory and injunctive relief. It declared that Florida’s ballot-order scheme violated the First and Fourteenth Amendments. And it permanently enjoined the Secretary and the 67 Supervisors of Elections from implementing the ballot-order statute. Based on the Secretary’s “responsibility for general supervision and administration of the election laws,” the district court ordered the Secretary to neither “enforce, nor permit enforcement of,” the statute. The district court also ordered the Secretary to “take all practicable measures within the scope of [her] official authority to ensure compliance with the terms of [its] Order.” And it enjoined any “supervisor of elections of any Florida county”— none of whom were named as defendants or served with process as parties to this lawsuit—from issuing “any ballot which is organized pursuant to” the statute. The district court also ordered the Secretary to “provide written guidance to the supervisors of elections of Florida’s counties informing them that this Court has declared the [statute] unconstitutional” and to “include a true and correct copy of this Court’s order in her written guidance.”

9

The district court did not require Florida to adopt a specific alternative method of ordering candidates on ballots; it instead explained that two kinds of alternative schemes would be constitutional and allowed Florida to choose an alternative scheme. The first group of permissible schemes it identified were “rotational schemes,” which “rotate candidates’ names within their office blocks on a county-by-county or precinct-by-precinct basis.” The district court explained that these schemes “equaliz[e] the burden on voting rights” by “distributing the candidate name order effects more evenly across all candidates.” The second group of permissible schemes the district court identified are those that “cleans[e] the partisan taint from the process,” such as ordering candidates alphabetically by last name, by the order in which they submit their qualifying paperwork, or by lottery.

II. STANDARD OF REVIEW

We review questions of subject-matter jurisdiction de novo. United States v.

Pavlenko, 921 F.3d 1286, 1289 (11th Cir. 2019).

III. DISCUSSION

Federal courts have an independent obligation to ensure that subject-matter jurisdiction exists before reaching the merits of a dispute. “For a court to pronounce upon . . . the constitutionality of a state or federal law when it has no jurisdiction to do so is, by very definition, for a court to act ultra vires.” Steel Co. v. Citizens for a Better Env’t, 523 U.S. 83, 101–02 (1998). If at any point a federal

10

court discovers a lack of jurisdiction, it must dismiss the action. See MSP Recovery, LLC v. Allstate Ins. Co., 835 F.3d 1351, 1357 (11th Cir. 2016).

Unfortunately, the district court took its obligation to ensure its jurisdiction far too lightly. It dismissed weighty challenges to the voters’ and organizations’ standing under Article III as a “hodgepodge” of “[p]reliminary [m]iscellanea” and proceeded to declare Florida’s ballot statute unconstitutional and enter an injunction against both the Secretary and the nonparty Supervisors. In doing so, the district court acted ultra vires by ordering relief that the voters and organizations had no standing to seek.

Article III of the Constitution limits the subject-matter jurisdiction of federal courts to “Cases” and “Controversies.” U.S. Const. art. III, § 2. “To have a case or controversy, a litigant must establish that he has standing,” which requires proof of three elements. United States v. Amodeo, 916 F.3d 967, 971 (11th Cir. 2019). The litigant must prove (1) an injury in fact that (2) is fairly traceable to the challenged action of the defendant and (3) is likely to be redressed by a favorable decision. Lujan v. Defs. of Wildlife, 504 U.S. 555, 560–61 (1992).

Because the elements of standing “are not mere pleading requirements but rather an indispensable part of the plaintiff’s case, each element must be supported . . . with the manner and degree of evidence required at the successive stages of the litigation.” Id. at 561. If an action proceeds to trial, the facts necessary

11

to establish standing “must be supported adequately by the evidence adduced at trial.” Id. (internal quotation marks omitted). And when plaintiffs seek prospective relief to prevent future injuries, they must prove that their threatened injuries are “certainly impending.” Clapper v. Amnesty Int’l USA, 568 U.S. 398, 401 (2013) (internal quotation marks omitted).

We divide our discussion of why the voters and organizations lack Article III standing in two parts. First, we explain that neither the voters nor the organizations proved an injury in fact. Second, we explain that even if they had proved an injury, that injury would be neither traceable to the Secretary nor redressable by relief against her.

A. Neither the Voters nor the Organizations Proved an Injury in Fact.

We divide our discussion of injury in two parts. We first explain that the

individual voters failed to prove an injury. We then explain that the organizations likewise failed to prove an injury.

1. The Voters Failed to Prove an Injury.

Two of the three voters never testified at trial or in a deposition. The record

contains no evidence about any injuries those two individuals suffered in the past or may suffer in the future. Indeed, we do not even know whether they plan to vote in future Florida elections.

12

When confronted with this lack of evidence, the district court reasoned that an “impact on the right to vote” is “common to all election laws,” so the voters must have standing. But the Supreme Court has made clear that “a person’s right to vote is individual and personal in nature,” so “voters who allege facts showing disadvantage to themselves as individuals have standing to sue.” Gill v. Whitford, 138 S. Ct. 1916, 1929 (2018) (internal quotation marks omitted). And of course, “[t]he facts necessary to establish standing . . . must not only be alleged at the pleading stage, but also proved at trial.” Id. at 1931. Because they failed to offer any evidence at trial showing disadvantage to themselves as individuals, these two voters failed to prove an injury.

The only voter who offered any evidence at trial was Nancy Jacobson.

Jacobson testified that she “always vote[s],” that she “go[es] out of [her] way to vote in every election,” and that she consistently votes for Democratic candidates. But Jacobson failed to identify any difficulty in voting for her preferred candidate or otherwise participating in the political process.

Although her brief is less than clear on this point, Jacobson appears to identify two threatened injuries from the ballot statute. The first is that some unidentified Democratic candidates for whom she will vote in future elections will lose those elections because of the primacy effect. The second injury is that— regardless of the outcome of any election—the ballot statute “dilutes” the votes of

13

Democrats relative to Republicans by allocating the windfall vote entirely to Republican candidates. We reject both theories of injury.

To the extent Jacobson contends that she will be injured if a Democratic candidate for whom she votes loses an election or is at increased risk of losing, we disagree. A candidate’s electoral loss does not, by itself, injure those who voted for the candidate. Voters have no judicially enforceable interest in the outcome of an election. See Raines v. Byrd, 521 U.S. 811, 819, 824, 830 (1997). Instead, they have an interest in their ability to vote and in their vote being given the same weight as any other.

Raines, which involved the standing of legislators to challenge the constitutionality of the Line Item Veto Act, is instructive. Id. at 814, 816. Several legislators who voted against the Act sued to challenge it. Id. at 814. The Supreme Court explained that passage of the Act did not injure the legislators who voted against it because “their votes were given full effect,” and the disappointed legislators “simply lost that vote.” Id. at 824. The Court made clear that legislators have standing to challenge the defeat or enactment of legislation only if the outcome of the vote changed because their votes were “nullified”—that is, not counted at all. Id. at 823 & n.6. Jacobson does not argue that the ballot statute nullifies her vote. Instead, her complaint is that less careful voters will vote for Republican candidates solely because they appear first on the ballot, which might

14

cause her preferred candidates to lose. Like the legislators in Raines, the first harm she identifies is an unfavorable electoral outcome, wholly apart from any allegation of vote dilution or nullification.

Although the voting rights of legislators and citizens are not identical, see Nev. Comm’n on Ethics v. Carrigan, 564 U.S. 117, 126 (2011), we conclude that absent any evidence of vote dilution or nullification, a citizen is not injured by the simple fact that a candidate for whom she votes loses or stands to lose an election. And two of our sister circuits agree. See Berg v. Obama, 586 F.3d 234, 240 (3d Cir. 2009) (“Berg’s wish that the Democratic primary voters had chosen a different presidential candidate . . . do[es] not state a legal harm.”); Becker v. Fed. Election Comm’n, 230 F.3d 381, 390 (1st Cir. 2000) (holding that a candidate’s decreased “chance of being elected” was “hardly a restriction on voters’ rights and by itself [was] not a legally cognizable injury sufficient for standing”). Jacobson’s first alleged injury is legally insufficient to establish Article III standing.

Insofar as Jacobson argues that the ballot statute will injure her by diluting her vote relative to the votes of Republicans, she failed to prove any such injury. Her theory of vote dilution appears to be that, because of Florida’s ballot order and the primacy effect, it takes a greater number of careful Democratic voters than careful Republican voters to elect their preferred candidates. The reason for this disparity is that some less careful voters will select Republican candidates solely

15

because they happen to appear first on the ballot, thereby diluting the votes of careful Democratic voters. Even assuming that this kind of “vote dilution” counts as an Article III injury, the evidence Jacobson offered is insufficient to prove it.

In Gill, the Supreme Court addressed whether voters had standing to challenge a partisan gerrymander based on the dilution of their votes. 138 S. Ct. at 1929–31. Partisan gerrymandering operates by placing voters of one party “in legislative districts deliberately designed to ‘waste’ their votes in elections where their chosen candidates will win in landslides (packing) or are destined to lose by closer margins (cracking).” Id. at 1930. The voters’ theory of injury was that the partisan gerrymander caused their votes to “carry less weight” than they would “in another, hypothetical district” that had not been packed or cracked. Id. at 1931. But instead of offering evidence that they lived in a packed or cracked district, which could have shown “disadvantage to themselves as individuals,” id. at 1930 (internal quotation marks omitted), the voters rested their case on a “theory of statewide injury to Wisconsin Democrats,” id. at 1932.

To prove partisan vote dilution, the voters in Gill relied on an “average measure” of “partisan asymmetry” that compared the “statewide sum of one party’s wasted votes” to “the statewide sum of the other party’s wasted votes.” Id. at 1933. The Supreme Court held that this average measure of the partisan effects of a gerrymander was insufficient to establish the voters’ standing because it did

16

not “address the effect that a gerrymander has on the votes of particular citizens.” Id. It instead “measure[d] something else entirely: the effect that a gerrymander has on the fortunes of political parties.” Id.

Jacobson similarly relies on a statewide average measure of the primacy effect in Florida elections to prove the injury of partisan vote dilution. Her experts testified, and the district court found, that candidates who appear first on the ballot in Florida receive an average primacy effect vote of about five percent. But the experts acknowledged that this average measure tells us nothing about the existence or size of the primacy effect in any given election. Dr. Krosnick agreed that his analysis did not “mean that every Republican candidate receive[s] a [five] percent advantage by being listed first.” As he explained, the primacy effect will be larger in some races and smaller in others. Indeed, because Jacobson relies solely on an average measure of the primacy effect, we cannot know how often the primacy effect is zero and how often it is much greater than five percent. Any estimates we might make about the variance in the primacy effect across races would be pure speculation.

As in Gill, the average measure of partisan advantage on which Jacobson relies is insufficient to prove that her individual vote will be diluted. “We need not doubt [Jacobson’s] math” to reach this conclusion. Id. The reason her calculations cannot establish standing is that they “are an average measure.” Id. “They do not

17

address the effect” that ballot order and the primacy effect have “on the votes of particular citizens” in any given election. Id. (emphasis added). Instead, like the average measures at issue in Gill, Jacobson’s calculations “measure something else entirely: the effect that [ballot order and the primacy effect have] on the fortunes of political parties” across many elections. Id. And complaints about that effect are based on nothing more than “generalized partisan preferences,” which federal courts are “not responsible for vindicating.” Id.

Much like the average measure of wasted votes in Gill, the average measure of the primacy effect treats all elections “as indistinguishable, even though their individual situations are quite different.” Id. In low-information races between Democrats and Republicans, the primacy effect may be quite pronounced. But in an especially competitive, high-information race, the primacy effect may be negligible or nonexistent. Likewise, some races in noncompetitive districts may have no Republican candidates on the ballot at all and, hence, no primacy effect. An average measure of the primacy effect across all elections cannot tell us whether ballot order has diluted or will dilute Jacobson’s or any other citizen’s vote in any particular election. See id. (explaining that statewide average measures of partisan advantage were incapable of distinguishing between the effects of a gerrymander on one citizen as opposed to another).

18

Jacobson and the other voters failed to prove that they have suffered or will suffer partisan vote dilution in any particular election. As in Gill, this lawsuit presents a dispute “about group political interests, not individual legal rights.” Id. The “generalized partisan preferences” on which the voters rely cannot provide an injury in fact sufficient for Article III standing. Id.

2. The Organizations Failed to Prove an Injury.

For their part, the organizations rely on two theories of injury. They seek to establish associational standing based on the injuries of their members, see Summers v. Earth Island Inst., 555 U.S. 488, 494 (2009), and organizational standing based on their own injuries, see Common Cause/Ga. v. Billups, 554 F.3d 1340, 1350–51 (11th Cir. 2009). But they failed to prove an injury under either theory.

To establish associational standing, an organization must prove that its members “would otherwise have standing to sue in their own right.” Friends of the Earth, Inc. v. Laidlaw Envtl. Servs. (TOC), Inc., 528 U.S. 167, 181 (2000). The organizations contend that they have standing based on injuries suffered by Democratic voters and candidates who are their members. But five of the six organizations failed to even allege, much less prove, that they have any members—voters or candidates. That failure is fatal to their associational standing. See Summers, 555 U.S. at 498.

19

The only organization that describes itself as having members is the Democratic National Committee, but it failed to identify any of its members, much less one who will be injured by the ballot statute. See id. (requiring organizations to establish “that at least one identified member” will suffer an injury); see also Ga. Republican Party v. Sec. & Exch. Comm’n, 888 F.3d 1198, 1203–04 (11th Cir. 2018). And even if we accept as true the allegation of the complaint that the Committee’s members include Democratic voters and candidates in Florida, the Committee still has not proved that one of those unidentified members will suffer an injury.

Any voters and candidates in Florida face the same problem as Jacobson.

That is, because the Committee relies solely on an average measure of the primacy effect, we have no basis to conclude that the primacy effect will impact any particular voter or candidate in any particular election. Cf. Summers, 555 U.S. at 497 (rejecting the argument that an organization could establish standing if there was “a statistical probability that some of [its] members are threatened with concrete injury”). And the Committee has not proved that at least one of its unidentified members “is certain to be injured by” the primacy effect. Ga. Republican Party, 888 F.3d at 1204.

The organizations argue that they have suffered an injury in their own right by diverting resources to combat the effects of the ballot statute. In Havens Realty

20

Corp. v. Coleman, the Supreme Court held that an organization could establish standing to sue under the Fair Housing Act if it alleged, and later proved, that the challenged actions of the defendants drained its resources and thereby impaired its other operations. 455 U.S. 363, 378–79 & n.21 (1982). The housing organization in Havens Realty alleged that the defendants’ discriminatory renting practices required it “to devote significant resources to identify and counteract” those practices, which “perceptibly impaired” the organization’s “ability to provide counseling and referral services for low- and moderate-income homeseekers.” Id. at 379 (internal quotation marks omitted). The Court concluded that these allegations were sufficient to establish standing at the pleading stage, but it warned that at trial the organization would have to prove “that it has indeed suffered impairment in its role of facilitating open housing before it will be entitled to judicial relief.” Id. at 379 & n.21. Because statutory standing under the Fair Housing Act “extend[s] to the full limits” of standing under Article III of the Constitution, id. at 372, we have applied the reasoning of Havens Realty to determine whether an organization has Article III standing based on the diversion of its resources. See, e.g., Fla. State Conference of NAACP v. Browning, 522 F.3d 1153, 1165 & n.14 (11th Cir. 2008).

Consistent with Havens Realty, our precedent holds that “an organization has standing to sue on its own behalf if the defendant’s illegal acts impair its ability

21

to engage in its projects by forcing the organization to divert resources to counteract those illegal acts.” Id. at 1165. In Browning, we ruled that the NAACP and another organization had standing to challenge a voting requirement because the organizations would “divert personnel and time” from other activities “to educating volunteers and voters on compliance with” the requirement. Id. at 1166. In a later decision, we held that the NAACP had standing to challenge a law that required voters to present photo identification because the organization was “actively involved in voting activities and would divert resources from its regular activities to educate and assist voters in complying with” the law. Common Cause/Ga., 554 F.3d at 1350.

To establish resource diversion, the organizations cite the testimony of Daniel Kazin, the director of campaigns for the Democratic Congressional Campaign Committee. When asked why he believed the ballot statute harms the Committee, Kazin responded that “[b]ecause of the primacy effect, we need to spend additional resources in the target districts that we have.” The organizations also rely on similar testimony from Guy Cecil, the chair of Priorities USA, who testified that the organization had to “invest more resources into [Florida] in order to compensate for” the primacy effect.

Although resource diversion is a concrete injury, neither Kazin nor Cecil explained what activities the Committee or Priorities USA would divert resources

22

away from in order to spend additional resources on combatting the primacy effect, as precedent requires. See Havens Realty, 455 U.S. at 379 n.21; see also Browning, 522 F.3d at 1166 (“These resources would otherwise be spent on registration drives and election-day education and monitoring.”); Common Cause/Ga., 554 F.3d at 1350 (explaining that resources would be diverted “from ‘getting voters to the polls’ to helping them obtain acceptable photo identification” (alteration adopted)); Ga. Latino All. for Human Rights v. Governor of Ga., 691 F.3d 1250, 1260 (11th Cir. 2012) (observing that an immigration organization “cancelled citizenship classes to focus on” increased inquiries about a new law). Based on Kazin’s testimony, we do not know what activities, if any, might be impaired by the Committee’s decision to allocate “additional resources” to target districts because of the primacy effect. And Cecil’s testimony likewise fails to identify any activities that will be impaired by Priorities USA’s decision to “invest more resources” into Florida. Their testimony fails to establish an injury based on diversion of resources.

The organizations also contend that the ballot statute injures them by harming their mission of electing Democrats, but that harm is not a cognizable injury. An organization’s general interest in its preferred candidates winning as many elections as possible is still a “generalized partisan preference[]” that federal courts are “not responsible for vindicating,” no less than when individual voters

23

assert an interest in their preferred candidates winning elections. Gill, 138 S. Ct. at 1933; see also id. at 1932 (rejecting a voter’s “hope of achieving a Democratic majority in the legislature” as “a collective political interest” that cannot establish standing). Harm to an organization’s generalized partisan preferences describes only “a setback to [its] abstract social interests,” which is insufficient to establish a concrete injury in fact. Havens Realty, 455 U.S. at 379; see also Arcia v. Fla. Sec’y of State, 772 F.3d 1335, 1342 (11th Cir. 2014) (requiring “a concrete and demonstrable injury, not an abstract social interest” for organizational standing (alteration adopted) (internal quotation marks omitted)).

We need not decide whether a political party would have standing to challenge an electoral practice that harmed one of its candidate’s electoral prospects in a particular election. See, e.g., Tex. Democratic Party v. Benkiser, 459 F.3d 582, 586 (5th Cir. 2006) (holding that the Texas Democratic Party had standing to challenge action that would reduce “its congressional candidate’s chances of victory” in upcoming election). As discussed, the average measure of partisan advantage on which the organizations rely tells us nothing about whether ballot order has affected or will affect any particular candidate in any particular election. And in any event, the organizations do not argue that a particular candidate’s prospects in a future election will be harmed. They instead contend that they have standing based on “systemic disadvantage to [their] party relative to

24

other political parties.” Because that kind of harm from ballot order is based on nothing more than “generalized partisan preferences,” it is insufficient to establish standing. Gill, 138 S. Ct. at 1933.

B. Any Injury from Ballot Order Is Neither Traceable to the Secretary nor Redressable by Relief Against Her.

Even if the voters and organizations had proved an injury in fact, they would still lack standing because any injury would be neither traceable to the Secretary nor redressable by relief against her. Instead, any injury would be traceable only to 67 Supervisors of Elections and redressable only by relief against them. The voters and organizations’ failure to join the Supervisors as defendants is an independent reason that they lack standing to maintain this action.

To satisfy the causation requirement of standing, a plaintiff’s injury must be “fairly traceable to the challenged action of the defendant, and not the result of the independent action of some third party not before the court.” Lujan, 504 U.S. at 560 (alterations adopted) (internal quotation marks omitted). The voters and organizations contend that they are injured because Republicans, not Democrats, appear first on the ballot in Florida’s general elections. So for them to have standing, the order in which candidates appear on the ballot must be traceable to the Secretary—the only defendant in this action. The problem for the voters and organizations is that Florida law tasks the Supervisors, independently of the Secretary, with printing the names of candidates on ballots in the order prescribed

25

by the ballot statute. Fla. Stat. § 99.121 (“The names of [candidates] shall be printed by the supervisor of elections upon the ballot in their proper place as provided by law.”). The Secretary is responsible only for “certify[ing] to the supervisor of elections of each county . . . the names of persons nominated.” Id. The voters and organizations have offered no contrary evidence to establish that the Secretary plays any role in determining the order in which candidates appear on ballots. “Because the [Secretary] didn’t do (or fail to do) anything that contributed to [their] harm,” the voters and organizations “cannot meet Article III’s traceability requirement.” Lewis v. Governor of Ala., 944 F.3d 1287, 1301 (11th Cir. 2019) (en banc).

Our conclusion that any injury from ballot order is not traceable to the Secretary rests on the reality that the Supervisors are independent officials under Florida law who are not subject to the Secretary’s control. The Supervisors are constitutional officers who are elected at the county level by the people of Florida; they are not appointed by the Secretary. Fla. Const. art. VIII, § 1(d); Fla. Stat. § 98.015(1). The Florida Department of State’s organic statute does not list the Supervisors among its divisions, Fla. Stat. § 20.10(2), and the Board of County Commissioners, not the Department, compensates the Supervisors. Id. § 98.015(2). Only the Governor of Florida, not the Secretary, may suspend county officials such as the Supervisors, and only the state senate may remove them from office. Fla.

26

Const. art. IV, § 7; see also, e.g., Fla. Exec. Order No. 19-19 (executive order suspending the Supervisor of Elections for Palm Beach County); Fla. Exec. Order No. 18-342 (executive order suspending the Supervisor of Elections for Broward County). Indeed, the only means of control the Secretary has over the Supervisors is through coercive judicial process: she may bring “actions at law or in equity by mandamus or injunction to enforce the performance of any duties of a county supervisor of elections.” Fla. Stat. § 97.012(14). That the Secretary must resort to judicial process if the Supervisors fail to perform their duties underscores her lack of authority over them. Because the Supervisors are independent officials not subject to the Secretary’s control, their actions to implement the ballot statute may not be imputed to the Secretary for purposes of establishing traceability.

Contrary to the reasoning of the district court, the Secretary’s position as “the chief election officer of the state,” id. § 97.012, with “general supervision and administration of the election laws,” id. § 15.13, does not make the order in which candidates appear on the ballot traceable to her. We recently rejected a similar argument en banc. See Lewis, 944 F.3d at 1300. In Lewis, two workers sued the Attorney General of Alabama to challenge a state law that preempted a local ordinance requiring employers to pay higher wages. Id. at 1293–94. We explained that the workers’ injury—receiving lower wages because of the state law—was not traceable to the Attorney General because he had never enforced or threatened to

27

enforce the law, and the law itself contemplated no role for the Attorney General. Id. at 1296, 1298–99. And of particular relevance to this appeal, we rejected the workers’ reliance upon “a host of provisions of the Alabama Code that generally describe the Attorney General’s [enforcement] authority” to establish traceability. Id. at 1300. In the absence of any evidence that the Secretary controls ballot order, the voters and organizations likewise cannot rely on the Secretary’s general election authority to establish traceability. See id. at 1298–1300. Florida law expressly gives a different, independent official control over the order in which candidates appear on the ballot. See Fla. Stat. § 99.121.

Because the Secretary will not cause any injury the voters and organizations might suffer, relief against her will not redress that injury—either “directly or indirectly.” See Lewis, 944 F.3d at 1301 (internal quotation marks omitted). An injunction ordering the Secretary not to follow the ballot statute’s instructions for ordering candidates cannot provide redress, for neither she nor her agents control the order in which candidates appear on the ballot. Nor can declaratory relief against the Secretary directly redress any injury to the voters and organizations. A declaratory judgment against the Secretary does not bind the Supervisors, “who are not parties” to this action. Id. at 1302 (internal quotation marks omitted). As nonparties, the Supervisors are not “obliged . . . in any binding sense . . . to honor an incidental legal determination [this] suit produce[s].” Id. (internal quotation

28

marks omitted). They remain lawfully entitled to print candidates’ names on the ballot in the order prescribed by Florida law unless and until they are made parties to a judicial proceeding that determines otherwise. See id. at 1302–03.

To be sure, the district court ordered the Secretary to “provide written guidance to the supervisors of elections of Florida’s counties informing them that this Court has declared the [statute] unconstitutional” and to include “a true and correct copy of this Court’s order in her written guidance.” But this “notice” theory of redressability contravenes the “settled principle[]” that “it must be the effect of the court’s judgment on the defendant—not an absent third party—that redresses the plaintiff’s injury.” Id. at 1301 (internal quotation marks omitted). Any persuasive effect a judicial order might have upon the Supervisors, as absent nonparties who are not under the Secretary’s control, cannot suffice to establish redressability. See id. at 1305 (“If courts may simply assume that everyone (including those who are not proper parties to an action) will honor the legal rationales that underlie their decrees, then redressability will always exist.” (quoting Franklin v. Massachusetts, 505 U.S. 788, 825 (1992) (Scalia, J., concurring in part and concurring in the judgment))). “Redressability requires that the court be able to afford relief through the exercise of its power, not through the persuasive or even awe-inspiring effect of the opinion explaining the exercise of its power.” Id. (quoting Franklin, 505 U.S. at 825 (Scalia, J., concurring in part and

29

concurring in the judgment)). Because the voters and organizations failed to sue the officials who will cause any future injuries, even the most persuasive of judicial opinions would have been powerless to redress those injuries.

Even if we consider the persuasive effect of the judgment on the nonparty Supervisors, the voters and organizations have not established that redress is likely “as a practical matter.” Utah v. Evans, 536 U.S. 452, 461 (2002). They have not proved that declaratory relief against the Secretary will “significantly increase the likelihood” that the Supervisors will ignore state law and follow a federal decree that does not bind them. Lewis, 944 F.3d at 1301. The Supervisors are obliged under state law to continue printing candidates’ names “upon the ballot in their proper place as provided by law” regardless of what a federal court might say in an action that does not involve them. Fla. Stat. § 99.121. The district court’s decision rests on the flawed notion that by declaring the ballot statute unconstitutional, it eliminated the legal effect of the statute in all contexts. But “federal courts have no authority to erase a duly enacted law from the statute books.” Jonathan F. Mitchell, The Writ-of-Erasure Fallacy, 104 Va. L. Rev. 933, 936 (2018); see also Steffel v. Thompson, 415 U.S. 452, 469 (1974) (“Of course, a favorable declaratory judgment . . . cannot make even an unconstitutional statute disappear.” (internal quotation marks omitted)). Our power is more limited: we may “enjoin executive officials from taking steps to enforce a statute.” Mitchell, supra, at 936. And we

30

can exercise that power only when the officials who enforce the challenged statute are properly made parties to a suit.

The district court apparently understood that relief against the Secretary would not redress any injury to the voters and organizations, so it enjoined the Supervisors too. Its injunction stated, “No supervisor of elections of any Florida county . . . shall issue any ballot which is organized pursuant to the [ballot statute].” And its opinion warned the Supervisors against “selectively interpret[ing] parts of” its order “or otherwise avoid[ing] compliance.”

The district court exceeded its authority by purporting to enjoin the Supervisors, none of whom have ever been parties to this lawsuit. Although a district court may bind nonparties “who are in active concert” with a defendant, Fed. R. Civ. P. 65(d)(2)(C), that rule applies only when a plaintiff validly invokes federal jurisdiction by satisfying the traceability and redressability requirements of standing against a defendant. See In re Infant Formula Antitrust Litig., 72 F.3d 842, 843 (11th Cir. 1995) (“The Federal Rules of Civil Procedure do not create federal jurisdiction.”). If a plaintiff sues the wrong defendant, an order enjoining the correct official who has not been joined as a defendant cannot suddenly make the plaintiff’s injury redressable. The district court was without jurisdiction to enjoin the lone defendant in this action, much less the nonparty Supervisors. See Zenith Radio Corp. v. Hazeltine Research, Inc., 395 U.S. 100, 112 (1969) (holding

31

that the district court erred when it enjoined a nonparty that was never determined to be in active concert with a defendant).

The district court also relied on an inapposite decision, Democratic Executive Committee of Florida v. Lee, 915 F.3d 1312, 1318 (11th Cir. 2019), to conclude that relief against the Secretary would redress any injury to the voters and organizations. In Lee, a motions panel of this Court ruled that the Florida Secretary of State was a proper defendant under Ex parte Young, 209 U.S. 123 (1908), in an action challenging an election procedure administered by the county Supervisors of Elections. 915 F.3d at 1316, 1318 (citing Fla. Stat. § 101.68). But Article III standing and the proper defendant under Ex parte Young are “[s]eparate[]” issues, Lewis, 944 F.3d at 1295, and Lee addressed only the latter. To be a proper defendant under Ex parte Young—and so avoid an Eleventh Amendment bar to suit—a state official need only have “some connection” with the enforcement of the challenged law. 209 U.S. at 157. In contrast, Article III standing requires that the plaintiff’s injury be “fairly traceable” to the defendant’s actions and redressable by relief against that defendant. Lewis, 944 F.3d at 1298, 1301 (internal quotation marks omitted). The district court erred by treating Lee as if it addressed—let alone resolved—the standing issues in this suit.

Because the voters and organizations lack standing to sue the Secretary, we have no occasion to consider whether the Secretary is a proper defendant under Ex

32

parte Young—the only issue Lee addressed. See id. at 1296, 1306. Nor need we decide whether Lee—which was issued by a motions panel instead of a merits panel—is even binding precedent. See Democratic Exec. Comm. of Fla. v. Nat’l Republican Senatorial Comm., 950 F.3d 790, 795 (11th Cir. 2020) (declining to vacate the opinion of the motions panel in Lee after the appeal became moot because “the necessarily tentative and preliminary nature of [the] stay-panel opinion precludes the opinion from having an effect outside that case”).

We cannot agree with our colleague’s partial dissent that whether the voters and organizations satisfied traceability and redressability are “difficult questions” or that “principles of judicial restraint” counsel against deciding them. Dissenting Op. at 68. The resolution of those issues is straightforward, and the Secretary has repeatedly asked federal courts in our circuit to decide them. And we are unaware of any principle of judicial restraint that counsels against addressing multiple elements of standing, a threshold issue of justiciability. The Supreme Court has decided multiple elements of standing in alternative holdings, see Clapper, 568 U.S. at 401–02 (injury and traceability), and so have we, see, e.g., City of Miami Gardens v. Wells Fargo & Co., 931 F.3d 1274, 1283–84 (11th Cir. 2019) (injury and traceability); Kawa Orthodontics, LLP v. Sec’y, U.S. Dep’t of the Treasury, 773 F.3d 243, 245–48 (11th Cir. 2014) (deciding all three elements). There is nothing unusual or untoward about our alternative holdings on traceability and

33

redressability, as the partial dissent argues. See Dissenting Op. at 91–95. To the contrary, they promote judicial economy by resolving issues that have percolated in our circuit for years and are likely to recur in future litigation.

The partial dissent says that the Secretary never advanced in this case the argument we adopt today, id. at 70, 95, but that assertion tells only half the story. As the Secretary mentioned at oral argument, Oral Argument at 34:40–35:08 (Feb. 12, 2020), her office has repeatedly, if unsuccessfully, argued to the district judge who presided over this litigation that the Secretary has highly limited authority over county election officials, including the Supervisors. See, e.g., Rivera Madera v. Detzner, 325 F. Supp. 3d 1269, 1275 (N.D. Fla. 2018) (Walker, C.J.) (rejecting the Secretary’s argument that “he has no relevant power over the county supervisors of elections”); Fla. Democratic Party v. Detzner, No. 4:16-cv-607- MW-CAS, 2016 WL 6090943, at *4–5 (N.D. Fla. Oct. 16, 2016) (Walker, J.) (rejecting the Secretary’s arguments that “he cannot direct the [county] canvassing boards to comply with any order issued by this Court” and that “he does not possess the power to issue orders [to county officials] directing compliance with Florida’s election laws”). The Secretary made clear at oral argument that her office has not changed its position on this issue, even if in this lawsuit she elected not to raise the argument yet again before a district court that had repeatedly rejected the Secretary’s own understanding of her authority under state law. Oral Argument at

34

34:55–35:08 (Feb. 12, 2020) (“[W]e do not think we’re the right defendant. We have made this argument on several occasions . . . and, quite frankly, in the Northern District of Florida we have not succeeded . . . .”). So our ruling today is consistent with the Secretary’s longstanding view about the scope of her powers.

The partial dissent next contends that the Secretary’s authority to prescribe rules about ballot layout, Fla. Stat. § 101.151(9)(a), and to provide written direction to the Supervisors, id. § 97.012(16), may make the order in which candidates appear on the ballot traceable to her, Dissenting Op. at 74–79, but we do not see how. That the Secretary has the power to prescribe rules and issue directives about ballot order, which the Supervisors might well be obliged to follow, says nothing about whether she “possess[es] authority to enforce the complained-of provision,” as the causation element of standing requires. Lewis, 944 F.3d at 1299 (emphasis added) (quoting Dig. Recognition Network, Inc. v. Hutchinson, 803 F.3d 952, 958 (8th Cir. 2015)). If rulemaking authority were sufficient to establish traceability, plaintiffs could presumably also challenge a law by suing the legislators who enacted it instead of the officials who execute it. Although in many cases the same official will both make and execute a challenged regulation, that arrangement is not present here. See Fla. Stat. § 99.121.

The partial dissent also contends that an injunction forbidding the Secretary to provide the Supervisors with any instructions about ballot order would likely

35

provide redress, Dissenting Op. at 80–81, 88, but we again do not see how. Florida law already directs the Supervisors to place candidates on the ballot in the order “provided by law,” Fla. Stat. § 99.121—that is, in the order prescribed by the ballot statute, see id. § 101.151(3)(a). An injunction ordering the Secretary to stay silent would do nothing to muzzle these two sections of the Florida code, which already bind the Supervisors to list candidates in a particular order. Indeed, one of the Supervisors testified at trial that they “apply the [ballot] statute” because it “is the law.” There is no contrary evidence to suggest that the Supervisors would suddenly begin to disregard state law in the absence of instructions from the Secretary.

Under the partial dissent’s theory of traceability and redressability, the only relief that might possibly redress any injuries from ballot order would be an injunction ordering the Secretary to promulgate a rule requiring the Supervisors to place candidates on the ballot in an order contrary to the ballot statute. But the voters and organizations never requested such extraordinary relief, and for good reason. Any such relief would have raised serious federalism concerns, and it is doubtful that a federal court would have authority to order it. See Va. Office for Prot. & Advocacy v. Stewart, 563 U.S. 247, 255 (2011) (explaining that the Ex parte Young exception to sovereign immunity “is limited to [the] precise situation” in which “a federal court commands a state official to do nothing more than refrain from violating federal law”); cf. New York v. United States, 505 U.S. 144, 177–78

36

(1992) (holding that the federal government could not commandeer states to enact or enforce a federal regulatory scheme).

In any event, it is also far from clear that ordering the Secretary to promulgate a rule that is contrary to the ballot statute would even make redress likely. The voters and organizations have not argued that the Supervisors are likely to ignore a state statute that obliges them to place candidates on the ballot in a particular order in favor of a regulation issued by the Secretary. Again, their hesitation is not without good reason: Florida law is clear that when a regulation and a statute conflict, the statute prevails. See Nicholas v. Wainwright, 152 So. 2d 458, 460 (Fla. 1963). The partial dissent asserts that in this scenario the Supervisors would likely follow the Secretary’s instructions over the statute. Dissenting Op. at 88–89. But “[w]e do not know what would justify” the partial dissent’s confidence when Florida law is to the contrary. Lujan, 504 U.S. at 570 (plurality opinion).

It bears emphasis that even the district court understood the traceability and redressability problems inherent in this lawsuit. In an attempt to avoid those problems, it took the truly remarkable step of enjoining nonparties. Although the decision to enjoin nonparties was unjustifiable, it makes clear what the partial dissent says is murky: the Secretary plainly is not the cause of any alleged injuries from ballot order, and relief against her cannot redress those injuries.

37

To satisfy traceability and redressability, the voters and organizations should have sued the Supervisors of Elections instead of the Secretary of State. That approach would have made for more defendants, but nothing prevented the voters and organizations from taking that course of action. See Socialist Workers Party v. Leahy, 145 F.3d 1240, 1243 (11th Cir. 1998) (explaining that the plaintiffs filed suit “against the Secretary of State and the sixty-seven county supervisors of elections”). Because they failed to sue the Supervisors, the voters and organizations lack Article III standing.

IV. CONCLUSION

By entering a judgment on the merits when it had no case or controversy before it, the district court offered “no more than an expression of opinion upon the validity of the [law] in question.” Muskrat v. United States, 219 U.S. 346, 362 (1911). That kind of advisory opinion is beyond the power of federal courts. The district court should have dismissed the action for lack of standing. It erred by reaching the merits and entering an injunction against nonparties whom it had no authority to enjoin. We VACATE the judgment against the Secretary and REMAND with instructions to dismiss for lack of jurisdiction.

38

WILLIAM PRYOR, Circuit Judge, concurring:

In addition to the voters’ and organizations’ lack of standing, this lawsuit suffers from another fatal jurisdictional defect: it presents a nonjusticiable political question. Complaints of unfair partisan advantage based on the order in which candidates appear on the ballot bear all the hallmarks of a political question under the recent decision of the Supreme Court in Rucho v. Common Cause, 139 S. Ct. 2484 (2019). No judicially discernable and manageable standards exist to determine what constitutes a “fair” allocation of the top ballot position, and picking among the competing visions of fairness “poses basic questions that are political, not legal.” Id. at 2500. And even if courts could agree on a standard for fairly ordering ballots, no objective measures exist to identify violations of that standard. See id. at 2501. This lawsuit asks us “to reallocate political power between the two major political parties, with no plausible grant of authority in the Constitution, and no legal standards to limit and direct” our decision. Id. at 2507. That kind of complaint is “outside the courts’ competence and therefore beyond the courts’ jurisdiction.” Id. at 2494.

To place this dispute in context, I first provide a brief history of ballot regulation in America. This history reveals that concerns about ballot order are not new. And the political branches of state governments have long taken a variety of approaches to addressing these perceived concerns.

39

At the Founding, Americans voted using their voices, a show of hands, or ballots prepared by individual voters, political parties, and party organizations. Joseph P. Harris, Election Administration in the United States 150–52 (1934); Burson v. Freeman, 504 U.S. 191, 200 (1992) (plurality opinion). The Southern states retained voice voting the longest; Kentucky did not abandon it until 1890. Harris, supra, at 151 & n.8. But because of the abuses associated with voice voting, including bribery and voter intimidation, most states began to use paper ballots within two decades of the Founding. Burson, 504 U.S. at 200 (plurality opinion).

As paper ballots became more widespread, some of the abuses associated with voice voting “reinfected the election process.” Id. Political parties printed their ballots on colorful paper, often with distinctive designs, so that the ballots could be easily distinguished at the polls. Harris, supra, at 151. This practice threatened ballot secrecy and made bribery and voter intimidation easier to accomplish, so state legislatures enacted laws that required the use of white paper or official envelopes for ballots. Id. at 151–52.

Other abuses that had not been possible with voice voting also arose. The party organizations that printed the ballots engaged in fraudulent practices. They would sometimes distribute fake ballots that bore the markings of one party but contained only a few of that party’s candidates—“just enough to fool the unwary.”

40

Id. at 152. And in some elections, the party organizations would decline to place the names of some qualified candidates on their ballots, which made it impossible for those candidates to be elected. Id.

These abuses led Americans to adopt the “Australian ballot”—an official ballot containing the names of all qualified candidates that election officials distribute at the polls. Id. at 152–54. As its name suggests, this kind of ballot first appeared in Australia in the 1850s, and American states rapidly adopted it between 1887 and 1900. Id. Although a “true Australian ballot” grouped the names of all candidates beneath the office for which they were running without identifying their party affiliation, most American states did not adopt the original form of the Australian ballot. Id. at 154. Instead, the states modified the Australian ballot “to retain the strength of the political parties.” Id. Many states grouped the candidates of each party into separate columns with a party circle at the top of each column that enabled voters to “vote a ‘straight ticket’ with a single mark.” Id. at 155. Others retained the Australian ballot’s grouping of candidates by office, adding only the party designation of the candidates. Id. at 154–55.

Concerns about the order in which candidates appear on the ballot have been with us since the adoption of the Australian ballot. By 1934, states followed several different practices for ordering their ballots. States that used party-column ballots determined the parties’ position on the ballot from left to right in one of five

41

ways: (1) alphabetically, (2) a definite order fixed by state law, with the party in power given the first column, (3) in order of the votes received for some office in the last election, (4) by the election officer charged with preparing the ballot, or (5) by lottery. Id. at 180. Among political parties, the left-most column was “most desired,” but the advantage gained from that position was viewed as “not great.” Id.

In states that used office-group ballots, a common view was that “the position at the top of a list of candidates is of material help to the candidate thus favored.” Id. at 181. States dealt with this perceived advantage for first-listed candidates in different ways. Some rotated the names of candidates by ballot or voting precinct, but others established a uniform ballot order based on the votes a party received in the last general election, candidate last name, the order in which nominating petitions were received, or lottery. Id. at 181–83.

Today, states continue to use different methods to order their ballots. Some states, like Florida, determine ballot order based on the results of the last election for Governor or another state office. Ariz. Rev. Stat. Ann. § 16-502(E); Conn. Gen. Stat. § 9-249a(a); Fla. Stat. § 101.151(3)(a); Ga. Code Ann. § 21-2-285(c); Ind. Code § 3-11-2-6(a)(1); Md. Code Ann., Elec. Law §§ 1-101(dd), 9-210(j)(2); Mich. Comp. Laws § 168.703; Mo. Rev. Stat. § 115.239(1); Neb. Rev. Stat. § 32- 815(1); N.Y. Elec. Law § 7-116(1); 25 Pa. Cons. Stat. § 2963(b); Tex. Elec. Code

42

Ann. § 52.091(b). Others determine it based on the party that currently holds a majority in the state legislature, Tenn. Code Ann. §§ 2-1-104(a)(11)–(12), 2-5- 208(d)(1), or the number of votes each party received in the last congressional election, Wyo. Stat. Ann. § 22-6-121(a). The Minnesota rule is similar to the NFL draft: candidates of the major party that won the fewest votes in the preceding election are listed first. Minn. Stat. § 204D.13(2). Delaware has the most straightforward rule: Democrats first, then Republicans. Del. Code Ann. tit. 15, § 4502(a)(5). And still other states order their ballots based on nonpartisan considerations. See, e.g., Ala. Code § 17-6-25 (alphabetical by candidate last name); Ark. Code Ann. § 7-5-207(c)(1) (random lottery); Ky. Rev. Stat. Ann. § 118.225(1) (rotating candidate names in each congressional district).

Against this wide array of state practice, voters and organizations brought this constitutional challenge to Florida’s 70-year-old law that assigns the top ballot position to candidates of the incumbent Governor’s party. They alleged violations of the First Amendment and the Equal Protection Clause of the Fourteenth Amendment as interpreted in a line of decisions beginning with Anderson v. Celebrezze, 460 U.S. 780 (1983), and Burdick v. Takushi, 504 U.S. 428 (1992). Although the voters and organizations also alleged a traditional equal-protection claim based on the ballot statute’s “disparate treatment” of similarly-situated political parties, the district court did not rule on that claim and it is not before us.

43

The voters and organizations’ complaint, in a nutshell, is that Florida’s ballot statute confers an impermissible partisan advantage on Republicans by virtue of the primacy effect. Because Republican candidates enjoy a “windfall vote” of approximately five percentage points from disinterested voters who reflexively pick the first candidate, the Democratic voters and organizations have a harder time electing their preferred candidates than if Florida distributed the windfall vote more evenly. They argue that this regime burdens their right to vote and should be evaluated using the approach established in Anderson and Burdick.

The recent decision of the Supreme Court in Rucho compels the conclusion that this complaint presents a nonjusticiable political question. This complaint shares the same critical feature that led the Supreme Court to hold complaints of partisan gerrymandering nonjusticiable in Rucho: neither this complaint of partisan advantage from ballot order nor complaints of partisan advantage from redistricting can be adjudicated using “judicially discernible and manageable” standards. Rucho, 139 S. Ct. at 2502.

In Rucho, the Supreme Court held that complaints of partisan gerrymandering are nonjusticiable for two main reasons. First, these complaints invariably rest on a threshold determination about what a “fair” apportionment of political power looks like. See id. at 2499–500. The Court reasoned that one possible standard of fairness—proportional representation—might have been

44

judicially manageable but had no basis in constitutional law or the history of the Republic. See id. at 2499. And absent proportional representation, the Court explained, “it is not even clear what fairness looks like in this context.” Id. at 2500. Fairness could mean creating the greatest number of competitive districts, districting to ensure that each party receives its proportional share of “safe” seats, or adhering to traditional districting criteria. Id. (internal quotation marks omitted). And choosing between these different visions of fairness “poses basic questions that are political, not legal.” Id.

Second, even if courts could agree on a standard of fairness, they would have to determine how much deviation from that standard in pursuit of partisan interests was permissible. Id. at 2501. Some amount of partisan gerrymandering is constitutional and inevitable. Id. at 2497. To hold that legislators could never consider partisan interests in districting “would essentially countermand the Framers’ decision to entrust districting to political entities.” Id. And in addition to the problem of deciding how much partisan intent is too much, complaints of partisan gerrymandering also present line-drawing problems concerning partisan effect—judges must decide “how much partisan dominance is too much.” Id. at 2498 (emphasis added) (internal quotation marks omitted). For example, to police partisan gerrymandering, courts would “have to decide the ideal number of seats for each party and determine at what point deviation from that balance went too

45

far.” Id. at 2501. Because the Constitution supplies neither an objective standard for the fair apportionment of political power nor any principled basis for identifying violations of that (nonexistent) standard, the Court concluded that complaints of partisan gerrymandering present nonjusticiable political questions. Id. at 2500–02.

Under the reasoning of Rucho, complaints of partisan advantage based on ballot order are likewise nonjusticiable political questions. The voters and organizations’ complaint is based on the notion that Florida’s ballot statute, by virtue of the primacy effect, confers an unfair partisan advantage on the party that last won the Governorship. But courts cannot rely on legal standards to adjudicate this kind of complaint because it does not allege any burden on individual voting rights. Instead, adjudicating this kind of complaint would require courts to pick among various conceptions of a politically “fair” ballot order that have no basis in the Constitution. For that reason, the complaint “poses basic questions that are political, not legal.” Id. at 2500. And even if a judicially discernable and manageable standard for fairly ordering a ballot existed, there are no standards for determining how much of a departure from an ideal ballot order amounts to a constitutional violation. See id. at 2501. As I will explain, Rucho cannot be persuasively distinguished from this appeal.

46

The basic problem with the voters and organizations’ complaint is that it is not based on the right to vote at all, so we cannot evaluate their complaint using the legal standards that apply to laws that burden the right to vote. As the voters and organizations correctly point out, we must evaluate laws that burden voting rights using the approach of Anderson and Burdick, which requires us to weigh the burden imposed by the law against the state interests justifying the law. Common Cause/Ga. v. Billups, 554 F.3d 1340, 1352 (11th Cir. 2009). But “we have to identify a burden before we can weigh it.” Crawford v. Marion Cty. Election Bd., 553 U.S. 181, 205 (2008) (Scalia, J., concurring in the judgment). And here it is impossible to identify a burden on voting rights imposed by the ballot statute that is susceptible to the balancing test of Anderson and Burdick.

The statute at issue here is unlike any law that this Court or the Supreme Court has ever evaluated under Anderson and Burdick. The statute does not make it more difficult for individuals to vote, see, e.g., Crawford, 553 U.S. at 198 (plurality opinion) (photo-identification law); Common Cause/Ga., 554 F.3d at 1354 (same), or to choose the candidate of their choice, see Burdick, 504 U.S. at 430 (prohibition on write-in voting). It does not limit any political party’s or candidate’s access to the ballot, which would interfere with voters’ ability to vote for and support that party or candidate. See, e.g., Timmons v. Twin Cities Area New Party, 520 U.S. 351, 353–54, 358–59 (1997) (law forbidding individuals to appear

47

on the ballot as the candidate of more than one party); Norman v. Reed, 502 U.S. 279, 288–89 (1992) (ballot-access law for new parties); Munro v. Socialist Workers Party, 479 U.S. 189, 199 (1986) (ballot-access law for minor-party candidates); Anderson, 460 U.S. at 782, 786 (early filing deadline for candidate paperwork); Fulani v. Krivanek, 973 F.2d 1539, 1539, 1543 (11th Cir. 1992) (ballot-access law for minor-party candidates). Nor does it burden the associational rights of political parties by interfering with their ability to freely associate with voters and candidates of their choosing. See, e.g., Wash. State Grange v. Wash. State Republican Party, 552 U.S. 442, 451–52 (2008); Clingman v. Beaver, 544 U.S. 581, 587, 593 (2005); Tashjian v. Republican Party of Conn., 479 U.S. 208, 213–14 (1986). And to state the obvious, the statute certainly does not create the risk that some votes will go uncounted or be improperly counted. See, e.g., Democratic Exec. Comm. of Fla. v. Lee, 915 F.3d 1312, 1318–20 (11th Cir. 2019) (challenge to signature-match procedures for absentee and provisional ballots); Wexler v. Anderson, 452 F.3d 1226, 1232 (11th Cir. 2006) (challenge to manual- recount procedures under which some ballots might “receive a different, and allegedly inferior, type of review in the event of a manual recount”). All the statute does is determine the order in which candidates appear in each office block on the ballot.

48

If the statute burdened voting or associational rights even slightly, we could apply legal standards to determine whether the burden was unconstitutional. Under Anderson and Burdick, we would weigh the burden imposed by the law against the state interests justifying that burden. See Common Cause/Ga., 554 F.3d at 1352. But because the statute does not burden the right to vote, we cannot engage in that kind of review. The voters and organizations ask us to decide not whether the ballot statute burdens identifiable voting or associational rights, but whether it confers an unfair partisan advantage on the Republican Party.

Instead of basing their complaint on individual voting or associational rights, the voters and organizations allege a novel complaint premised on the idea that the extra votes that flow from top ballot position should be distributed “fairly” between the major political parties. The “crux of [their] constitutional claim,” they explain, “is the way in which” the ballot statute distributes the primacy vote “between similarly-situated major parties.” In their view and the district court’s, fairness means distributing the primacy vote either evenly between the major parties or on some apolitical basis, like random lottery or alphabetically by candidate last name.

But sensible as those approaches might be, they are hardly the only ways to conceive of a “fair” ballot order. As in Rucho, “it is not even clear what fairness looks like in this context.” 139 S. Ct. at 2500. Instead of splitting the primacy

49

effect between the two major parties, perhaps Florida should ensure that each political party on the ballot—including minor parties—has an equal number of its candidates listed first for office. After all, parties that have qualified to be on the ballot are similarly situated with respect to any right to be first on the ballot. Or, because that approach might give an undue advantage to minor parties with few supporters, perhaps Florida should distribute the primacy effect proportionately based on the number of registered voters in each party. That is, if 20 percent of registered voters belong to one political party, that party’s candidates should appear first on 20 percent of the ballots, and so on. Maybe Minnesota’s approach is fairest: award the primacy effect entirely to one party—the party that received the fewest votes in the last election. Minn. Stat. § 204D.13(2). One could imagine many other ballot schemes that plausibly claim to be the fairest way, or at least a fair way, to distribute the primacy effect, including the one Florida adopted nearly 70 years ago: giving all parties the chance to win the primacy effect at each gubernatorial election.

As in the partisan gerrymandering context, picking among these alternatives “poses basic questions that are political, not legal.” Rucho, 139 S. Ct. at 2500. “There are no legal standards discernible in the Constitution for making such judgments, let alone limited and precise standards that are clear, manageable, and politically neutral.” Id. “Any judicial decision on what is ‘fair’ in this context

50

would be an ‘unmoored determination’ of the sort characteristic of a political question beyond the competence of the federal courts.” Id. (quoting Zivotofsky v. Clinton, 566 U.S. 189, 196 (2012)).

And even if courts could discern in the Constitution a standard of fairness for evaluating ballot-order regimes, they would run headlong into the second problem the Supreme Court identified in Rucho. There are no discernable and manageable standards “to answer the determinative question”: How much partisan advantage from ballot order is too much? See id. at 2501; see also id. at 2498 (asking “how much partisan dominance is too much” (internal quotation marks omitted)). It is impossible to ensure that each candidate or party in a particular election appears at the top of the ballot an equal number of times. Election officials cannot know in advance how many ballots will be cast in a given race, let alone how many ballots will be cast in each county or voting precinct or which counties and precincts have the largest numbers of disinterested voters. Relatedly, how large must the primacy effect be to create a constitutional problem? Two percent of voters? Five percent? Some greater share? If the standard is an “outcome determinative” number of voters, then any disparity in allocating the primacy effect could violate the Constitution in close races. Would awarding the primacy effect to a single political party be constitutional in a noncompetitive state where it would make no difference to electoral outcomes but unconstitutional in a

51

battleground state? As with partisan gerrymandering, even if courts “knew which version of fairness to be looking for, there are no discernible and manageable standards for deciding whether there has been a violation.” Id. at 2501.

At bottom, the voters and organizations’ challenge to the ballot statute rests on the notion “that each party must be influential in proportion to its number of supporters.” Id. Their complaint is that some voters who are neither Democrats nor Republicans will vote for the Republican candidate solely because the Republican is listed first, giving Republicans an advantage beyond their actual number of supporters. But the Supreme Court has never accepted that baseline as providing a justiciable standard in any context. It has instead emphatically rejected the idea that federal courts are “responsible for vindicating generalized partisan preferences.” Id. (quoting Gill v. Whitford, 138 S. Ct. 1916, 1933 (2018)).

The federal judiciary’s “constitutionally prescribed role is to vindicate the individual rights of the people appearing before it.” Id. (quoting Gill, 138 S. Ct. at 1933). Where an election law does not burden the right to vote in any way, we cannot vindicate individual rights. And we “have no license to reallocate political power between the two major political parties, with no plausible grant of authority in the Constitution, and no legal standards to limit and direct [our] decisions.” Id. at 2507. The complaints of partisan gerrymandering in Rucho cannot be persuasively distinguished from the voters and organizations’ complaint. A

52

complaint that the order in which candidates appear on a ballot confers an impermissible partisan advantage to one party presents a nonjusticiable political question.

One possible response to the preceding analysis is that because the voters and organizations have not alleged any burden on voting rights, their complaint fails on the merits though it remains justiciable. But a complaint can both fail to state a constitutional violation and be nonjusticiable if there are no judicially discernible and manageable standards to adjudicate it. Take complaints of partisan gerrymandering. In the light of Rucho, we know that any complaint that a redistricting plan is unconstitutionally partisan must be dismissed as nonjusticiable—even if the challenged plan is so fair that it could not possibly violate the Constitution. Nor must a particular practice even be capable of violating the Constitution for challenges to that practice to be nonjusticiable. Our guide, again, is Rucho. We do not know whether partisan gerrymandering can ever violate the Constitution; in its 46 years of attempting to adjudicate those complaints, the Supreme Court never declared a single redistricting plan unconstitutionally partisan. Id. at 2491, 2497–98, 2507. But even though partisan gerrymandering may not violate the Constitution, challenges to that practice are nonetheless nonjusticiable because they are unsuited for resolution by federal

53

courts. Id. at 2507–08. The same is true for complaints of partisan advantage based on ballot order.

The voters and organizations’ arguments that their complaint is justiciable are unconvincing. To make their case, they attempt to distinguish Rucho, invoke a host of inapposite precedents, and posit hypothetical laws that bear no resemblance to the challenged law in this action. None of their arguments have merit.

The voters and organizations first suggest that Rucho is distinguishable because the Supreme Court searched for a judicially manageable standard to police partisan gerrymandering “for decades” without success. But Rucho makes clear that complaints of partisan gerrymandering have always been nonjusticiable; it did not impose a requirement that courts first struggle to identify a justiciable standard for some period of time before declaring a complaint nonjusticiable. Complaints of partisan gerrymandering did not become nonjusticiable only after the Court tried and failed to develop a standard. See Lester v. United States, 921 F.3d 1306, 1312 (11th Cir. 2019) (W. Pryor, J., respecting the denial of rehearing en banc) (“[W]e should be mindful of the difference between a change in judicial doctrine and a change in law.” (emphasis added)). Instead, the Court’s inability to discern a manageable standard was evidence that these complaints had always been “outside the courts’ competence and therefore beyond the courts’ jurisdiction.” Rucho, 139 S. Ct. at 2494; see also Lester, 921 F.3d at 1313 (“Without distinguishing between

54

judges’ understanding of the law and the law itself, . . . the Supreme Court [could not] meaningfully describe a past decision of its own as ‘wrong the day it was decided.’” (quoting Planned Parenthood of Se. Penn. v. Casey, 505 U.S. 833, 863 (1992) (joint opinion of O’Connor, Kennedy, and Souter, JJ.))). As discussed below, the judiciary’s experience with partisan ballot-order complaints provides similar evidence that no judicially discernible and manageable standards exist to adjudicate them—that is, that these complaints have always been nonjusticiable.

The voters and organizations also argue that Rucho is distinguishable because some amount of partisan gerrymandering is constitutionally permissible in redistricting, but partisan considerations are off limits in the realm of election administration. And if partisan considerations are forbidden in election administration, that reality arguably eliminates the line-drawing problem the Supreme Court faced in Rucho—how much partisanship is too much? In the voters and organizations’ view, any partisanship is too much partisanship in this context. Cf. id. at 2502 (explaining that complaints of racial gerrymandering can rightly ask “for the elimination of a racial classification” but that complaints of partisan gerrymandering “cannot ask for the elimination of partisanship”).

This argument has at least two problems. First, partisan considerations are not entirely off limits in election administration. Partisan motivations do not doom a nondiscriminatory election law if “valid neutral justifications” also support the

55

law. Crawford, 553 U.S. at 204 (plurality opinion); see also Common Cause/Ga., 554 F.3d at 1355. But even if partisan motivations were entirely off limits in election administration, that fact would not eliminate the line-drawing problems inherent in the voters and organizations’ complaint, which is based solely on the partisan effects of the ballot statute. The voters and organizations have never argued that the Democrat-led legislature and Democratic governor that enacted the statute were motivated by impermissible partisan intent. Their complaint does not ask for the elimination of partisan intent in ballot order. It asks for a fair allocation of the primacy vote, much like the complaints of partisan gerrymandering in Rucho asked for “a [fair] level of political power and influence.” 139 S. Ct. at 2499.

The voters and organizations next contend that because other challenges to election regulations are justiciable, theirs must be too. They point to Williams v. Rhodes, 393 U.S. 23, 24, 28 (1968), which held that a challenge to laws that “made it virtually impossible” for certain political parties to access the ballot was justiciable. But Williams provides no support for their position.

Laws that limit the ability of candidates or parties to access the ballot burden “voters’ freedom of choice and freedom of association.” Anderson, 460 U.S. at 806 (emphasis added); see also Socialist Workers Party, 479 U.S. at 193 (explaining that ballot-access restrictions “impinge upon the rights of individuals to associate for political purposes, as well as the rights of qualified voters to cast their votes

56

effectively”). Standards exist to assess the burdens imposed by restrictions on ballot access. See Socialist Workers Party, 479 U.S. at 193–99. But no standards exist to judge challenges to the partisan advantage conferred by ballot order.

The voters and organizations contend that if we determine their complaint is nonjusticiable, other more nefarious ballot laws will be insulated from judicial review. They offer examples of hypothetical laws that require a “thumbs-up” or asterisk symbol next to candidates of the Governor’s party, or that require the names of those candidates to appear in larger font, bold print, or a different color. Because challenges to these laws should be justiciable, they argue, so also should challenges to laws that govern ballot order.

A ruling that this lawsuit is nonjusticiable would not mean that challenges to these kinds of ballot laws are also nonjusticiable. The Elections Clause, which commits the regulation of the “Times, Places and Manner of holding Elections” to state legislatures, U.S. Const. art. I, § 4, cl. 1, provides a judicially discernable and manageable standard to evaluate nonprocedural laws about ballot content. The Supreme Court has held that the Elections Clause establishes the boundaries of state authority over elections. See Cook v. Gralike, 531 U.S. 510, 523 (2001) (“[T]he States may regulate the incidents of such elections, including balloting, only within the exclusive delegation of power under the Elections Clause.”). In Cook, the Court invalidated a Missouri law that placed a pejorative designation

57

next to candidates who refused to support term limits because the law did not regulate the time, place, or manner of elections but instead sought to disparage or endorse particular candidates. Id. at 523–26. “Thumbs-up” laws could be evaluated under that standard, as could other laws that arguably do not regulate the manner of elections, like laws that provide favorable font choices for certain candidates. But Cook and the Elections Clause provide no standard to evaluate laws that govern ballot order. Unlike the law at issue in Cook or a “thumbs-up” law, laws that govern ballot order plainly regulate the manner of elections and are within the power of states to enact. To use an Australian ballot, Florida, like every other state, necessarily had to decide the order in which candidates’ names appear on the ballot. And the choice of what order to adopt cannot be evaluated using legal standards because it “poses basic questions that are political, not legal.” Rucho, 139 S. Ct. at 2500.

One might think that holding the voters and organizations’ complaint to be nonjusticiable would mean that all challenges to ballot order are nonjusticiable, but that is not so. Rucho makes clear that one kind of challenge to a law can be justiciable and another nonjusticiable depending on whether judicially discernable and manageable standards exist to adjudicate the complaint. The Court explained that challenges to a redistricting plan based on racial gerrymandering or violations of the one-person, one-vote principle are justiciable because manageable standards

58

exist to adjudicate those complaints, even though challenges to the same redistricting plan based on its partisan effects are nonjusticiable. See id. at 2501– 02. Similarly, if the voters and organizations’ complaint were that Florida’s ballot order somehow made it more difficult for Democrats to vote for their candidate of choice, their complaint would be justiciable, and we would have to weigh the burden imposed by the law against the state’s regulatory interests. See Burdick, 504 U.S. at 434; Common Cause/Ga., 554 F.3d at 1352. It is conceivable that a law governing ballot order could have that effect: a law that required Democrats to be placed on the last page of the ballot and all other candidates to appear on the earlier pages might make it more difficult for Democratic voters to find and select their preferred candidate. But that is not the kind of complaint the voters and organizations brought. They instead ask us to fairly apportion the primacy vote among the political parties, and that complaint falls squarely within Rucho’s definition of a political question.

The voters and organizations also argue that the decisions of other courts adjudicating complaints of partisan advantage based on ballot order prove that their complaint is justiciable. But the relevant decisions all predate Rucho. See, e.g., Libertarian Party of Va. v. Alcorn, 826 F.3d 708, 717 (4th Cir. 2016); Green Party of Tenn. v. Hargett, 767 F.3d 533, 550–51 (6th Cir. 2014); Koppell v. N.Y. State Bd. of Elections, 153 F.3d 95, 96 (2d Cir. 1998); McLain v. Meier, 637 F.2d 1159,

59

1167 (8th Cir. 1980); Sangmeister v. Woodard, 565 F.2d 460, 465 (7th Cir. 1977). And none of the decisions addressed whether complaints of partisan advantage based on ballot order are justiciable. More fundamentally, that courts have attempted to adjudicate a complaint does not mean the complaint is justiciable. Indeed, before Rucho, numerous lower courts had crafted some standard to adjudicate complaints of partisan gerrymandering. See Rucho, 139 S. Ct. at 2502– 06.

Even taken on their own terms, these decisions support, rather than undermine, the conclusion that the voters and organizations’ complaint is nonjusticiable. They provide evidence that the voters and organizations’ complaint is inherently standardless, much as the many prior decisions attempting to adjudicate complaints of partisan gerrymandering did in Rucho. See id. at 2497–98. Because complaints of partisan advantage based on ballot order are not based on the right to vote at all, the courts in each of these decisions were forced to decide what constituted a fair method of allocating of the top ballot position and then determine whether the challenged law so departed from that standard of fairness that it violated the Constitution.

Unsurprisingly, the courts settled on different and sometimes contradictory standards. The Fourth Circuit, for example, concluded that “facially neutral and nondiscriminatory” ballot-order laws “impose[] only the most modest burdens” on

60

voting and associational rights and for that reason survive scrutiny under Anderson and Burdick. Libertarian Party of Va., 826 F.3d at 717. The Eighth Circuit, in contrast, held that “position advantage must be eliminated as much as is possible” and decided that the “fairest remedy” was “some form of ballot rotation whereby ‘first position’ votes are shared equitably by all candidates.” McLain, 637 F.2d at 1169 (emphasis added). The Seventh Circuit did not require rotation of the top ballot spot among all candidates; instead, it held that laws governing ballot order pose no constitutional problem so long as they are “neutral” in character and do not intentionally favor one class of candidates over another. Sangmeister, 565 F.2d at 465–68. And at least one court has concluded that even a “neutral” method of assigning ballot position—alphabetically by candidate last name—violated the state constitutional rights of a candidate whose name would never allow him to appear at the top of the ballot. Kautenburger v. Jackson, 333 P.2d 293, 294–95 (Ariz. 1958). These decisions strengthen the conclusion that there are no judicially discernable and manageable standards for adjudicating complaints of partisan advantage based on ballot order. Such complaints present competing visions of fairness that are “unguided and ill suited to the development of judicial standards.” Rucho, 139 S. Ct. at 2501 (internal quotation marks omitted). Federal judges have no business deciding them.

61

Finally, the voters and organizations contend that the Supreme Court’s summary affirmance in Mann v. Powell, 398 U.S. 955 (1970) (mem.), establishes that their complaint is justiciable, but that is plainly wrong. The law at issue in Mann placed candidates on the ballot in the order they submitted their nominating petitions and gave the Illinois Secretary of State unfettered discretion to break ties if he received multiple petitions simultaneously. 314 F. Supp. 677, 678–79 (N.D. Ill. 1969), aff’d, 398 U.S. 955. When the Secretary received two or more petitions simultaneously, he chose to break the tie in favor of incumbents. Id. A three-judge district court enjoined that practice, id. at 679, and the Supreme Court summarily affirmed, Mann, 398 U.S. at 955. But because neither court addressed whether it had jurisdiction to hear the dispute, Mann is not precedential as to justiciability. See Steel Co. v. Citizens for a Better Env’t, 523 U.S. 83, 91 (1998) (“We have often said that drive-by jurisdictional rulings of this sort . . . have no precedential effect.”); United States v. More, 7 U.S. (3 Cranch) 159, 172 (1805) (Marshall, C.J., at oral argument) (“No question was made, in that case, as to the jurisdiction. It passed sub silentio, and the court does not consider itself as bound by that case.”).

Not only is Mann a nonprecedential drive-by ruling on the issue of justiciability, but it also provides no basis to conclude that the Supreme Court has ever adjudicated a complaint based on the partisan effects of ballot order. The Supreme Court has cautioned that we must not overread its summary affirmances:

62

“the precedential effect of a summary affirmance extends no further than the precise issues presented and necessarily decided by those actions. A summary disposition affirms only the judgment of the court below, and no more may be read into our action than was essential to sustain that judgment.” Anderson, 460 U.S. at 784 n.5 (internal quotation marks omitted); see also Mandel v. Bradley, 432 U.S. 173, 176 (1977) (“Because a summary affirmance is an affirmance of the judgment only, the rationale of the affirmance may not be gleaned solely from the opinion below. When we summarily affirm, without opinion, we affirm the judgment but not necessarily the reasoning by which it was reached.” (alteration adopted) (internal quotation marks omitted)). It is impossible to tell whether the Supreme Court based its summary affirmance in Mann on the district court’s rationale—that the Secretary’s decision to “favor incumbents in breaking ties” violated the plaintiffs’ “right to fair and evenhanded treatment,” 314 F. Supp. at 679—or on some alternative basis. The Court could just as well have affirmed on an alternative ground—for example, the ground urged by the plaintiffs “that the statute [was] unconstitutional for failure to set out standards to guide the Secretary’s action in breaking ties.” Id. at 678. In short, the voters and organizations’ reliance on Mann is completely misplaced.

The voters and organizations’ attempts to escape the reasoning of Rucho are all unavailing. Despite their many protests, Rucho compels the conclusion that

63

complaints of unfair partisan advantage based on ballot order present nonjusticiable political questions. Although Rucho may seem counterintuitive to federal judges who are used to usurping the authority of state legislatures to regulate elections, it should not. The Constitution commits the “Times, Places and Manner” of holding congressional elections to legislatures—the state legislatures in the first instance, subject to any regulations Congress prescribes. U.S. Const. art. I, § 4, cl. 1. Our founding charter never contemplated that federal courts would dictate the manner of conducting elections—in this lawsuit, down to the order in which candidates appear on a ballot.

Alexander Hamilton explained in Federalist 59 that “a discretionary power over elections ought to exist somewhere,” but that somewhere was not the federal judiciary. The Federalist No. 59, at 306 (Alexander Hamilton) (George W. Carey & James McClellan eds., 2001). Instead, Hamilton identified “only three ways in which this power could have been reasonably organized.” Id. It could be “lodged wholly in the national legislature, or wholly in the state legislatures, or primarily, in the latter, and ultimately in the former.” Id. The Constitution, of course, adopted the third option. But the district court in this action assumed for itself the “discretionary power over elections” that the Constitution assigns to the state and federal legislatures, in contravention of clear Supreme Court precedent that should

64

have prevented it from reaching the merits of this dispute. Its decision to do so was error.

65

JILL PRYOR, Circuit Judge, concurring in part and dissenting in part:

For the past 20 years, the Republican candidate’s name has been listed first on every general election ballot in every race in every contested partisan election in the state of Florida. In this case, individuals and organizations sued Florida’s Secretary of State to challenge as unconstitutional the state statute governing ballot ordering in general elections. Florida law requires the names of candidates from the governor’s party to be listed first for each office on the general election ballot. See Fla. Stat. § 101.151(3)(a). The district court found after a bench trial that this ballot-ordering scheme has awarded Republican candidates a “small but statistically significant advantage” due to the tendency of some voters to select a candidate simply because his name is listed first (a phenomenon known as the “primacy effect” or “candidate name order effect”). Doc. 202 at 2. 1 As a result, the court concluded, the scheme violated the First and Fourteenth Amendments.

The merits question in this appeal is whether Florida’s ballot-order law violates the Constitution by awarding the advantage created by the primacy effect to candidates based on their affiliation with the governor’s political party. But before we can address the merits, we must be sure that the plaintiffs have standing to bring this challenge. See Warth v. Seldin, 422 U.S. 490, 498 (1975) (“[T]he

1 “Doc. #” refers to the numbered entry on the district court’s docket.

66

question of standing is whether the litigant is entitled to have the court decide the merits of the dispute or of particular issues.”).

The Constitution limits the power of the judiciary to deciding “Cases” and “Controversies.” U.S. Const. art. III, § 2, cl. 1. To satisfy the case-or-controversy requirement, a plaintiff must have standing to sue. See Spokeo, Inc. v. Robins, 136 S. Ct. 1540, 1547 (2016). To establish standing, a plaintiff must show (1) an injury in fact; (2) a causal connection between the injury and the causal conduct, meaning that the injury is fairly traceable to the defendant’s conduct; and (3) a likelihood that the injury will be redressed by a favorable decision. Id.; MSP Recovery, LLC v. Allstate Ins. Co., 835 F.3d 1351, 1357 (11th Cir. 2016). Failure to demonstrate any one of these three elements defeats a plaintiff’s standing. The majority opinion concludes that the plaintiffs lack standing to sue the Secretary of State because at trial they failed to prove all three: (1) that any plaintiff suffered an injury in fact; (2) that any injury a plaintiff suffered, if one existed, was fairly traceable to the Secretary’s conduct, and (3) that any injury a plaintiff suffered, if one existed, could be redressed by a judgment against the Secretary.

I agree with the majority that the plaintiffs lack standing because at trial they failed to prove that any plaintiff suffered an injury in fact. In my view, the plaintiffs’ strongest argument is that plaintiff Priorities USA—which describes itself as a progressive organization that works to increase participation of

67

underrepresented communities in civil life and the political process and supports Democrats running for office around the country—suffered an injury in fact under a diversion-of-resources theory. The plaintiffs argue that to overcome the advantage the ballot-order scheme grants Republican candidates as a result of the primacy effect, Priorities USA had to devote more resources to the candidates it supported in Florida, which drained its resources. But I must agree with the majority that even Priorities USA’s evidence is insufficient. The problem with the plaintiffs’ reliance on a diversion-of-resources theory is that they presented no evidence of any activities Priorities USA “divert[ed] resources away from in order to spend additional resources on combatting the primacy effect.” Maj. Op. at 21- 22 (citing Havens Realty Corp. v. Coleman, 455 U.S. 363, 379 n.21 (1982)). Because the plaintiffs failed to prove that any plaintiff suffered an injury in fact, I join Part III-A of the majority opinion holding that the plaintiffs failed to establish standing and thus the district court lacked jurisdiction to reach the merits.

The absence of injury in fact defeats standing; there is no need to go further.

But the majority does not stop here. In Part III-B, the majority opinion announces alternative holdings that the plaintiffs lack standing because, even assuming they proved injury, their injury was not fairly traceable to the Secretary of State or redressable in a lawsuit against her because under Florida law different government officials—county election supervisors—are responsible for preparing

68

the ballots listing the candidates in order.2 I believe that the questions of Florida law we must resolve to decide traceability and redressability—in ordering candidates’ names on ballots, what role does the Secretary of State play and does she exercise sufficient authority over county election supervisors?—are considerably harder than the majority makes them out to be. I cannot agree with the majority’s decision to forge ahead nonetheless and reach alternative holdings that depend on resolving unsettled and difficult questions of state law about the scope of a state official’s role and the extent of her power. What’s more, in holding that the plaintiffs failed to establish traceability or redressability, the majority unveils a new understanding of those concepts that imposes a heavier burden on the plaintiffs than our precedent supports and creates a split with authority from other circuits. I respectfully dissent from Part III-B of the majority opinion because I believe that principles of judicial restraint counsel us to refrain from addressing traceability and redressability.3

2 There is no question that the majority opinion reaches alternative holdings here because any one of these three independent grounds would mean that the plaintiffs lacked standing to pursue the case. See Evans v. Ga. Reg’l Hosp., 850 F.3d 1248, 1255-56 (11th Cir. 2017).

3 In his separate concurrence, Judge William Pryor offers another reason why this lawsuit should be dismissed, that it presents a non-justiciable political question. He argues that the Supreme Court’s recent decision in Rucho v. Common Cause, 139 S. Ct. 2484 (2019), “compels the conclusion that this complaint presents a nonjusticiable political question.” Concurrence at 44. I cannot agree that Rucho compels that conclusion when the Supreme Court indicated that its reasoning was specific to the gerrymandering context, which the Court described as one of the “most intensely partisan aspects of American political life.” Rucho, 139 S. Ct. at 2507. Because Judge William Pryor’s concurrence reflects only his own views and not those of the majority, I address the application of the political question doctrine no further.

69

I.

As the majority opinion reflects, the issues of traceability and redressability turn on our interpretation of Florida law. To establish that their injury was traceable to the Secretary of State’s conduct, the plaintiffs had to prove a “causal connection” between their injuries and the conduct they complained of. Focus on the Family v. Pinellas Suncoast Transit Auth., 344 F.3d 1263, 1273 (11th Cir. 2003). To establish that their injury was redressable by suing the Secretary, the plaintiffs had to prove that it was “likely, as opposed to merely speculative” that their hypothetical injury would “be redressed by a favorable decision” against her. Loggerhead Turtle v. Cty. Council of Volusia Cty., 148 F.3d 1231, 1253 (11th Cir. 1998) (internal quotation marks omitted). In deciding whether the plaintiffs could satisfy the traceability and redressability requirements, then, we must consider the extent of the Secretary of State’s authority under Florida state law when it comes to ballot ordering.

According to the majority opinion, the plaintiffs failed to establish redressability or traceability because Florida law tasks county election supervisors, “independently of the Secretary, with printing the names of candidates on ballots in the order prescribed by the ballot statute.” Maj. Op. at 24-25. The majority opinion interprets Florida law as (1) placing all responsibility for the ordering of candidates on the ballots with the county election supervisors, thus giving the

70

Secretary no “role in determining the order in which candidates appear on ballots,” and (2) giving the Secretary no control over the county election supervisors. Id. at 25-26. Notably, no Florida court has ever held that the Secretary of State’s authority is so limited. Perhaps even more remarkable, the majority’s argument about the Secretary of State’s authority is one that she herself never raised in this case, even though, as the majority opinion demonstrates, it would have been to her advantage.4 And yet the majority opinion concludes that this case presents a straightforward question about the proper interpretation of Florida’s Election Code.

I urge that we abstain from alternative holdings on traceability and redressability because these issues turn on questions about how the state of Florida has structured its government to divide power between state and local officials in the crucial function of holding elections. These are questions that are best answered by Florida state courts. Reading Florida’s Election Code in the absence of such guidance, however, I think the better answer to the question whether the Secretary of State plays a sufficient role in setting ballot order and exercises adequate control over the county election supervisors to support standing is “yes.”

4 The majority opinion downplays the significance of the Secretary of State’s decision not to argue traceability or redressability in this case by suggesting that it was simply a strategic call not to raise these arguments before a district court judge who had repeatedly rejected them in other cases. But this supposition fails to account for the Secretary’s decision not to argue traceability and redressability in her briefing on appeal. See Ouachita Watch League v. Jacobs, 463 F.3d 1163, 1169-70 (11th Cir. 2006) (recognizing that a party may raise a jurisdictional issue for the first time on appeal).

71

In this section, I review the provisions in Florida’s Election Code defining the scope of the Secretary of State’s authority, with emphasis on three provisions that the majority opinion seems to misapprehend. Importantly, if the majority opinion’s understanding of Florida law is wrong, so are its holdings on traceability and redressability.

A.

To understand the scope of the Secretary of State’s authority, we must interpret Florida’s Election Code, following Florida’s rules of statutory construction. See Robbins v. Garrison Prop. & Cas. Ins. Co., 809 F.3d 583, 586 (11th Cir. 2015). Those rules provide that “legislative intent is the most important factor that informs our analysis.” Quarantello v. Leroy, 977 So. 2d 648, 651 (Fla. Dist. Ct. App. 2008). Legislative intent must be gleaned “primarily from the text of the statute,” focusing on “the actual language used by the Legislature.” Id. (internal quotation marks omitted). In examining statutory text, courts in Florida “will not look merely to a particular clause in which general words may be used, but will take in connection with it the whole statute.” Id. (internal quotation marks omitted); see also Orange Cty. v. Singh, 268 So. 3d 668, 671 n.4 (Fla. 2019) (“In construing the Florida Election Code, it is necessary to read all provisions in pari materia.”).

72

When we look at Florida’s Election Code as a whole, we see that the Florida Legislature has divided responsibility for administering elections among state and local officials. The Secretary of State, appointed by the governor, serves as the head of the Department of State, oversees its Division of Elections, and is charged with “general supervision and administration of the election laws.” See Fla. Stat. §§ 15.13; 20.10(1), (2)(a). The Secretary is the “chief election officer of the state” responsible for “[o]btain[ing] and maintain[ing] uniformity in the interpretation and implementation of the election laws.” Id. § 97.012(1). To maintain this uniformity, the Department of State may “adopt by rule uniform standards for the proper and equitable interpretation and implementation of the requirements of chapter 97 through 102 and 105 of the Election Code.” Id. 5 The ballot-order statute is found in chapter 101.

Although the Secretary of State plays a role in overseeing elections across the state, most of the work in administering elections occurs at the county level. Each of Florida’s 67 counties elects its own election supervisor who oversees how elections in her county are conducted. See id. § 98.015(1). Each supervisor is

5 This provision excludes two chapters of the Election Code, 103 and 104, from the Department of State’s power to adopt uniform standards for the interpretation and implementation of the requirements of the other chapters. Chapter 103 primarily addresses the procedures that govern the electors who cast Florida’s votes for President of the United States in the electoral college, Fla. Stat. §§ 103.011-103.141, and Chapter 104 sets forth criminal penalties for violating Election Code provisions, id. §§ 104.011-104.43. Neither chapter is relevant to the issues before us.

73

responsible for appointing an election board, comprised of poll workers for each precinct in the county, that conducts the voting in each precinct on election day. See id. § 102.012(1), (4). The supervisor’s responsibilities also include “updat[ing] voter registration, enter[ing] new voter registrations into the statewide voter system, and act[ing] as the official custodian of documents” related to elector registration “and changes in voter registration status.” Id. § 98.015(3).

Most relevant here, county election supervisors print the ballots that voters use. Before a general election, the Department of State certifies to each county election supervisor the names of the candidates running for office that are to appear on ballots in that county. Id. § 99.121. The Election Code then directs that the “names of such persons shall be printed by the supervisor of elections upon the ballot in their proper place as provided by law.” Id. Based solely on this language, the majority opinion concludes that election supervisors set the order of the candidate’s names on the ballot. From there, the majority opinion concludes that the Secretary of State exercises no control over how Florida election supervisors carry out their duty to order ballots because under Florida’s state constitution and the Election Code the county officials are elected by the voters, and the Secretary of State does not appoint them, does not compensate them, may not suspend them, and may not remove them from office.

74

My concern is that the majority opinion’s analysis of Florida law is incomplete because it reads provisions of Florida’s Election Code in isolation, contrary to Florida’s rules of statutory construction. In particular, the majority’s interpretation fails to appreciate the effect of three relevant provisions of the Code, which suggest that the Florida Legislature intended for the Secretary of State to play a substantive role in setting the ballot order and overseeing how election supervisors carry out their duties in this regard.

The first provision the majority opinion largely overlooks is the one in which the Florida Legislature charges the Department of State with “adopt[ing] rules prescribing a uniform primary and general election ballot.” See id. § 101.151(9)(a). These rules must incorporate the requirements of § 101.151— which includes the ballot-order scheme in subsection (3)(a)—and may “prescribe additional matters” including rules governing “[i]ndividual race layout.” See id. § 101.151(9)(a). Among other things, the Department’s “rules must graphically depict a sample . . . general election ballot form.” Id. § 101.151(9)(b). The Department’s form ballots incorporate the ballot-ordering scheme. See, e.g., Form Official General Election Ballot, DS-DE 207 (eff. Sep. 12, 2018), available at https://www.flrules.org/Gateway/reference.asp?No=Ref-06441 (last accessed April 28, 2020); see also Fla. Admin. Code Ann. r. 1S-2.032(15)(b) (2020) (stating that the ballot form is incorporated by reference into the Secretary’s rules). This is

75

consistent with the Secretary of State’s explanation at oral argument that after the primary elections, “we have a ballot order that the [Secretary of] State provides to the [county election] supervisors . . . then they design and set the ballot per the order that is provided by the State.” Oral Argument Recording at 36:10-36:46.6 Indeed, the Elections Code’s use of the terms “prescribing” and “prescribe”

when describing the Secretary of State’s power to make rules governing general election ballots and individual race layout confirms that the Florida Legislature granted the Secretary of State authority to direct election supervisors when they perform the task of preparing ballots, including the ordering of candidates. The plain meaning of “prescribe” is “[t]o make an authoritative ruling.” Prescribe, The Oxford English Dictionary (online ed.) (last accessed April 27, 2020); see Nehme v. Smithkline Beecham Clinical Labs., Inc., 863 So. 2d 201, 205 (Fla. 2003) (explaining that under Florida law, “[w]hen necessary, the plain and ordinary meaning of words can be ascertained by reference to a dictionary” (internal quotation marks omitted)). In trying to discover what the Florida Legislature intended when it adopted the Election Code, I cannot imagine that when it directed

6 A fuller quotation provides insight into the division of responsibility between the Secretary of State and county election supervisors:

[A]fter the primary we have a ballot order that the [Secretary of] State provides to the supervisors. And they set their ballots choosing their preferred printer, their preferred software, their preferred machines that have all been approved. And then they design and set the ballot per the order that is provided by the State.

Oral Argument Recording at 36:10-36:46.

76

the Secretary to adopt rules incorporating the requirements of the ballot-order statute in “prescribing” general election ballots, it contemplated that county supervisors administering those elections would not be required to follow them.

This should have been enough to give the majority pause, but there is a second provision that the majority opinion misapprehends. Section 97.012(16) authorizes the Secretary of State to “[p]rovide written direction and opinions to the supervisors of elections on the performance of their official duties with respect to the Florida Election Code or rules adopted by the Department of State.” Fla. Stat. § 97.012(16). This provision appears to flatly contradict the majority’s opinion that the Secretary plays no role in and has no authority over the election supervisors’ “performance of their official duties with respect to the Florida Election Code” when it comes to the Code’s ballot-order provision. Id.

The power to issue written “direction” to election supervisors, according to the term’s plain and ordinary meaning, is the power to “instruct[]” the election supervisors on “how to proceed or act” in carrying out their official duties and to give them “authoritative guidance.” Direction, The Oxford English Dictionary (online ed.) (last accessed April 27, 2020).7 Again, why would the legislature

7 As an example, the Secretary of State recently issued a directive instructing how county election supervisors are to carry out their statutory duties under Fla. Stat. § 101.657 to select sites for early voting. See Fla. Dep’t of State, Directive 2020-01—Early Voting Sites on College & University Campuses and Fla. Stat. 101.657(1)(a) (Apr. 2, 2020), https://dos.myflorida.com/media/702989/directive-2020-01.pdf. Under Florida law, election supervisors must operate early voting sites. See Fla. Stat. § 101.657(1)(a). An election

77

include such a provision if it intended that the election supervisors had no obligation to follow the Secretary’s directions and opinions?

This brings me to the third provision of the Florida Election Code that the majority opinion neglects to afford the significance I believe is due. As the majority opinion points out, § 97.012(14) gives the Secretary of State the power to bring an action at law or in equity by mandamus or injunction “to coerc[e]” a county supervisor of elections to perform any duties with respect to the Election Code or to comply with any rule adopted by the Department of State. Maj. Op. at 26; see Fla. Stat. § 97.012(14). The majority opinion views this provision as evidence that the Secretary of State lacks authority over the election supervisors because she must rely on the judicial process to coerce an election supervisor to comply.

supervisor may conduct early voting only at certain locations, such as a main or branch office of the election supervisor, a city hall, a permanent public library facility, a fairground, a civic center, or a courthouse. Id. In selecting early voting sites, the election supervisor must “provide all voters in the county an equal opportunity to cast a ballot, insofar as is practicable” and ensure that there is “sufficient nonpermitted parking to accommodate the anticipated amount of voters.” Id.

In Directive 2020-01, the Secretary of State instructed election supervisors how to perform these duties. She explained to election supervisors that they were not required to limit early voting sites to locations that “have a certain number of nonpermitted parking” spots but must ensure that the “early voting sites collectively within a county” provide sufficient nonpermitted parking spots to accommodate the anticipated number of early voters. Directive 2020-01 at ¶¶ 7-8. She then listed factors for election supervisors to consider when determining whether the early voting sites offered sufficient parking. Id. at ¶ 8. Although the county election supervisors are elected county officials who operate outside the Department of State, this directive demonstrates that the Secretary of State issues binding written directions to instruct them on the performance of their official duties under the Election Code.

78

I lack the majority’s confidence that this provision signals the Secretary’s lack of authority over the election supervisors. I find it significant that the Florida Legislature expressly gave the Secretary of State a cause of action, particularly a mandamus action—an “extraordinary remedy”—to compel an election supervisor to follow the Department of State’s rules. State ex rel. Perkins v. Lee, 194 So. 315, 317 (Fla. 1940). After all, it is well-established under Florida law that a writ of mandamus is available only when the duty sought to be coerced is “ministerial in nature” and the “respondent is under a clear legal duty to act.” State ex rel. Cherry v. Stone, 265 So. 2d 56, 57 (Fla. Dist. Ct. App. 1972) (emphasis added). If county election supervisors are under a clear legal duty to follow her Department’s rules, then it cannot be true that the Secretary lacks the authority to direct them. Rather than supporting the majority opinion’s conclusion, § 97.012(14) appears to undercut it by showing that the Secretary of State possesses the authority to compel election supervisors to perform their duties in accord with her rules and directives.8

8 It is true that the Secretary of State does not appoint county election supervisors and has no power to suspend them or remove them from office, but I think the majority infers too much from these facts. The Florida Legislature designed a system of government in which the Secretary of State lacks these particular powers but nonetheless possesses the authority to oversee and direct how local officials carry out their duties, to ensure compliance with state election law and maintain uniformity of election procedures throughout the state.

Furthermore, although mandamus may seem like an indirect and inefficient remedy, after a more fulsome look at the Secretary’s authority, I think it is reasonable to assume that the mandamus power would need to be exercised only rarely.

79

When all of these provisions are read together as a unified whole, it appears to me that the Code gives the Secretary of State the power to set ballot-order rules and control how election supervisors organize ballots. To be clear, I am not answering these questions of Florida state law, which, as I explain below, should be answered by Florida state courts. But I am confident that the majority opinion’s conclusions about the limits of the Secretary of State’s authority rest on a woefully incomplete analysis of Florida’s Election Code.

B.

If the majority opinion is wrong about the scope of the Secretary of State’s authority under Florida law, that would mean that the plaintiffs satisfied the standing requirements of traceability and redressability. Let me explain. First, traceability: If the Secretary plays a role in ordering candidates’ names on general election ballots following the ballot-order statute, any hypothetical injury the plaintiffs suffered as a result of Florida’s ballot-order law would, at a minimum, “flow indirectly from” the Secretary’s actions. Focus on the Family, 344 F.3d at 1273 (recognizing that the traceability inquiry is “concerned with something less than the concept of proximate cause” (internal quotation marks omitted)).

The fact that the plaintiffs’ hypothetical injuries also could be fairly traced to the county election supervisors does not change the analysis. An injury cannot be “the result of the independent action of some third party not before the court.”

80

Lujan v. Defs. of Wildlife, 504 U.S. 555, 560 (1992) (emphasis added) (alterations adopted) (internal quotation marks omitted). But standing “is not defeated merely because the alleged injury can be fairly traced to the actions of both parties and non-parties.” Loggerhead Turtle, 148 F.3d at 1247.

Now, redressability: A hypothetical injury arising from the challenged law would have been redressed by the district court’s injunction, which, among other things, directed the Secretary of State not to “enforce . . . the ballot order scheme described in section 101.151(3)(a).” Doc. 202 at 72. Under this injunction, the Secretary would have to cease providing county election supervisors with form ballots and promulgating rules and regulations that effectuated the Election Code’s ballot-order scheme—meaning that when preparing ballot forms the Secretary would have to use a different method for ordering the candidates for each office. She could have selected any method other than putting candidates from the governor’s political party first in every race. As the Secretary has explained, her department provides each county election supervisor with a list of candidates in the order required by the ballot-order statute. Given that the Secretary provides the lists and oversees and directs how the county election supervisors carry out their duties, it seems “likely, as opposed to merely speculative” that any hypothetical injury the plaintiffs suffered as a result of enforcement of the ballot-order statute

81

would be redressed by the district court’s relief. Loggerhead Turtle, 148 F.3d at 1253 (quoting Lujan, 504 U.S. at 561).

The majority opinion’s primary argument about traceability and redressability is that the Secretary of State lacks a sufficient connection to Florida’s ballot-order scheme because she plays no role in setting ballot order and exercises no control over county election supervisors who set ballot order. As I explained above, the majority opinion reaches this conclusion only by ignoring, for the most part, three key provisions of Florida’s Election Code. When the majority opinion finally gets around to acknowledging these three provisions, it shifts gears and raises an entirely new argument—that when a state official exercises authority conferred on her by state law to promulgate rules and regulations under a statute, the official does not “enforce” the statute. Maj. Op. at 34 (emphasis omitted) (internal quotation marks omitted). The majority opinion warns that if we were to conclude that the Secretary’s rule-making power gives her the authority to enforce the Election Code, “plaintiffs could presumably also challenge a law by suing the legislators who enacted it instead of the officials who execute it.” Id.

I disagree that the statutory scheme reveals that the Secretary of State does not enforce the ballot-order statute. To me, the Secretary of State’s role in elections, specifically ballot ordering, points more clearly to the conclusion that she enforces the statute. The Secretary prepares and provides to county election

82

supervisors uniform ballot forms that incorporate the ballot-order scheme, promulgates rules under the Election Code including the ballot-order statute, and oversees how county election supervisors carry out their duties, all in fulfilling her responsibility (hers alone) to maintain uniformity in the interpretation and implementation of the Code throughout the state. See Fla. Stat. §§ 97.012(1), (14), (16); 101.151(9). The majority opinion accepts that the Secretary “instruct[s]” supervisors about ballot order. Maj. Op. at 34. Yet it cites no authority supporting its conclusion that a state official afforded these sorts of responsibilities does not enforce the statute. Nor does it cite any authority suggesting that an executive- branch state official who carries out such responsibilities has a similar relationship to the enforcement of the statute as a state legislator who voted to enact it.

Given all of this, I think the better conclusion is that the Secretary of State’s enforcement connection with the ballot-order statute is sufficient to establish that any injury the plaintiffs suffered “flow[ed]” at least “indirectly” from her actions and that it is “likely” that any such injury would be redressed by injunctive relief against the Secretary. Focus on the Family, 344 F.3d at 1273. I recognize that in this case the issues of traceability and redressability both turn on the Secretary’s role in enforcing the ballot-order statute. But this is hardly surprising because often “redressability and traceability overlap as two sides of a causation coin.” Nova Health Sys. v. Gandy, 416 F.3d 1149, 1159 (10th Cir. 2005) (internal

83

quotation marks omitted); see also Dynalantic Corp. v. Dep’t of Def., 115 F.3d 1012, 1017 (D.C. Cir. 1997) (same).

To support its argument that the Secretary of State lacks a sufficient connection to the statute’s enforcement, the majority opinion relies on our recent en banc decision in Lewis v. Governor of Alabama, 944 F.3d 1287 (11th Cir. 2019) (en banc). Lewis concerned the Birmingham city council’s passage of an ordinance raising the minimum wage for workers in the city. Id. at 1292. In response, the Alabama Legislature adopted a statewide minimum-wage law, effectively nullifying Birmingham’s ordinance. Id. at 1292-93. Employees who worked in Birmingham, along with several public interest groups, sued the Attorney General of Alabama, claiming racial discrimination under multiple theories. The plaintiffs sought as relief a declaration that the state statute was unconstitutional and an injunction ordering the Attorney General to notify the legislature and the public of the statute’s invalidity. Id. at 1294-95. In considering traceability, we concluded that the Attorney did not enforce the statute because it “envision[ed] no role for the Attorney General.” Id. at 1299. And in reviewing redressability, we reasoned that because the “Attorney General played no enforcement role whatsoever” with respect to the minimum wage law, a judgment against the Attorney General would not “directly redress” the plaintiffs’ injury. Id. at 1301-02 (internal quotation marks omitted).

84

The majority opinion argues that our reasoning in Lewis shows that the Secretary of State does not enforce the ballot-order statute. But this case is not Lewis because here the Secretary of State plays a substantial role in the statutory scheme at issue. Lewis does not help the majority in going further; once we concluded that the Alabama Attorney General had no role in enforcing the statute, we did not address the type of enforcement role a state official must have to satisfy traceability or redressability.

The majority opinion seeks to fill this silence by making new rules about the role a state official must have with respect to a challenged statute to establish traceability and redressability. But neither Supreme Court nor this Circuit’s precedent imposes such a heavy burden on plaintiffs challenging state laws. I note further that when confronted with cases in which defendant state officials carried out similar responsibilities with respect to challenged laws, our sister circuits have concluded that the officials were enforcing the law sufficiently to confer standing. See OCA-Greater Houston v. Texas, 867 F.3d 604 (5th Cir. 2017); Calzone v. Hawley, 866 F.3d 866 (8th Cir. 2017).

In a case strikingly similar to this one, the Fifth Circuit considered whether the plaintiff established traceability and redressability for standing purposes in a lawsuit against Texas’s Secretary of State. The court resolved both issues by concluding that the Secretary had a sufficient “enforcement connection with” a

85

challenged state statute regarding the administration of elections. 9 See OCA- Greater Houston, 867 F.3d at 613-14 (internal quotation marks omitted). When a citizen with a limited ability to communicate in English sought to have her son serve as an interpreter for her while she voted, local officials refused, citing a Texas statute allowing an interpreter to assist a voter only if the interpreter was registered to vote in the voter’s county of residence. Id. at 607-09. In a lawsuit against the Secretary of State challenging the Texas statute on federal preemption grounds, id. at 609, the Secretary argued that the voter’s injury was neither fairly traceable to him nor redressable in a lawsuit against him and instead was the result of actions by county officials who applied the statute to prohibit her son from serving as an interpreter. Id. at 612-13. The Fifth Circuit rejected his argument.

To determine whether the plaintiff’s injury was fairly traceable to the Secretary and redressable in litigation against the Secretary, the court considered whether under Texas law the Secretary had a role in enforcing the challenged statute. See id. at 613-14. The court explained that a state official had “no enforcement connection with the challenged statute” when he had no “duty or ability to do anything” with respect to the challenged law. Id. (emphasis omitted)

9 In an amicus brief filed in this case, Texas’s Attorney General emphasized the similarities in how Florida and Texas have chosen to administer elections. In both states local officials, who operate outside the Department of State and may not be removed from office by the Secretary of State, prepare ballots, while the Secretary of State is tasked with obtaining and maintaining uniformity in the application of the state’s election laws.

86

(internal quotation marks omitted). Because by law the Secretary of State was the “chief election officer of the state” and was “instructed by statute to obtain and maintain uniformity in the application, operation, and interpretation” of Texas’s election code, the court concluded, he had a sufficient “enforcement connection with the challenged statute” in the election code to establish traceability and redressability. Id. at 613-14 (internal quotation marks omitted).

Florida’s Secretary of State enjoys the same powers and responsibilities as the Texas Secretary. At the risk of beating a dead horse, I reiterate that she serves as Florida’s “chief election officer,” Fla. Stat. § 97.012; is instructed by statute to “obtain and maintain uniformity in the interpretation and implementation” of Florida’s Election Code, id. § 97.012(1); and is empowered by statute to promulgate rules to implement the statute in questions, id. § 101.151(9). The majority opinion’s holding that she lacks a sufficient enforcement connection with the ballot-order statute to satisfy traceability and redressability is directly contrary to the Fifth Circuit’s holding in OCA-Greater Houston.

The majority opinion’s determination that the Secretary of State does not enforce the ballot-order statute also is in tension with a decision from the Eighth Circuit, albeit one outside of the election context. The Eighth Circuit held that for the purpose of standing a state official played a sufficient role in enforcing a challenged statute when state law authorized her to promulgate rules and

87

regulations to implement the statute. See Calzone, 866 F.3d at 870. In Calzone, a truck driver sued, among others, the superintendent of Missouri’s state highway patrol after the driver was cited for refusing to permit a state highway patrol officer to inspect his vehicle under a Missouri law that authorized officers to stop any commercial vehicle to police its compliance with size and weight restrictions. Id. at 869. The driver challenged the state statute as unconstitutional. Id.

On appeal, the Eighth Circuit held that the driver had standing to sue the superintendent. Id. at 870. Although the superintendent was not involved in the stop or citation, the court held that the driver’s injuries were traceable to her and redressable in a lawsuit against her because the driver was stopped under the state statute that authorized patrol officers to stop commercial vehicles, and the superintendent had adopted rules and regulations to implement the statute. Id. (citing Mo. Rev. Stat. § 304.230.1). The court accepted that the superintendent’s adoption of rules and regulations led the officer to “implement the statute by conducting vehicle inspections,” which caused the driver’s injury. The driver’s injuries thus were traceable to her and redressable against her.10 Id. Under

10 The Eighth Circuit drew the conclusion that the superintendent’s promulgation of rules and regulations led the highway patrol officer to conduct the stop, even though another provision of the state statute separately authorized the officers to conduct suspicionless stops, meaning that the officer could have conducted the stop regardless of any rules or regulations the superintendent adopted. See Calzone, 866 F.3d at 870-71 (citing Mo. Rev. Stat. § 340.230.2). Despite the highway patrol officers’ independent statutory authorization to perform suspicionless stops, the court concluded that the driver’s injury bore a sufficient causal connection to the superintendent’s actions to be traceable to and redressable against her. See id. at 870.

88

Calzone’s reasoning, the fact that Florida law authorizes the Secretary of State to promulgate rules and regulations to implement the ballot-order statute is sufficient to establish that she enforces the law.

The majority opinion creates a circuit split on the connection a state official must have with a challenged state statute for a plaintiff to satisfy traceability and redressability. Although “intercircuit splits on points of law are not all bad,” we should have good reasons for creating one. Pub. Health Tr. of Dade Cty. v. Lake Aircraft, Inc., 992 F.2d 291, 295 n.4 (11th Cir. 1993) (“[W]e do not create intercircuit splits lightly.”). I remain unconvinced that there are good reasons here.

The majority opinion raises an additional argument regarding redressability:

that an injunction directed to the Secretary of State would not redress the plaintiffs’ injuries because the relief would not alter the conduct of the county election supervisors who print the ballots. According to the majority opinion, there is nothing to suggest that any relief directed to the Secretary would change how county election supervisors prepare their ballots. I disagree. If the district court directed the Secretary of State to cease providing ballot forms that list candidates from the governor’s party first for every office, it is likely that county election supervisors would follow the Secretary’s official guidance under the authority granted her by state law. True, I cannot definitely say that when faced with a conflict between the Secretary of State’s directions and the statute, the county

89

election supervisors would not reverse their customary course, stop relying on form ballots or lists from the Secretary, ignore the Department of State’s rules and directives, and follow the statute instead. Maybe in practice some would. But that does not make redress from the court’s order “speculative” as a legal matter.11 In reaching the opposite conclusion, the majority opinion splits from the Fourth Circuit, which—when faced with analogous facts—found a sufficient likelihood that local officials would follow the state official’s instructions, regardless of the statute. Bostic v. Schaefer, 760 F.3d 352, 370-71 (4th Cir. 2014). In Bostic, two same-sex couples brought a constitutional challenge to Virginia’s state statutes and state constitutional amendment that prohibited same-sex marriage. Id. at 367-68. The couples sued two defendants: the clerk of a circuit court, who had denied one of the couples a marriage license, and the state registrar for vital records, who was responsible for developing a marriage license application form and distributing it to circuit court clerks throughout Virginia. Id. at 369, 371.

11 To bolster its opinion, the majority opinion cites to the trial testimony of a single former county election supervisor that he applied the ballot statute because it is the law. The former election supervisor gave this testimony while recounting that voters occasionally asked him why Republican candidates appeared at the top of their ballots. He would respond to the voter that the order was set by the statute and he was applying the statute. No testimony at trial addressed what this or any other county election supervisor would do if the ballot order from the Secretary of State did not follow the statute. Regardless, such testimony is irrelevant to the legal question of whether state law gives the Secretary enforcement authority over county election supervisors sufficient to meet the standards for traceability and redressability.

90

On appeal, the Fourth Circuit held that the couples had standing to sue the state registrar for vital records because the registrar’s “promulgation of a marriage license application form that does not allow same-sex couples to obtain marriage licenses” resulted in the couples being denied marriage licenses. 12 Id. at 371-72. The Fourth Circuit determined that the registrar was “enforc[ing]” Virginia’s marriage laws by developing and circulating license forms that did not allow same- sex couples to obtain marriage licenses. Id. at 372. The court explained that the registrar’s actions “resulted in” local officials denying marriage license requests from same-sex couples, id.—despite the fact that if the registrar had distributed forms that permitted same sex-couples to apply for marriage licenses, Virginia law still would have prohibited local officials from issuing marriage licenses to same- sex couples, see id. at 367-68 (cataloguing Virginia laws prohibiting same-sex marriage).

Applying the majority opinion’s logic, even if a federal court ordered the Virginia registrar to cease issuing marriage application forms that barred same sex marriage, a local clerk who issued marriage licenses could have refused to issue a same-sex couple a marriage license on the ground that Virginia law continued to

12 The Fourth Circuit also concluded that the plaintiffs had standing to sue the clerk who had denied one of the couples a marriage license. But the court made clear that the inclusion of the clerk as a defendant did not establish that the couples had standing to sue the registrar, because the standing requirements had to be satisfied as “to each defendant.” Bostic, 760 F.3d at 370-71.

91

bar same-sex marriages. See id. at 368. But the Fourth Circuit did not see it that way. To state the obvious, the Secretary of State plays a similar role in issuing sample ballots and lists of candidates as the Virginia registrar did in issuing marriage license forms.

To wrap up, there are strong reasons to doubt the majority opinion’s interpretation of Florida law and resulting conclusions about the role the Secretary of State plays in implementing Florida’s ballot-order statute. If the majority opinion is wrong about Florida law—meaning that in fulfilling her duties the Secretary of State enforces the ballot-order statute—then in my view she has a sufficient connection to the statutory scheme to satisfy traceability or redressability. The majority opinion’s argument that a state official who performs these duties does not “enforce” the law lacks any support in our precedent and conflicts with decisions from several other circuits.

II.

Because it is far from clear that Florida law limits the Secretary of State’s authority in the way the majority outlines, I believe that principles of judicial restraint direct us to abstain from issuing alternative holdings on redressability or traceability and to decide this case solely on the basis that the plaintiffs failed to prove an injury in fact. The “cardinal principle of judicial restraint” is “if it is not necessary to decide more, it is necessary not to decide more.” PDK Labs. Inc. v.

92

Drug Enf’t Admin., 362 F.3d 786, 799 (D.C. Cir. 2004) (Roberts, J., concurring). Of course, we do not always adhere to this prudential principle. At times principles of judicial economy or other considerations may guide us to resolve an appeal by giving multiple, alternative holdings. See Evans v. Ga. Reg’l Hosp., 850 F.3d 1248, 1255-56 (11th Cir. 2017); see also Karsten v. Kaiser Found. Health Plan of Mid-Atl. States, Inc., 36 F.3d 8, 11 (4th Cir. 1994).

But just because we may offer alternative holdings does not mean that we should. The Supreme Court has cautioned that when a court can “readily” dispose of a case on one threshold ground, it should refrain from offering alternative holdings resting on other threshold grounds that are “difficult to determine.” Sinochem Int’l Co. v. Malay. Int’l Shipping Corp., 549 U.S. 422, 436 (2007). Putting its advice into practice, in Lujan, a majority of the Court declined to resolve the issue of standing on different grounds through alternative holdings. In Lujan, Justice Scalia authored an opinion that concluded in Part III-A that the plaintiffs failed to establish an injury fact and in Part III-B that the plaintiffs failed to establish redressability. 504 U.S. at 561-71. Although Justice Scalia delivered the opinion of the Court in Part III-A, Part III-B was a plurality opinion in which only three justices joined. See id. at 556. In a separate concurrence, Justice Kennedy, joined by Justice Souter, explained that he joined Part III-A but not Part III-B of Justice Scalia’s opinion because “[i]n light of the conclusion that

93

respondents have not demonstrated a concrete injury here sufficient to support standing,” there was no need to reach redressability. Id. at 579-80 (Kennedy, J., concurring in part and concurring in the judgment). I would follow the Justices’ lead in Lujan and resolve this case on only one standing ground.

This principle applies with even greater force when the alternative holding rests on difficult questions of state law. State courts are “the ultimate expositors of state law,” Mullaney v. Wilbur, 421 U.S. 684, 691 (1975), and so “it is axiomatic that [a state’s supreme court] is the best one to decide issues” of that state’s law.13 Blue Cross & Blue Shield of Ala., Inc. v. Nielsen, 116 F.3d 1406, 1413 (11th Cir. 1997). When we decide questions of state law, unique federalism concerns arise. We risk creating “needless friction with state policies,” trampling on the “rightful independence of the state governments,” and upsetting “harmonious relation[s] between state and federal authority.” Quackenbush v. Allstate Ins. Co., 517 U.S. 706, 717-18 (1996) (internal quotation marks omitted); see Knick v. Township of Scott, 139 S. Ct. 2162, 2188 (2019) (Kagan., J, dissenting) (discussing the Supreme Court’s long-held understanding that “federal courts should refrain whenever possible from deciding novel or difficult state-law questions”). We therefore have

13 We have characterized our decisions interpreting state law as written “in faint and disappearing ink.” LeFrere v. Quezada, 582 F.3d 1260, 1262 (11th Cir. 2009) (internal quotation marks omitted). “[O]nce the state supreme court speaks[,] the effect of anything we have written vanishes like the proverbial bat in daylight, only faster.” Id. (internal quotation marks omitted).

94

recognized that principles of “[f]ederalism and comity” require us to “defer[] to state courts on ultrasensitive state law matters.” Nielsen, 116 F.3d at 1413.

Federalism concerns that counsel us to abstain from deciding such state law issues are even more compelling when the state law issue “sound[s] to the heart of a state’s self-government.” Id. After all, it is “[t]hrough the structure of its government, and the character of those who exercise government authority, [that] a State defines itself as a sovereign.” Gregory v. Ashcroft, 501 U.S. 452, 460 (1991). When a federal court decides novel and difficult state law questions about how the state government is organized, the federal court may fail to afford due respect to the state’s sovereignty.

I recognize that we retain broad discretion to decide whether to resolve a case through alternative holdings. And whether we choose to exercise our discretion in a particular case depends, of course, on the circumstances. Given the circumstances here, I view this case as a textbook example of when principles of judicial restraint caution us against making alternative holdings.

In this case, there are three potential threshold grounds on which we can resolve this appeal. One of the grounds—that the plaintiffs failed to establish an injury in fact—turns on a relatively straightforward question of federal law: Did the plaintiffs present evidence that any plaintiff suffered an injury in fact sufficient to establish standing under Supreme Court and Eleventh Circuit precedent? In

95

contrast, the majority opinion’s alternative holdings that the plaintiffs failed to prove their injuries were traceable to the Secretary of State and redressable in a lawsuit against her require us to sort out the complex interplay of state statutes and delve into the division of authority among state and local officials, issues that go to the very structure of state government. And in addressing these issues, the majority opinion lays out a theory delineating the authority of a high-ranking state official that the official never advanced in this case.

Here, I believe that the appropriate course is to abstain from alternative holdings on traceability and redressability. By deciding novel questions of state law about the structure and organization of Florida’s state government, I fear that the majority opinion fails to demonstrate proper respect for state sovereignty and risks creating needless friction with state authority. See Gregory, 501 U.S. at 460. On top of that, the majority opinion applies new standards for traceability and redressability that impose heavier burdens than Supreme Court and Circuit precedent requires, all while creating a circuit split. I respectfully dissent from Part III-B of the majority opinion.