Before CLEMENT, HO, and DUNCAN, Circuit Judges. EDITH BROWN CLEMENT, Circuit Judge: Appellant Diane Haddock sued the seven district judges of Tarrant County’s family law courts (the “District Judges”) in their official capacities, District Judge Patricia Baca-Bennett in her personal capacity, and the County under 42 U.S.C. § 1983, alleging that she was fired for refusing to support a political candidate and for her husband’s political activity. Holding that Haddock was both a policymaking and confidential employee lawfully subject to patronage termination, the district court dismissed her suit. We AFFIRM. FACTS AND PROCEEDINGS Tarrant County family courts are presided over by seven elected district judges, who, in turn, are assisted by seven appointed associate judges. Haddock was an associate judge for nearly twenty years. Because they serve more than one district judge, Texas law requires Tarrant County associate judges be appointed with the unanimous approval of the district judges; they can be removed, however, by a majority vote. TEX. FAM. CODE §§ 201.001(d), 204(b). In 2016, Haddock and fellow associate judge James Munford indicated interest in running for a district judge position. It was believed they would run against one another for the 322nd district seat. Around the same time, the grandparents of a child who died while in her mother’s custody—after Haddock had signed the order giving the mother custody—circulated claims that Haddock had mishandled the case, going so far as to allege that she had taken a bribe.[1] Munford’s wife allegedly repeated these harsh allegations publicly, presumably to gain political advantage for her husband. Haddock decided not to run, but she and her husband do not appear to have reconciled with Munford and his wife. During the campaign, although Haddock herself allegedly did not engage in any overt political activity, her husband campaigned against Munford. Mr. Haddock and a political group with which he was associated accused Munford of being a “RINO” (Republican In Name Only), violating the Second Amendment by signing protective orders requiring litigants to surrender their firearms on inadequate evidence, physically abusing and sexually assaulting his first wife, and terrifying his current wife by threatening her and a male friend of hers with a gun. District Judge Patricia Baca-Bennett, who supported Munford’s candidacy, allegedly sought to put a stop to Mr. Haddock’s opposition by demanding that Haddock publicly support Munford and “get her husband under control.” Haddock refused to do either. Baca-Bennett allegedly subjected Haddock to “badgering, threats, back-biting, undermining and maligning, and a campaign to orchestrate the termination of [Haddock's] employment.” She also allegedly sought to intimidate Haddock’s husband by reminding him “who Diane works for” and spread rumors about Haddock resigning that “undermined [Haddock's] authority as a judge.” During the campaign, Haddock also learned that the district judge for her own District 233 was retiring. Kenneth Newell won the Republican primary (he then ran unopposed, meaning he knew then that he would become District 233′s district judge), so he spoke with Haddock about her future as the District 233 associate judge. He indicated that he was concerned about the political situation and had “not made a decision about what to do with” Haddock. Following unsuccessful complaints to Tarrant County’s human resources department, Haddock eventually sued Baca-Bennett and Tarrant County for subjecting her to a hostile work environment in retaliation for her husband’s political activity and her own refusal to support Munford. Fewer than ninety days later, she was terminated by a majority of the seven district judges, including Newell. She amended her complaint to address her termination, add the District Judges in their official capacities as defendants, and demand reinstatement or front pay in lieu thereof. The district court dismissed Haddock’s claims for money damages against the District Judges in their official capacity under Rule 12(b)(1), holding that the suit is barred by the Eleventh Amendment because the District Judges are state officials, meaning “the state was the real, substantial party in interest,” and the state has not waived sovereign immunity. See Va. Off. for Prot. & Advoc. v. Stewart, 563 U.S. 247, 255 (2011) (cleaned up). Haddock does not appeal this ruling. The district court also dismissed Haddock’s claim for injunctive relief against the District Judges under Rule 12(b)(6). The First Amendment generally prohibits adverse employment actions against government employees based on political affiliation, Elrod v. Burns, 427 U.S. 347, 373 (1976), but, where “an employee’s private political beliefs would interfere with the discharge of [her] public duties, [her] First Amendment rights may be required to yield to the State’s vital interest in maintaining governmental effectiveness and efficiency,” Branti v. Finkel, 445 U.S. 507, 517 (1980). Sometimes called the Elrod/Branti exception, this maxim most often applies to employees in policymaking or confidential positions. Finding that Haddock’s position involved both policymaking and confidential relationships with the District Judges and, “[t]herefore, an associate judge’s political ideology, associations, and activities may rationally influence a district judge’s assessment of the individual’s suitability for a position as an associate judge,” the district court held that she had failed to state a claim on which relief could be granted against the District Judges and dismissed Haddock’s demands for injunctive relief under Rule 12(b)(6). Haddock v. Tarrant Cnty., No. 4:18-cv-00817-O, 2019 WL 7944073, at *7–8 (N.D. Tex. Sept. 11, 2019). The district court dismissed all claims against Tarrant County under Rule 12(b)(6), both because Haddock had failed to allege an underlying constitutional violation and because she had failed to allege a county policy or policymaker that caused the alleged violation. Finally, the district court dismissed all claims against Baca-Bennett under Rule 12(b)(6) on the basis of qualified immunity. Haddock timely appealed. STANDARD OF REVIEW We review a dismissal on the pleadings under Rules 12(b)(1) or 12(b)(6) de novo, “accepting all well-pleaded facts as true and viewing those facts in the light most favorable to the plaintiffs.” Wolcott v. Sebelius, 635 F.3d 757, 762–63 (5th Cir. 2011) (citation omitted). “Generally, a court ruling on a 12(b)(6) motion may rely on the complaint, its proper attachments, documents incorporated into the complaint by reference, and matters of which a court may take judicial notice.” Id. at 763 (cleaned up). III. DISCUSSION A. Haddock argues on appeal that the district court erred in applying the Elrod/Branti exception to her First Amendment claims because she claims that she is neither a policymaker nor a confidential employee. She also argues that her intimate association claim (allegedly, Baca-Bennett retaliated against Haddock for her husband’s speech, not her own) is—categorically—not subject to the Elrod/Branti exception. We disagree. Haddock also argues that the Supreme Court’s balancing test in Pickering v. Board of Education, 391 U.S. 563 (1968), would be more appropriate than an Elrod/Branti analysis. We need not analyze this argument in any great depth; where the Government’s interest in political loyalty is weighed against an employee’s First Amendment interests, the tests frequently merge. See Maldonado v. Rodriguez, 932 F.3d 388, 392 (5th Cir. 2019) (“This court’s decisions have melded the Supreme Court’s discussion of these principles in Branti v. Finkel with the broader but similar Pickering–Connick test.”). Generally speaking—and applicable here—if the Elrod/Branti exception applies, the Pickering analysis is also concluded. We also note that the test, strictly speaking, is not about whether an employer is a policymaker or confidential employee. “[R]ather, the question is whether the hiring authority can demonstrate that party affiliation is an appropriate requirement for the effective performance of the public office involved.” Branti, 445 U.S. at 518. That said, “where a public employee . . . occupies a confidential or policymaking role, the employer’s interests more easily outweigh the employee’s First Amendment rights.” Maldonado, 932 F.3d at 392 (alteration in original) (quoting Gentry v. Lowndes Cnty., 337 F.3d 481, 486 (5th Cir. 2003)). (1) Haddock’s pleadings, combined with Texas law, make clear that she is a policymaker subject to the Elrod/Branti exception, and political affiliation is relevant to her qualification for the associate judge position. The reason the Elrod/Branti exception typically applies to policymakers is that such employees are uniquely positioned to frustrate the policy agendas of the elected officials for whom they work. As our colleagues on the Seventh Circuit have explained, “it would undermine the democratic process to hold that the winners at the polls may not employ those committed to implementing their political agenda.” Kurowski v. Krajewski, 848 F.2d 767, 770 (7th Cir. 1988). “Policymakers are ‘public employees whose responsibilities require more than simple ministerial competence, whose decisions create or implement policy, and whose discretion in performing duties or in selecting duties to perform is not severely limited by statute, regulation, or policy determinations made by supervisors.’” Garza v. Escobar, 972 F.3d 721, 729 (5th Cir. 2020) (quoting Aucoin v. Haney, 306 F.3d 268, 273 (5th Cir. 2002)). “An employee with responsibilities that are not well defined or are of broad scope more likely functions in a policymaking position.” Stegmaier v. Trammell, 597 F.2d 1027, 1033 (5th Cir. 1979). Haddock argues that judges, categorically, cannot be policymakers because they merely apply the law to the facts of a case. Although we appreciate this aspirational view of the judiciary generally, both the structure of the judiciary in Texas and Haddock’s pleadings refute this argument. Haddock relies heavily on a case recently reversed by the Supreme Court, in which the Third Circuit held that “a judicial officer, whether appointed or elected, is not a policymaker.” See Adams v. Governor of Del., 922 F.3d 166 (3d Cir. 2019), rev’d sub. nom. Carney v. Adams, 141 S. Ct. 493 (2020) (reversing on standing grounds without comment on whether judges are policymakers). Adams is unpersuasive for reasons beyond its reversal. First, the categorical pronouncement was mere dicta; the context of the case was Delaware’s constitutional structure, which required consideration of political party when appointing judges. This structure itself, the Third Circuit reasoned, demonstrated “that political loyalty is not an appropriate job requirement for Delaware judges” because it required the Governor to occasionally “nominate judges who belong to a different political party.” Id. at 179. In contrast, the Texas constitution leaves the selection of judges to the electorate, with no requirement or expectation that voters ever knowingly select a judge with whom they disagree. Second, we are guided by the unanimous opinion of our colleagues on other Circuits that judicial officers can be (and often are) policymakers. See, e.g., Mumford v. Basinski, 105 F.3d 264, 272 (6th Cir. 1997) (family law referee’s “political ideology, associations, and activities may rationally influence a judge’s assessment of an individual’s suitability for a position as his referee”); Kurowski, 848 F.2d at 770 (“A judge both makes and implements governmental policy. A judge may be suspicious of police or sympathetic to them, stern or lenient in sentencing, and political debates rage about such questions.”); cf. Hawkins v. Steingut, 829 F.2d 317, 318 (2d Cir. 1987) (granting qualified immunity for dismissal of Workers’ Compensation referee “referred to by the Board as ‘Workers’ Compensation Law judges’”). Particularly where, as here, judges are elected based on both personal and political qualifications, we see no reason why they or their appointees should be categorically excluded as policymakers. In Texas, as “[i]n most states[,] judges are elected, implying that the office has a political component.” Kurowski, 848 F.2d at 770. Finally, the specific facts of this case illustrate that the associate judge position was a policymaking role. The Sixth Circuit’s opinion in Mumford is particularly illuminating. Mumford was a Domestic Relations Court referee, with authority to “conduct [ ] hearings on the matters referred to him, [ ] issu[e] [ ] subpoenas, [ ] swear[ ] and examin[e] [ ] witnesses, . . . promulgat[e] [ ] evidentiary rulings and . . . [enter] certain pretrial, discovery, temporary restraining, and other orders necessary to regulate the proceedings, all without judicial ratification.” Mumford, 105 F.3d at 272. Similarly, once a case is referred to a Tarrant County associate judge, they can hear “any aspect of a suit over which the court has jurisdiction . . . including any matter ancillary to the suit.”[2] TEX. FAM. CODE § 201.005(a). Most of an associate judge’s decisions are subject to de novo review by the presiding district judge, but associate judges can also issue final orders in cases in which the parties have waived the right to a de novo hearing. TEX. FAM. CODE § 201.007. Even where a party requests de novo review, an associate judge’s orders remain in full effect unless and until they are reversed. TEX. FAM. CODE § 201.013. Like the referees in Mumford, Tarrant County associate judges “effectively make[ ] policy for, or suggest[ ] policy to, the court on each occasion that [they] resolve[ ] a dispute in the court’s name or recommend[ ] a disposition to a judge.” 105 F.3d at 272. There can be no question that Haddock was entrusted with the type of broad discretion that paradigmatically characterizes a policymaker. More crucially, Haddock’s complaint shows that the policymaking functions of an associate judge were directly relevant during judicial elections. Munford’s performance as an associate judge—including degree of party fealty (whether he was a “RINO”) and attitude toward political hot- button topics like gun rights—were key campaign issues. Haddock, by her own allegations, was fired at least in part (if not entirely) because of her husband’s speech on those specific topics. Haddock herself had planned to run for a district judgeship until controversy over her own decision-making as an associate judge led her to drop out of the race. As the Sixth Circuit explained, “judges are policymakers because their political beliefs influence and dictate their decisions on important jurisprudential matters.” Newman v. Voinovich, 986 F.2d 159, 163 (6th Cir. 1993). Judicial temperament (for example, willingness to issue protective orders) is directly relevant to the job of Tarrant County family court associate judges and is an important aspect of the political qualifications—and electoral fortunes—of the district judges they represent. Haddock herself notes the importance of associate judges understanding and respecting what she terms district judge’s “preferences.” For example, one “district judge will nearly always order a batterer’s intervention course. Another will almost never order a social study in child custody cases.” The voters of Tarrant County should not have to wonder whether the district judges they elect will be able to carry out the will of the electorate without constant oversight of their associate judges. Instead, district judges are entitled to select associate judges they trust to carry out their policy preferences. Haddock was a policymaker, so, to the extent that her claims are premised on perceived political disloyalty—whether because she refused to support Munford, was believed to agree with her husband’s anti-Munford advocacy, or for whatever other reason—her termination was constitutional under the Elrod/Branti doctrine. (2) Haddock was also a confidential employee. “A government employee may be ‘confidential’ ‘if he or she stands in a confidential relationship to the policymaking process, e.g., as an advisor to a policymaker, or if he or she has access to confidential documents or other materials that embody policymaking deliberations and determinations, e.g., as a private secretary to a policymaker.’” Garza, 972 F.3d at 729 (quoting Maldonado, 932 F.3d at 393). If a superior official would be unable to carry out her duties as efficiently or to delegate sensitive tasks when she did not feel she could trust an employee to keep her confidences, that is likely a confidential employee. Associate judges are “privy to confidential”—and, given the nature of family law matters, often extremely sensitive—”litigation materials and internal court communications in the discharge of [their] duties, and further maintain[ ] a personal confidential relationship with the judge(s) which [they] serve[ ].” Mumford, 105 F.3d at 272. Whether in private conversation with district judges or in writing when they “resolve[ ] a dispute in the court’s name or recommend[ ] a disposition to a judge,” the associate judges serve as advisors and confidants to the district judges, aiding them in the execution of their duties. Id. Haddock argues that she cannot be a confidential employee because seven associate judges working for seven district judges results in “forty-nine independently developing working relationships”—too many relationships, she argues, to implicate the sort of close, personal relationships characteristic of confidential employees. First, Haddock’s math is misguided—this case has nothing to do with her relationships with the other associate judges. Only seven working relationships are relevant—between Haddock and her superiors, the district judges. We suspect all of our twenty-five colleagues on this court would agree that judges can reasonably be expected to maintain at least seven close, yet professional working relationships. Second, this numerical argument is firmly foreclosed by precedent. See, e.g., Gentry, 337 F.3d at 486 (“[I]f a public employee’s loyalty is owed to a [five-]member governing board, he cannot choose political favorites or enemies among the board because shifting coalitions or electoral victories may too easily render the employee’s decisions, made in accord with personal preference, at odds with the board majority view.”); Kinsey v. Salado Indep. Sch. Dist., 950 F.2d 988, 996 (5th Cir. 1992) (en banc) (school superintendent’s loyalty may be required by a seven-member school board). Further, Haddock’s pled facts—which at this stage, we must presume to be true—make clear that the associate judges and district judges developed close, personal relationships that involved the exchange of confidences, including on politically sensitive and policy-oriented topics. Haddock discussed electoral politics and her own prospective campaign with District Judge William Harris—her supervising District 233 judge prior to Newell’s election. She ultimately decided not to run for office based, in part, on his advice. We also know that Newell replaced Haddock with a close associate (the friend who “emceed” his investiture). Our colleagues on the Seventh Circuit note that, where personal interactions are an important part of the work environment, “[p]olitical animosity . . . can in practice create a hostile work environment where face to face contact and cooperation are essential,” in some cases harming the efficiency of the office. See Meeks v. Grimes, 779 F.2d 417, 423 (7th Cir. 1985). This is precisely what happened here. Haddock alleges that she accused Baca-Bennett of unethical judicial conduct—specifically, “violat[ing] the canons governing active judges”—by openly campaigning for Munford. The Haddocks and Munfords lobbed vitriolic campaign rhetoric at each other that might have made the Hatfields and McCoys blush—the allegations ranged from sexual assault and other domestic violence to taking bribes and leaving a child to die in an unsafe home. Although Haddock alleged that “all seven associate judges serve all seven district judges,” it’s difficult to imagine a healthy working relationship between Haddock and at least two of the judges, which, all else being equal, makes her a less effective employee than an associate judge who can work amicably with all seven. Haddock also alleges that Baca-Bennett’s role in the dispute “undermine[d] respect for [Haddock's] judicial authority,” which presumably impacted Haddock’s effectiveness on the bench, even when serving the remaining five judges. Ultimately, although Haddock alleges she believed Newell otherwise wished to retain her, she was left with the impression that he felt “she would be difficult to keep despite her qualifications due to the political situation.” In short, the political dispute disrupted Tarrant County family court operations, caused several of the elected district judges to lose faith in Haddock’s ability to do her job, impeded Haddock’s ability to assert her authority in court, and compromised her trustworthiness as an employee in the eyes of at least two of the seven district judges she was duty-bound to serve. The Elrod/Branti exception is not about labels like “policymaker” or “confidential,” but about preventing precisely this type of disruption. (3) Finally, Haddock argues that some of the specific First Amendment rights upon which she bases her claims cannot be subject to Elrod/Branti analysis. Specifically, she argues that Elrod/Branti may apply to reprisals for an employee who actively campaigns against her superior, but—because the speech at issue was her husband’s, not her own (she, allegedly, refused to campaign for or against anyone)—she is being punished for her association with her spouse and for refusing to campaign. In other words, Haddock argues that the First Amendment rights of intimate association and freedom from compelled speech should not be subject to the Elrod/Branti exception. Our precedent firmly establishes that Elrod/Branti applies to refusal to speak. See, e.g., Stegmaier, 597 F.2d at 1030, 1040 (holding confidential employee could be discharged for failing to support elected officeholder’s candidacy under Elrod). A policymaker who refuses to endorse a winning candidate may be discharged as readily as one who endorses a loser. We also join the unanimous opinion of our sister Circuits in holding that intimate association claims can be subjected to Elrod/Branti analysis. See, e.g., Simasko v. Cnty. of St. Clair, 417 F.3d 559 (6th Cir. 2005); McCabe v. Sharrett, 12 F.3d 1558, 1572 (11th Cir. 1994); Soderbeck v. Burnett Cnty., 752 F.2d 285 (7th Cir. 1985) (Posner, J.). There may be reason to doubt the effectiveness of either policymaking or confidential employees when they are intimately associated with an elected official’s political opponents. Haddock refused to endorse Munford and indicated that she would take no action to curtail her husband’s campaigning. Her husband spent (or was believed by Baca-Bennett to have spent) between $30,000 and $300,000 campaigning against Munford. Haddock’s husband appears to have campaigned against Munford, at least in part, as a form of retaliation for Munford’s wife’s campaign against Haddock. When a policymaker refuses to endorse a candidate, her spouse spends or is believed to have spent a large sum of money opposing the candidate, and there is reason to believe the policymaker shares her spouse’s animosity based on personal history, it is reasonable for an elected official to doubt the policymaker’s political loyalty. See Soderbeck, 752 F.2d at 288 (“Mrs. Soderbeck was the political enemy of her husband’s political enemy.”). As a policymaker, Haddock could be terminated, under these circumstances, for her husband’s political activity because the District Judges had reason to doubt that she was committed to their policy agendas or judicial philosophies—that is, the agendas and philosophies chosen by the voters. The case is even stronger that a confidential employee may be discharged for intimate associations that cause an elected official to question the employee’s loyalty. In McCabe, the Eleventh Circuit held that an elected police chief could demote his confidential secretary to a non-confidential position because she was married to one of his officers. McCabe did not involve any allegations that the plaintiff had campaigned against the new police chief or had ever violated his trust. To the contrary, “[e]vidence produced by both parties demonstrate[d]” that the plaintiff “ actually breached no confidences during the brief period she served as” the defendant’s secretary, there was no reason to believe she had ever breached the prior chief’s confidences, and the odds her ever doing so “may not have been overwhelming.” McCabe, 12 F.3d at 1572–73 & n.17. Nonetheless, her job required her to have access to the chief’s confidential communications, including communications about personnel complaints and officer discipline. If there were a complaint against her husband or one of his colleagues, she would see it first. The McCabe court reasoned that “[i]t is a matter of common experience that spouses tend to possess a higher degree of loyalty to their marital partners than to their superiors, and often discuss workplace matters with one another, even matters that a superior has designated as confidential.” Id. at 1572. The elected official was uncomfortable “having the wife of an officer under [his] command function[ ] as [his] confidential Executive Secretary,” for fear (based on nothing more than the fact of her marriage to her husband) that her loyalty would be elsewhere, so he was constitutionally permitted to demote her. Id. Similarly, here, as a matter of common experience and the loyalty that spouses (hopefully) feel toward one another, there is reason to believe that Haddock’s loyalty would be to her husband first and to the District Judges second. So long as this created no conflict, it was fine; when Haddock’s husband became several judges’ fierce political enemy, it became a problem. Consider, for example, the campaign allegation that Munford did not adequately respect gun rights. Assume, hypothetically, that it’s true. Judges have a great deal of discretion with respect to protective orders. The voters chose Munford—and his judicial preferences. If, however, Munford wished to circulate a memo to the associate judges indicating his preference that, when he delegates a case to them, they exercise their discretion broadly in favor of protective orders requiring litigants to surrender their firearms, he would have to ask himself first whether he wanted to risk the memo ending up in a campaign ad against him during the next election cycle. He would have to consider that one of the associate judges was married to his political enemy, and any preferences he expressed, in confidence, might be repeated to someone who was looking for ammunition to use against him in the next election. A reasonable person in Munford’s position would question whether he could confidentially discuss, develop, or express policy, philosophy, or jurisprudential preferences to Haddock without undue personal risk. The District Judges—Baca-Bennett and Munford especially—had reason to doubt that they could trust Haddock with confidential policy- related materials or conversations. They had reason to doubt that she agreed with their policy preferences, because her husband had campaigned against Munford, in part, on policy grounds, and she had refused to attempt to curtail his campaigning or take a position herself. That Haddock alleges she had not violated any confidences or knowingly gone against any district judge’s policy preferences is of no moment “because we do not require employers to wait until their office is disrupted before taking action.” Garza, 972 F.3d at 732. Haddock was in a policymaking and confidential role, and, under the Elrod/Branti exception, could constitutionally be discharged for the exercise of rights that would otherwise by protected by the First Amendment. B. Haddock alleges that the district court erred by dismissing her claims against Tarrant County. Although Tarrant County, as a municipal entity, can be held liable under § 1983 when an “action pursuant to official municipal policy of some nature caused a constitutional tort,” it “cannot be held liable under § 1983 on a respondeat superior theory.” Monell v. Dep’t of Soc. Servs., 436 U.S. 658, 691 (1978). For municipal liability to attach, a plaintiff must prove “three elements: a policymaker; an official policy; and a violation of constitutional rights whose moving force is the policy or custom.” Zarnow v. City of Wichita Falls, 614 F.3d 161, 166 (5th Cir. 2010) (quoting Piotrowski v. City of Hous., 237 F.3d 567, 578 (5th Cir. 2001)). As explained above, because the Elrod/Branti exception applies to Haddock’s claims, she has failed to plead a constitutional violation. We therefore do not need to examine whether she has pled a county policymaker or official policy. The district court correctly dismissed Haddock’s claims against Tarrant County. C. Haddock also takes issue with the district court’s holding that Baca- Bennett has qualified immunity. “Qualified immunity shields federal and state officials from money damages unless a plaintiff pleads facts showing (1) that the official violated a statutory or constitutional right, and (2) that the right was clearly established at the time of the challenged conduct.” Ashcroft v. al-Kidd, 563 U.S. 731, 735 (2011) (cleaned up). These questions can be answered in either order. Pearson v. Callahan, 555 U.S. 223, 242 (2009). As explained above, Baca-Bennett did not violate Haddock’s constitutional rights; this is enough for Baca-Bennett to be entitled to qualified immunity. Even if Haddock’s rights had been violated, however, Baca-Bennett certainly did not have “fair warning that [her] conduct violate[d] a constitutional right.” Clarkston v. White, 943 F.3d 988, 993 (5th Cir. 2019) (quoting Delaughter v. Woodall, 909 F.3d 130, 140 (5th Cir. 2018)). Closely on-point authority from our sister Circuits indicated that the Elrod/Branti exception applies to positions very much like Haddock’s. See, e.g., Mumford, 105 F.3d 264. The case that Haddock primarily relies on for the proposition that judges are categorically not policymakers was (1) decided in another Circuit (2) after Haddock’s termination and (3) was reversed by the Supreme Court. See Adams, 141 S. Ct. 493. Baca-Bennett is entitled to qualified immunity. CONCLUSION The district court correctly held that Haddock, as both a policymaker and a confidential employee, was subject to the Elrod/Branti exception, and had therefore failed to allege a constitutional violation. AFFIRMED. JAMES C. HO, Circuit Judge, concurring in the judgment: I concur in the judgment and agree with much of what Judge Clement writes in her typically thoughtful opinion. I write separately to make just one observation. As Judge Clement explains, the plaintiff in this case should be afforded the same constitutional status as those that our court and other courts have previously regarded as “confidential employees” under the First Amendment. See, e.g., Garza v. Escobar, 972 F.3d 721, 731 (5th Cir. 2020) (holding that a Crime Victims Unit Coordinator was a confidential employee); Gentry v. Lowndes Cnty., 337 F.3d 481, 488 (5th Cir. 2003) (holding that a road manager and county administrator occupied confidential positions because the county board of supervisors “must be assured of the trust and loyalty of the road manager and administrator and must be able to assume the confidentiality, when necessary, of their mutual dealings”); Kinsey v. Salado Indep. Sch. Dist., 950 F.2d 988, 996 (5th Cir. 1992) (holding that a school superintendent “occupied a confidential relationship” with the school board because he was the custodian of the school’s confidential records and advised the board on confidential matters); Soderstrum v. Town of Grand Isle, 925 F.2d 135, 141 (5th Cir. 1991) (holding that a police chief’s secretary was a confidential employee); Stegmaier v. Trammell, 597 F.2d 1027, 1040 (5th Cir. 1979) (holding that a deputy circuit clerk was a confidential employee). See also, e.g., Mumford v. Basinski, 105 F.3d 264, 272 (6th Cir. 1997) (“Unquestionably, the inherent duties of an Ohio domestic relations court referee entail a relationship of confidence between the referee and the judge(s) which he serves.”). It is on that basis that I would affirm. Accordingly, I concur in the judgment.
