dissenting in part.
As in Republican Party of Minnesota v. White, 536 U.S. 765, 122 S.Ct. 2528, 153 L.Ed.2d 694 (2002) (White I), this case can and should be resolved in favor of the judicial candidate on the principal disputed issues without deciding whether “the First Amendment requires campaigns for judicial office to sound the same as those for legislative office.” Id. at 783, 122 S.Ct. 2528. Even assuming the Supreme Court declines to adopt Justice Kennedy’s view that “content-based restrictions that do not fall within any traditional exception should be invalidated without inquiry into narrow tailoring or compelling government interests,” id. at 793, 122 S.Ct. 2528 (Kennedy, J., concurring), the endorsement clause and the personal solicitation clause of the Minnesota Code of Judicial Conduct fail to withstand strict scrutiny.
The endorsement clause, 52 Minn. Stat. Ann., Code of Judicial Conduct, Rule 4.1(A)(3), eliminates one useful way for a judicial candidate to associate with other candidates for office and to communicate to the voters his or her outlook on issues of the day. The appellees’ principal justification for the ban on endorsements is that it furthers a compelling interest in maintaining impartiality toward parties appearing before the judicial candidate if he is elected. But the rule is barely tailored to achieve this objective, and when an endorsement does frustrate the asserted state interest, the problem almost always can be remedied by recusal. This prohibition on political speech is thus not narrowly tailored to further a compelling government interest.
Despite the endorsement clause, a judicial candidate remains free to make positive statements about another candidate for office. At oral argument, counsel for appellees explained, by way of example, that the endorsement clause prohibits the following language: “I endorse Michelle Bachmann for the U.S. Congress.” Oral Argument at 40:14-40:24 (Jan. 10, 2011), available at http://www.ca8.uscourts.gov/ cgi-bin/new/cmECFArg.pl?case_num=091578. Judicial Candidate X, however, may announce that “I agree with Senator Y on all of the issues of the day.” Judicial Candidate X may declare that “Senator Y has done an excellent job, and I am going to vote for Senator Y.” Id. at 41:10-42:40, 46:00-46:07. And Judicial Candidate X may speak directly in favor of reelecting Senator Y as long as the speech comes before Senator Y officially becomes a “candidate” for reelection. Contrary to the plurality’s suggestion, ante, at 1027, these statements are not forbidden by Rule 4.1(A)(10), which applies only to statements that would impair “the fairness of a matter or impending matter.” 52 Minn. Stat. Ann., Code of Judicial Conduct, Rule 4.1(A)(10). Rule 4.1(A)(10) restrains the judicial candidate only in the event that Senator Y is a party to litigation that is pending or imminent in the court on which the judicial candidate would serve if elected. If the plurality were correct about the breadth of Rule 4.1(A)(10), then the endorsement clause would be superfluous.
To prohibit the judicial candidate from using the magic words of endorsement does not further the State’s interest in maintaining the impartiality of judges in any material way. If a judicial candidate’s formal endorsement of a legislator would establish disqualifying bias in favor of the legislator, then so too would numerous other statements of praise, support, or association that are not forbidden by the Code of Judicial Conduct. See Oral Argument at 42:40-42:53. The purported marginal value of forbidding a formal endorsement is not sufficient to constitute a “compelling” interest in maintaining impartiality.
*1058Even assuming, in the alternative, that a formal endorsement creates a bias that is not generated by other permissible public statements, recusal is a less restrictive means of achieving the objective of judicial impartiality. The Supreme Court in Caperton v. A.T. Massey Coal Co., 556 U.S. 868, 129 S.Ct. 2252, 2267, 173 L.Ed.2d 1208 (2009); id. at 2268-69 (Roberts, C.J., dissenting), and Justice Kennedy in White I, 536 U.S. at 794, 122 S.Ct. 2528 (Kennedy, J., concurring), emphasized that States are free to adopt recusal standards more rigorous than the Due Process Clause requires. The appellees present hypothetical scenarios in which recusal would pose administrative difficulties, such as the rural trial judge who endorses the prosecuting attorney who regularly appears before the judge. But the endorsement clause is not targeted carefully at these scenarios. It broadly forbids all endorsements, no matter how unlikely that a candidate would ever appear in court before the judicial candidate, and no matter how simple it would be to remedy a problem of bias through recusal in a single, isolated instance. The breadth of the clause belies the assertion that the appellees are truly concerned about serious practical problems with recusal.
In support of its view that the endorsement clause is constitutional, the plurality mistakenly crafts an overbroad “party-issue” dichotomy based on the Supreme Court’s distinction in White I between “speech for or against particular parties,” and “speech for or against particular issues.” 536 U.S. at 776, 122 S.Ct. 2528. Based on this dichotomy, the plurality concludes that the appellees are free to ban speech by a judicial candidate for or against any other candidate for public office, because the speech is not aimed at an “issue,” but rather at a “party.” Ante, at 1024-25. The announce clause at issue in White I, however, forbade speech only about issues that were “likely to come before the candidate if he is elected judge.” 536 U.S. at 771, 122 S.Ct. 2528. The Court’s parallel reference to speech about “parties” must be understood likewise to mean, at most, speech about parties who are likely to come before the candidate if he is elected. The Supreme Court surely did not signal that a State is free to suppress a judicial candidate’s speech about persons who are unlikely ever to set foot in a Minnesota courtroom.
The appellees’ asserted interest in “judicial independence” also does not justify the prohibition on political speech. In my view, the opinion concurring in the judgment, ante, at 1031, has it essentially backwards. In the real world of politics, the primary risk of “dependence” comes not from a judicial candidate’s endorsement of other candidates for public office, but from the judicial candidate’s receipt of the endorsements of those who can deliver votes: political parties, unions, trade associations, powerful elected officials, leaders of demographic blocs, political interest groups, and the like. Candidates depend on these people and their endorsements for election to public office. The Russian judges described in the concurrence, ante, at 1034-35, presumably would understand this phenomenon: they did not suspect that Justice Breyer engaged in telephone justice because he endorsed a powerful politician, but the other way around.
Yet Minnesota does not and could not forbid political actors from endorsing judicial candidates and working for their election. By opting for judicial elections, the people of Minnesota have determined that the benefits of making judges accountable to the people outweigh the risks that judges will be unduly dependent on political actors and will decide cases based on the wishes of “political cronies” who endorse them rather than on the rule of law. This is not the independent judiciary that *1059the Founders envisioned for the federal courts, but neither is it a judiciary dependent on the King’s will alone for tenure and subsistence, see The Declaration of Independence para. 11 (U.S. 1776), or on legislative fIat, see Proceedings and Debates of the Virginia State Convention of 1829-30, at 616, 619 (remarks of John Marshall). It is a judiciary that is dependent on approval of the people. Having chosen to accept a judiciary that is “dependent” in that sense, Minnesota cannot at the same time forbid political speech by judicial candidates on the ground that it creates the potential for a much more attenuated form of political dependence. Strict scrutiny requires the court to focus on the extent to which the endorsement prohibition accomplishes an asserted purpose, White I, 536 U.S. at 777 n. 7, 122 S.Ct. 2528; cf. ante, at 1035 (Loken, J., concurring in the judgment), and the tailoring in this instance is anything but narrow.
The other interests asserted by the appellees and the plurality are insufficient to justify the endorsement clause. The commentary to the governing rule states that the prohibition on endorsements by judicial candidates is designed “to prevent them from abusing the prestige of judicial office to advance the interests of others.” 52 Minn. Stat. Ann., Code of Judicial Conduct, Rule 4.1, cmt. 4. But the rule is not narrowly tailored to serve that interest, because it also applies to candidates who do not hold judicial office. See White I, 536 U.S. at 796, 122 S.Ct. 2528 (Kennedy, J., concurring). And I agree with Judge Beam and Judge Loken that the plurality’s reliance on an interest in “maintaining the appearance of impartiality,” ante, at 1022, is so broad and general that it would justify restrictions that were declared unconstitutional by the Supreme Court in White I, so it must be wrong.
The second clause at issue, the personal solicitation clause, 52 Minn. Stat. Ann., Code of Judicial Conduct, Rule 4.1(A)(6), infringes on the ability of a judicial candidate to engage in speech to raise funds that the candidate believes are necessary to mount a competitive election campaign. As the Sixth Circuit recently observed:
Prohibiting candidates from asking for money suppresses speech in the most conspicuous of ways and, in the process, favors some candidates over others— incumbent judges (who benefit from their current status) over non-judicial candidates, the well-to-do (who may not need to raise any money at all) over lower-income candidates, and the well-connected (who have an army of potential fundraisers) over outsiders. For these reasons, it is tempting to say that any limitation on a candidate’s right to ask for a campaign contribution is one limitation too many.
Carey v. Wolnitzek, 614 F.3d 189, 204-205 (6th Cir.2010). The panel in Carey acknowledged that a prohibition on “face-to-face solicitations, particularly by sitting judges,” id. at 205 (emphasis added), tested that interpretation, but found it unnecessary to decide whether such a ban could pass muster, because the Kentucky provision at issue also prohibited a range of other solicitations. Id.
The appellees’ argument here boils down to this: The Minnesota rules forbid a judicial candidate to learn who makes contributions to his campaign committee. If the candidate is permitted to solicit funds personally, then he will find out who has contributed and how much, and who has declined to contribute. A ban on solicitation is thus necessary to further a compelling government interest in protecting against judges who are biased in favor of campaign contributors or against non-contributors, or an appearance of such bias. *1060The appellees’ amici add that “there is no way to have meaningful campaign solicitations where a candidate can freely solicit contributions in one-on-one meetings with prospective donors without a substantial likelihood of learning, at least in many instances, the outcome of the ‘ask.’ ” Brief of Amici Curiae Former Governor and Chief Justices in Support of Petition for Rehearing En Banc, at 3.
The personal solicitation clause is not narrowly tailored to advance the asserted government interests. The prohibition is not limited to solicitation by sitting judges, so it is not narrowly tailored to prevent an asserted interest in the abuse of judicial office. See White I, 536 U.S. at 796, 122 S.Ct. 2528 (Kennedy, J., concurring). The appellees’ case, rather, hinges on the assumption that a judicial candidate who personally solicits funds will learn who has contributed and how much, and then develop a bias or perceived bias as a result. The parties debate whether a judicial candidate is likely to learn through other means who contributed to his campaign. The answer to that question is not dispositive, however, because either way, the personal solicitation ban cannot withstand strict scrutiny.
There is good reason to doubt the effectiveness of the Minnesota rules that are designed to prevent judges from learning the identities of contributors. Information about contributors is a matter of public record, available on the Internet. It is hard to believe that a vigilant press would not report the fact that a party appearing before a particular judge was a financial contributor to the judge’s campaign. And judges are not forbidden to read the news. See Republican Party of Minn. v. White, 416 F.3d 738, 765 n. 16 (8th Cir.2005) (en banc) (“White II ”). There are other practical considerations. If the small town law firm in a remote corner of Minnesota does battle before the local trial judge against a big city law firm, is there really much mystery about which law firm likely contributed to the trial judge’s campaign? If judges are likely to learn or surmise the identities of contributors through open sources or simple deduction, then the appellees’ primary rationale for the personal solicitation clause is nullified, and the clause is substantially underinclusive. See Carey, 614 F.3d at 205.
On the other hand, if only definitive knowledge of contributors is relevant, and Minnesota’s rules are truly effective, then the rules prevent the judicial candidate from acquiring the information that supposedly justifies the ban on personal solicitations. The rules forbid a candidate personally to accept a contribution after a personal solicitation, 52 Minn. Stat. Ann., Code of Judicial Conduct, Rule 4.1(A)(6), and a candidate must direct his campaign committee not to disclose to the candidate the identities of campaign contributors or solicited persons who refuse to contribute. Id., Rule 4.4(B)(3). These provisions are not challenged. Under the unchallenged rules, therefore, the most that a judicial candidate could possibly learn as the result of a one-on-one personal solicitation is that a prospective donor has made an unenforceable and unverifiable statement of present intention to contribute or not to contribute. Whether the prospective donor follows through on any statement of present intention is unknown to the candidate.
Even the premise of the appellees’ argument, moreover, is unsupported. It is largely speculative to assume that homeowners who answer a knock on the door by a judicial candidate will typically volunteer a decision about whether to contribute. The rule forbids personal solicitation even by a judicial candidate who expressly requests that the solicited person respond only by mail to the candidate’s campaign *1061committee. Assuming the candidate does request a reply in person, people say all sorts of things to doorway solicitors: “I will think it over.” “I need to talk to my spouse.” “I gave at the office.” The most that the appellees’ amici are prepared to assert (and this despite no apparent personal experience with making one-on-one personal solicitations in a judicial campaign) is that there is a “substantial likelihood” in “many instances” that the solicited person will offer a response. And that is only with respect to one-on-one solicitations. The personal solicitation clause is not so limited. It also forbids one-on-nineteen and one-on-eighteen and one-on-seventeen solicitations, and the appellees offer nothing to suggest that a group of nineteen hearing a pitch at the local coffee shop is likely to volunteer its intentions to a candidate. If the appellees were genuinely concerned with one-on-one solicitations, then they could have fashioned a rule aimed directly at the perceived danger. That they chose, in the wake of White II, to prohibit solicitations of groups of nineteen suggests that the appellees simply wish to restrict solicitation as much as the courts will permit.
The appellees also ignore again the message of Caperton and Justice Kennedy’s concurrence in White I that the States are free to adopt recusal standards that are more rigorous than due process requires. They argue primarily that motions to disqualify rarely succeed under current rules. But if the appellees are concerned that judges will be biased or apparently biased in favor of contributors or against non-contributors, then they can implement rules that require recusal when certain contributors appear before the judge, or on motion of a solicited person who has declined to contribute. This approach would avoid infringing on First Amendment rights, eliminate the asserted bias and appearance of bias, and remove any potential for coercion of prospective donors, who could be assured that the judicial candidate will never preside over a solicited person’s case. See, e.g., Genelle I. Belmas & Jason M. Shepard, Speaking from the Bench: Judicial Campaigns, Judges’ Speech, and The First Amendment, 58 Drake L. Rev. 709, 734 (2010) (suggesting a set of recusal standards, and arguing that “this is a better approach than attempting to stifle either judges’ or donors’ speech rights”); John Copeland Nagle, The Recusal Alternative to Campaign Finance Legislation, 37 Harv. J. on Legis. 69, 83-84, 90 (2000) (suggesting a legislative recusal requirement that “prevents both corruption and appearance of corruption” and “avoids the First Amendment problems posed by restrictions on campaign contributions and expenditures,” and arguing that “[i]f the popular perception of the corrupting influence of campaign contributions is correct, then the appropriate response is to extend the recusal requirement to the judicial context”). There is some inconvenience in managing a broader recusal rule, but this is a First Amendment case in which strict scrutiny applies. The State may not opt to prohibit political speech when there is a less drastic alternative available, even if that alternative creates some manageable administrative burdens.
* * ❖
The Framers of the Federal Constitution designed a system for federal judges in which the appointing authorities are politically accountable to the people, but in which judges are granted protections on tenure and compensation in order to promote judicial independence. See The Federalist Nos. 76-79 (Alexander Hamilton). As Justice Kennedy observed in White I, “[t]here is general consensus that the design of the Federal Constitution, including lifetime tenure and appointment by nomination and confirmation, has preserved the *1062independence of the Federal Judiciary.” 536 U.S. at 795, 122 S.Ct. 2528 (Kennedy, J., concurring). But Minnesota is free to adopt a different system, and the people of Minnesota have not seen fit to implement the federal model. Having made its choice for judicial elections, the State is constrained by the constitutional rule that content-based restrictions on speech of judicial candidates must, at a minimum, withstand strict scrutiny under the First Amendment. The endorsement clause and the personal solicitation clause do not pass this demanding test, and I would reverse the judgment of the district court in relevant part.