Easterbrook, Circuit Judge.

Keri-Lyn J. Krafthefer, Attorney, Ancel, Glink, Diamond, Bush, Dicianni & Krafthefer, P.C., Chicago, IL, for Plaintiffs–Appellant.

John B. Murphey, Attorney, Rosenthal, Murphey, Coblentz & Donahue, James P. Nally, Attorney, Marie D. Spicuzza, Attorney, Office of the Cook County State's Attorney, Civil Actions Bureau, Chicago, IL, Burton S. Odelson, Attorney, Odelson & Sterk, Ltd, Evergreen Park, IL, for Defendants–Appellees.

Before Wood, Chief Judge, and Easterbrook and Sykes, Circuit Judges.

Easterbrook, Circuit Judge.Thaddeus Jones, an alderman in Calumet City, Illinois, wants to be mayor. One of his supporters, Stevon Grant (plus others who formed a committee), tried to prevent the incumbent, Michelle Markiewicz-Qualkinbush, mayor since 2003, from running for reelection in spring 2017. The means: a referendum that would have set a three-term limit on the City’s mayor. Grant gathered enough signatures to put that referendum on the ballot in November 2016. But it did not appear on that ballot, because the City itself proposed three referenda for that election, and the City’s proposals were certified before Grant’s. Illinois law limits to three the number of referenda on any ballot. See 10 ILCS 5/28-1. The parties call this the "Rule of Three." Illinois law creates a possibility that displaced referenda will roll over to the next election. 10 ILCS 5/28-5. Grant did not ask that his proposal do so.

One of the City’s proposals was a term-limits rule that would prevent the election as mayor of anyone who has served four or more consecutive terms as either mayor or alderman. That did not block Markiewicz-Qualkinbush from running but did bar Jones, who had been elected as an alderman in 1997 and was in his fifth term. That referendum passed, and Jones was removed from the ballot for the April 2017 mayoral race. Markiewicz-Qualkinbush was reelected. (Jones says that the City’s other two proposals also were aimed at him, but they do not require discussion.)

Jones filed two lawsuits—one in federal court under federal law, the other in state court under state law. In each he sought an injunction against the application of the Rule of Three and an order removing the City’s term-limits referendum from the ballot or nullifying the voters' approval of that referendum. In the federal suit, which was joined by Grant and the citizens group, plaintiffs also sought damages. Jones lost the state suit. Jones v. Calumet City , 2017 IL App (1st) 170236, 420 Ill.Dec. 371, 96 N.E.3d 456. (The defendants in the federal suit have not invoked preclusion, even though Jones deliberately split his claims.) The district court denied Jones’s motion for a preliminary injunction, and we affirmed. Jones v. Markiewicz-Qualkinbush , 842 F.3d 1053 (7th Cir. 2016). Now we have the appeal from the district court’s final decision in defendants' favor. The request for an injunction against the application of the Rule of Three in November 2016 is moot, but the requests for damages plus prospective relief that would knock out the Rule of Three in the future, and abrogate the term-limits rule that prevents Jones from running for mayor, remain live.

According to Jones (as we now call the three federal plaintiffs collectively), the Rule of Three violates the First Amendment (applied to the states by the Fourteenth Amendment) because it disables him from asking voters to support his proposal. The Rule of Three selects the first three proposals to be certified, and a municipality can reach that goal with as little as 48 hours' notice, see 5 ILCS 120/2.02(a), while a private citizen’s proposal depends on acquiring enough signatures. 10 ILCS 5/28-7. This means that a city observing a signature-gathering campaign in progress can get its own proposals on the ballot first—even if the real goal of those proposals is just to prevent the private ones from appearing. Jones contends that this is what happened in 2016 and maintains that any system barring private proposals from the ballot—whether directly or by allowing a unit of government to fill the available slots—violates the First Amendment.

This assumes that the ballot is a public forum and that there is a constitutional right to place referenda on the ballot. But there is no such right. Nothing in the Constitution guarantees direct democracy. The Constitution establishes the United States as an indirect democracy, in which elected representatives make the law. The nation’s founders thought that direct democracy would produce political instability and contribute to factionalism. See, e.g., Federalist No. 10 (Madison). There has never been a federal referendum. Nor has any federal court ever concluded that the ballot is a public forum that must be opened to referenda, let alone to as many referenda as anyone cares to propose.

To the contrary, many courts have held that private citizens lack a right to propose referenda or initiatives for any ballot, federal or state. See Molinari v. Bloomberg , 564 F.3d 587, 597 (2d Cir. 2009) ("[T]he right to pass legislation through a referendum is a state-created right not guaranteed by the U.S. Constitution"); Kendall v. Balcerzak , 650 F.3d 515, 523 (4th Cir. 2011) ("The referendum is a form of direct democracy and is not compelled by the Federal Constitution"); Taxpayers United for Assessment Cuts v. Austin , 994 F.2d 291, 295 (6th Cir. 1993) ("[W]e conclude that ... the Constitution does not require a state to create an initiative procedure"); Dobrovolny v. Moore , 126 F.3d 1111, 1113 (8th Cir. 1997) ("Clearly, the right to a state initiative process is not a right guaranteed by the United States Constitution, but is a right created by state law"); Angle v. Miller , 673 F.3d 1122, 1133 (9th Cir. 2012) ("There is no First Amendment right to place an initiative on the ballot"); Petrella v. Brownback , 787 F.3d 1242, 1259 (10th Cir. 2015) ("[W]e have repeatedly held that there is no First Amendment right to propose a voter initiative"); Biddulph v. Mortham , 89 F.3d 1491, 1497–98 (11th Cir. 1996) ("[T]he right to place a citizen initiative proposal on the ballot is a state-created right (and thus, by implication, not a right guaranteed by the First Amendment)."). See also Timmons v. Twin Cities Area New Party , 520 U.S. 351, 363, 117 S.Ct. 1364, 137 L.Ed.2d 589 (1997) ("Ballots serve primarily to elect candidates, not as forums for political expression."); John Doe No. 1 v. Reed , 561 U.S. 186, 212, 130 S.Ct. 2811, 177 L.Ed.2d 493 (2010) (Sotomayor, J., concurring, joined by Stevens & Ginsburg, JJ.) ("[W]e must be mindful of the character of initiatives and referenda. These mechanisms of direct democracy are not compelled by the Federal Constitution"); Georges v. Carney , 691 F.2d 297, 300 (7th Cir. 1982) ("[T]he parties to this litigation agree that there is no constitutional right to use the ballot box as a forum for advocating a policy").

Many of these decisions have cited Meyer v. Grant , 486 U.S. 414, 108 S.Ct. 1886, 100 L.Ed.2d 425 (1988). Meyer concerned a challenge to a Colorado statute that criminalized the payment of petition circulators. Among other things, the state argued that, because the federal Constitution did not prohibit Colorado’s government from completely eliminating citizens' state-created right to propose ballot initiatives, the statute did not significantly burden protected speech. The Court rejected this argument and ultimately struck down the statute—a state that does open the ballot cannot impose unconstitutional conditions—but did not reject the premise that the right to propose initiatives is an exclusively state-created right that the First Amendment does not guarantee. Id . at 424–25, 108 S.Ct. 1886.

So is the Rule of Three an unconstitutional condition on the exercise of a state-created right? Because the Rule of Three does not distinguish by viewpoint or content, the answer depends on whether the rule has a rational basis, not on the First Amendment. There can be little doubt that the Rule of Three is rationally related to a legitimate state objective. Each voter knows that other people likely will determine the outcome. That leads to free riding: each voter is tempted to allow others to do the work of reading about candidates, studying proposals, and making hard decisions. Many voters will do the work out of civic spirit, and others will do it out of self-interest (if some candidate or proposal could have a big effect on that voter), but the more complex the ballot the less attention each candidate and proposal receives. So the Supreme Court has said that states have a strong interest in simplifying the ballot. See, e.g., Munro v. Socialist Workers Party , 479 U.S. 189, 194–96, 107 S.Ct. 533, 93 L.Ed.2d 499 (1986) ; Illinois State Board of Elections v. Socialist Workers Party , 440 U.S. 173, 184–85, 99 S.Ct. 983, 59 L.Ed.2d 230 (1979) ; Bullock v. Carter , 405 U.S. 134, 145, 92 S.Ct. 849, 31 L.Ed.2d 92 (1972).

Limiting the number of referenda improves the chance that each will receive enough attention, from enough voters, to promote a well-considered outcome. There’s nothing magical about three; it may be too low (or too high; remember that the cap in federal elections is zero); but the benefit of some limit is plain. That is enough to show that the rule used in Illinois is valid. Indeed, Georges sustains the Rule of Three against the sort of arguments that Jones has advanced.

If we take Jones as objecting to the order in which proposals are placed on the ballot—with a municipality’s coming first as a practical matter, creating the possibility that all private proposals will be excluded—nothing changes. Because the ballot is not a public forum, the Constitution does not prevent a state from reserving the referendum process for its own communication, asking the voters to give thumbs up or down to municipal proposals while preventing any other access. That is indeed how many nonpublic forums operate, as means for communication by a single speaker or group of favored speakers. See, e.g., Perry Education Association v. Perry Local Educators' Association , 460 U.S. 37, 103 S.Ct. 948, 74 L.Ed.2d 794 (1983).

Jones makes a different kind of challenge to the referendum that knocked him out of the race for mayor. He contends that placing the City’s referendum on the ballot violated the Equal Protection Clause of the Fourteenth Amendment. It does not draw any suspect or forbidden line; it covers all persons who have served four or more terms in local offices. No one thinks that term limits are uniformly unconstitutional. But Jones says that this referendum was aimed at him, specifically, and therefore treated him as a prohibited class of one. See Village of Willowbrook v. Olech , 528 U.S. 562, 120 S.Ct. 1073, 145 L.Ed.2d 1060 (2000) ; Del Marcelle v. Brown County , 680 F.3d 887 (7th Cir. 2012) (en banc).

The Appellate Court of Illinois observed that this is false in fact. Three aldermen, not just Jones, were in their fourth or fifth terms and thus were ruled out of the mayoral race. 2017 IL App (1st) 170236 at ¶6, 420 Ill.Dec. 371, 96 N.E.3d 456. The referendum also prevents Markiewicz-Qualkinbush from running for reelection in 2021. It does not identify Jones by name; he does not contend that it should be treated as a bill of attainder. Still, he says that because he was the only person affected by the referendum in 2017 (the other two aldermen did not want to run for mayor), it should be treated as if it specified him as a class of one.

Olech holds that governmental action in class-of-one situations requires a rational basis. See also Nixon v. Administrator of General Services , 433 U.S. 425, 468–84, 97 S.Ct. 2777, 53 L.Ed.2d 867 (1977). (Whether there is also a state-of-mind requirement divided the court in Del Marcelle , which did not produce a majority for any rationale.) Jones does not deny that term-limits rules in general are supported by rational bases; instead he contends that this referendum lacks an adequate basis for knocking him , in particular, out of the 2017 race. The proposal designed to eliminate Markiewicz-Qualkinbush from the 2017 election may or may not have been a political dirty trick, but Jones tells us that the City’s referendum definitely was a political dirty trick rather than a piece of neutral legislation.

That’s politics: if Senators from State A propose to cut off a project valuable to State B, and Senators from B then support a tariff that hurts producers in State A, courts don't use the Equal Protection Clause to regulate the outcome. Politics is a rough-and-tumble game, where hurt feelings and thwarted ambitions are a necessary part of robust debate. See Manley v. Law , 889 F.3d 885 (7th Cir. 2018). It is impossible to imagine the judiciary attempting to decide when a politically retaliatory step goes "too far" without displacing the people’s right to govern their own affairs and making the judiciary just another political tool for one faction to wield against its rivals. The right response is political: Jones could have campaigned against the City’s referendum, and if the people wanted him to be mayor they could have defeated the proposed term limit. Instead it received about 65% of the votes cast.

Engquist v. Oregon Department of Agriculture , 553 U.S. 591, 128 S.Ct. 2146, 170 L.Ed.2d 975 (2008), holds that a fired public employee cannot use a class-of-one claim to contest the discharge or otherwise to ask a federal court to govern management of the workplace. The Court observed that a public employer, like a private employer, must exercise control that is bound to ruffle some feelings and produce economic injury. It added that judicial intervention under a class-of-one approach would substantially displace hierarchical (or civil-service) management practices and crimp management in ways that would do more harm than good. See also Garcetti v. Ceballos , 547 U.S. 410, 126 S.Ct. 1951, 164 L.Ed.2d 689 (2006) (for similar reasons, the First Amendment does not apply to speech that is part of a public job).

Everything that Engquist and Ceballos said about using constitutional law to regulate personnel management in a public workforce goes double about using class-of-one litigation to regulate political infighting. Any effort by the judiciary to stop one politician from proposing and advocating steps that injure another politician would do more to violate the First Amendment (the right to advocate one’s view of good policy is the core of free speech) than to vindicate the Equal Protection Clause. Laws with general effects must have the support of a rational basis, but as we observed earlier the Rule of Three has such a basis, and Jones does not contest the validity of term limits. A class-of-one claim cannot be used to attack political practices that are valid as a general matter but bear especially hard on one politician. Cf. Washington v. Davis , 426 U.S. 229, 96 S.Ct. 2040, 48 L.Ed.2d 597 (1976) (only disparate treatment can violate the Equal Protection Clause; disparate impact does not); Personnel Administrator of Massachusetts v. Feeney , 442 U.S. 256, 99 S.Ct. 2282, 60 L.Ed.2d 870 (1979) (same). The price of political dirty tricks must be collected at the ballot box rather than the courthouse.

AFFIRMED