POSNER, Chief Judge.
The Illinois Republican Party and its chairman appeal from the dismissal, 959 F.Supp. 982 (C.D.I11.1997), on grounds of justiciability, of their suit for a declaration that the method prescribed by the state constitution for electing justices of the Supreme Court of Illinois in Cook County violates the equal protection clause of the Fourteenth Amendment. See 111. Const. Art. VI, §§ 2, 3. As amended in 1970, the constitution divides the state into five districts for purposes of electing supreme court justices. One of the districts is Cook County, which includes Chicago and surrounding suburbs. Three of the seven justices are elected by Cook County voters in at-large elections; each of the remaining four is elected by the voters of one of the other districts. This system violates the principle of “one person, one vote,” but that is not the plaintiffs’ complaint—and for the excellent reason that the principle is inapplicable to judicial elections challenged as denials of equal protection. Wells v. Edwards, 409 U.S. 1095, 93 S.Ct. 904, 34 L.Ed.2d 679 (1973), aff'g 347 F.Supp. 453 (M.D.La.1972) (three-judge district court). The complaint is that the use of the at-large method in Cook County denies members of a particular group, namely Republicans, a fair opportunity to elect candidates of their choice. White v. Regester, 412 U.S. 755, 765-70, 93 S.Ct. 2332, 37 L.Ed.2d 314 (1973); Republican Party v. Martin, 980 F.2d 943 [1062]*1062(4th Cir.1993); cf. Barnett v. Daley, 32 F.3d 1196, 1198-99 (7th Cir.1994); Garza v. County of Los Angeles, 918 F.2d 763, 771 (9th Cir.1990). Cook County as a whole has more Democratic than Republican voters, and although Republicans are sometimes elected to county office, no Republican has been elected to the supreme court from Cook County since the 1970 amendment of the state constitution. The suburban areas of the county, however, are heavily Republican, with the consequence that if Cook County were divided into three districts for the election of supreme court justices (which the state constitution forbids, Cincinnati Ins. Co. v. Chapman, 181 Ill.2d 65, 229 Ill.Dec. 264, 268-69, 691 N.E.2d 374, 378-79 (1998)), then, provided the districts were not gerrymandered, the Republicans would have a good shot at electing one justice. So an effect of the at-large system is to deprive Republicans in Cook County of the power they would have under a districted system to elect a Republican justice.
The equal protection clause has been interpreted to require some degree, often a high degree (notably in legislative reapportionment cases, where the rule of “one person one vote” reigns), of equality in voting power, in the sense that each voter’s vote should have to the extent feasible the same weight in the political process as every other voter’s vote. Of course, the goal of equality of voting power must remain to a considerable degree aspirational. The system for electing United States Senators, a system not only embedded in the Constitution but expressly placed beyond the power of alteration by Article V, has the effect of weighting votes for Senators in states that have a small population more heavily than the votes for Senators in states that have a large population. And voters in swing districts have more effective voting power than voters in districts that are politically lopsided. These disparities cannot be corrected. Others can be. The Court long ago held racial gerrymandering unconstitutional. Gomillion v. Lightfoot, 364 U.S. 339, 81 S.Ct. 125, 5 L.Ed.2d 110 (1960); see also id. at 349, 81 S.Ct. 125 (concurring opinion). Later it held that state legislative districts must have the same population. Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964). Then that gerrymanders aimed at diluting the voting power of supporters of one of the political parties were, like racial gerrymanders, justiciable. Davis v. Bandemer, 478 U.S. 109, 106 S.Ct. 2797, 92 L.Ed.2d 85 (1986). And later still that malapportionment of judicial election districts could violate the Voting Rights Act. Chisom v. Roemer, 501 U.S. 380, 111 S.Ct. 2354, 115 L.Ed.2d 348 (1991). The Supreme Court has not yet had a case in which the use of at-large elections to fill judicial offices is challenged as a denial of equal protection because aimed at preventing the election of candidates of one of the parties. But like the Fourth Circuit in Republican Party v. Martin, supra—the only case similar to this one that we’ve found—we cannot see an objection in principle to what would be after all only a modest extension of existing law. The at-large election is a familiar device, as we know from such cases as Cane v. Worcester County, 35 F.3d 921, 925-27 (4th Cir.1994), and Westwego Citizens for Better Government v. City of Westwego, 946 F.2d 1109, 1116-23 (5th Cir.1991), for reducing the voting power of a minority, whether racial or political; and the reduction is the same whatever the office, whether legislative, executive, or judicial, is to be filled by election.
Suppose, to take a simplified version of the present case, that a state is divided into four legislative districts of which one has 50 percent of the population and the other three 16.67 percent each. The large district elects 3 legislators, all at large, and the small ones elect 1 legislator each. Party A is supported by 60 percent of the voters in the large district, Party B by the other 40 percent. Therefore Party A’s candidates win all three seats. But suppose that support for the two parties is uneven across the large district; if that district were divided into three equal parts without regard to politics, Party A would command the support of 80 percent of the voters in each of two subdistricts but Party B would command the support of 80 percent of the voters in the third subdistriet. (To see this, imagine that there is a total of 1,200 voters in the district as a whole, 400 in each subdistrict. Party A will then have 720 voters in total (.6 x 1,200) of whom 320 will be in the first subdistrict (.8 x 400), 320 in [1063]*1063the second, and therefore only 80 (720 - 320 - 320) in the third: 20 percent.) So A would end up with two seats and B with one. B’s supporters would still have less effective voting power than A’s, because B with 40 percent of the votes would elect only 33 percent of the legislators from the district and A with 60 percent would elect 67 percent of the legislators. Still, that is far different from 40 percent of the electorate electing zero, percent of the legislators and 60 percent of the electorate electing 100 percent of the legislators.
Given the experience that the courts have had with challenges to at-large elections, including at-large elections for judges (Chisom), we cannot understand the basis for the district court’s holding that the challenge to the at-large system for electing state, supreme court justices from Cook County is not justiciable. Judicial reluctance to enter the political thicket of state electoral systems was overcome many years ago, and we cannot see what difference it makes that the election is of judges rather than of legislators.
Free access — add to your briefcase to read the full text and ask questions with AI
POSNER, Chief Judge.
The Illinois Republican Party and its chairman appeal from the dismissal, 959 F.Supp. 982 (C.D.I11.1997), on grounds of justiciability, of their suit for a declaration that the method prescribed by the state constitution for electing justices of the Supreme Court of Illinois in Cook County violates the equal protection clause of the Fourteenth Amendment. See 111. Const. Art. VI, §§ 2, 3. As amended in 1970, the constitution divides the state into five districts for purposes of electing supreme court justices. One of the districts is Cook County, which includes Chicago and surrounding suburbs. Three of the seven justices are elected by Cook County voters in at-large elections; each of the remaining four is elected by the voters of one of the other districts. This system violates the principle of “one person, one vote,” but that is not the plaintiffs’ complaint—and for the excellent reason that the principle is inapplicable to judicial elections challenged as denials of equal protection. Wells v. Edwards, 409 U.S. 1095, 93 S.Ct. 904, 34 L.Ed.2d 679 (1973), aff'g 347 F.Supp. 453 (M.D.La.1972) (three-judge district court). The complaint is that the use of the at-large method in Cook County denies members of a particular group, namely Republicans, a fair opportunity to elect candidates of their choice. White v. Regester, 412 U.S. 755, 765-70, 93 S.Ct. 2332, 37 L.Ed.2d 314 (1973); Republican Party v. Martin, 980 F.2d 943 [1062]*1062(4th Cir.1993); cf. Barnett v. Daley, 32 F.3d 1196, 1198-99 (7th Cir.1994); Garza v. County of Los Angeles, 918 F.2d 763, 771 (9th Cir.1990). Cook County as a whole has more Democratic than Republican voters, and although Republicans are sometimes elected to county office, no Republican has been elected to the supreme court from Cook County since the 1970 amendment of the state constitution. The suburban areas of the county, however, are heavily Republican, with the consequence that if Cook County were divided into three districts for the election of supreme court justices (which the state constitution forbids, Cincinnati Ins. Co. v. Chapman, 181 Ill.2d 65, 229 Ill.Dec. 264, 268-69, 691 N.E.2d 374, 378-79 (1998)), then, provided the districts were not gerrymandered, the Republicans would have a good shot at electing one justice. So an effect of the at-large system is to deprive Republicans in Cook County of the power they would have under a districted system to elect a Republican justice.
The equal protection clause has been interpreted to require some degree, often a high degree (notably in legislative reapportionment cases, where the rule of “one person one vote” reigns), of equality in voting power, in the sense that each voter’s vote should have to the extent feasible the same weight in the political process as every other voter’s vote. Of course, the goal of equality of voting power must remain to a considerable degree aspirational. The system for electing United States Senators, a system not only embedded in the Constitution but expressly placed beyond the power of alteration by Article V, has the effect of weighting votes for Senators in states that have a small population more heavily than the votes for Senators in states that have a large population. And voters in swing districts have more effective voting power than voters in districts that are politically lopsided. These disparities cannot be corrected. Others can be. The Court long ago held racial gerrymandering unconstitutional. Gomillion v. Lightfoot, 364 U.S. 339, 81 S.Ct. 125, 5 L.Ed.2d 110 (1960); see also id. at 349, 81 S.Ct. 125 (concurring opinion). Later it held that state legislative districts must have the same population. Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964). Then that gerrymanders aimed at diluting the voting power of supporters of one of the political parties were, like racial gerrymanders, justiciable. Davis v. Bandemer, 478 U.S. 109, 106 S.Ct. 2797, 92 L.Ed.2d 85 (1986). And later still that malapportionment of judicial election districts could violate the Voting Rights Act. Chisom v. Roemer, 501 U.S. 380, 111 S.Ct. 2354, 115 L.Ed.2d 348 (1991). The Supreme Court has not yet had a case in which the use of at-large elections to fill judicial offices is challenged as a denial of equal protection because aimed at preventing the election of candidates of one of the parties. But like the Fourth Circuit in Republican Party v. Martin, supra—the only case similar to this one that we’ve found—we cannot see an objection in principle to what would be after all only a modest extension of existing law. The at-large election is a familiar device, as we know from such cases as Cane v. Worcester County, 35 F.3d 921, 925-27 (4th Cir.1994), and Westwego Citizens for Better Government v. City of Westwego, 946 F.2d 1109, 1116-23 (5th Cir.1991), for reducing the voting power of a minority, whether racial or political; and the reduction is the same whatever the office, whether legislative, executive, or judicial, is to be filled by election.
Suppose, to take a simplified version of the present case, that a state is divided into four legislative districts of which one has 50 percent of the population and the other three 16.67 percent each. The large district elects 3 legislators, all at large, and the small ones elect 1 legislator each. Party A is supported by 60 percent of the voters in the large district, Party B by the other 40 percent. Therefore Party A’s candidates win all three seats. But suppose that support for the two parties is uneven across the large district; if that district were divided into three equal parts without regard to politics, Party A would command the support of 80 percent of the voters in each of two subdistricts but Party B would command the support of 80 percent of the voters in the third subdistriet. (To see this, imagine that there is a total of 1,200 voters in the district as a whole, 400 in each subdistrict. Party A will then have 720 voters in total (.6 x 1,200) of whom 320 will be in the first subdistrict (.8 x 400), 320 in [1063]*1063the second, and therefore only 80 (720 - 320 - 320) in the third: 20 percent.) So A would end up with two seats and B with one. B’s supporters would still have less effective voting power than A’s, because B with 40 percent of the votes would elect only 33 percent of the legislators from the district and A with 60 percent would elect 67 percent of the legislators. Still, that is far different from 40 percent of the electorate electing zero, percent of the legislators and 60 percent of the electorate electing 100 percent of the legislators.
Given the experience that the courts have had with challenges to at-large elections, including at-large elections for judges (Chisom), we cannot understand the basis for the district court’s holding that the challenge to the at-large system for electing state, supreme court justices from Cook County is not justiciable. Judicial reluctance to enter the political thicket of state electoral systems was overcome many years ago, and we cannot see what difference it makes that the election is of judges rather than of legislators. The district judge expressed understandable reluctance to enter the “quagmire” of electoral reform, but did not explain why the quagmire is any deeper when the election is of judges. It is true as he remarked that judges perform different functions from legislators, but we do not see how this bears on justiciability.
The concept of justiciability is designed to confine the federal courts to the traditional core judicial functions of Anglo-American judiciaries. So when there is no “case” in that traditional sense, because of mootness, lack of adversity, lack of standing in the sense either that the plaintiff has suffered no injury from the defendant’s alleged wrongdoing or the court cannot grant relief that will confer a benefit on the plaintiff, lack of any law to apply (as when the court is asked to decide a “political question” in the sense of a question that lies outside judicial competence because of lack of judicially administrable standards or an unwillingness to “take on”’ a coequal branch of government in sensitive areas such as foreign relations), the suit will be dismissed as nonjusticiable. E.g., Steel Co. v. Citizens for a Better Environment, — U.S. -, 118 S.Ct. 1003, 1016-20, 140 L.Ed.2d 210 (1998); Spencer v. Kemna, — U.S. -, 118 S.Ct. 978, 140 L.Ed.2d 43 (1998); Nixon v. United States, 506 U.S. 224, 113 S.Ct. 732, 122 L.Ed.2d 1 (1993); United States v. Johnson, 319 U.S. 302, 63 S.Ct. 1075, 87 L.Ed. 1413 (1943) (per curiam). None of these conditions is present in this case. The Illinois Republican Party and its chairman have a real dispute with the State of Illinois; the equal protection clause provides a legal framework for its resolution; a judicial decree (for example splitting Cook County into three compact districts equal in population, each to elect one supreme court justice) could be entered that would provide the plaintiffs with real relief from the harm inflicted on them by the wrong that they allege; and redistricting (essentially what the plaintiffs are.requesting) is no longer considered to present a “political question” blocking the exercise of federal jurisdiction.
It is true that the “one person, one vote” rule is inapplicable to a vote-dilution claim. Chisom v. Roemer, supra, 501 U.S. at 403 n. 32, 111 S.Ct. 2354. Electoral districts cannot be configured in such a way that every racial, ethnic, political, or other group that might have strong political preferences has a proportionately equal chance of electing the candidates of its choice. But the Supreme Court crossed this bridge—the bridge that spans the objections to justiciability that are based on the absence of easily administrable legal standards—as long ago as Gomillion, where it first entertained a vote-dilution claim. And in Chisom the Court held that the notably vague standard of the Voting Rights Act—“totality of circumstances”—was nonetheless sufficiently administrable by courts to make claims of vote dilution in judicial elections justiciable. 501 U.S. at 402-03, 111 S.Ct. 2354. See also Republican Party v. Martin, supra, 980 F.2d at 950-52; cf. Voter Information Project, Inc. v. City of Baton Rouge, 612 F.2d 208 (5th Cir.1980). And Chisom was decided after the Supreme Court had held that the one person, one vote rule does not apply to judicial elections. If vote dilution in judicial elections is justiciable under an open-ended “totality of circumstances” law, it is justiciable under the no more open-ended equal protection clause.
[1064]*1064We conclude that this suit is justiciable and proceed to the merits, noting first that to be actionable a denial of equal protection must be intentional. E.g., Washington v. Davis, 426 U.S. 229, 96 S.Ct. 2040, 48 L.Ed.2d 597 (1976); Personnel Administrator v. Feeney, 442 U.S. 256, 99 S.Ct. 2282, 60 L.Ed.2d 870 (1979); People Who Care v. Rockford Board of Education, 111 F.3d 528, 534 (7th Cir.1997); Majeske v. Fraternal Order of Police, 94 F.3d 307, 311 (7th Cir.1996); Hayden v. Grayson, 134 F.3d 449, 452-53 (1st Cir.1998); United States v. Changco, 1 F.3d 837, 840 (9th Cir.1993). It follows that disparate impact—a law’s unintentionally bearing harder on one group than another—■ is not a permissible basis for finding a denial of equal protection. Columbus Board of Education v. Penick, 443 U.S. 449, 464, 99 S.Ct. 2941, 61 L.Ed.2d 666 (1979); United States v. Washington, 109 F.3d 335, 338 (7th Cir.1997); Majeske v. Fraternal Order of Police, supra, 94 F.3d at 311; Hayden v. Grayson, supra, 134 F.3d at 453. Disparities can be used as evidence of intentional discrimination, Personnel Administrator v. Feeney, supra, 442 U.S. at 279 n. 24, 99 S.Ct. 2282, and sometimes, as was the case in Gomillion v. Lightfoot, supra, 364 U.S. at 341, 81 S.Ct. 125, they are so glaring and so patently without justification as to give rise to an irresistible inference that they are the consequence of intentional discrimination. See also Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 266, 97 S.Ct. 555, 50 L.Ed.2d 450 (1977). But that is not the case here. Given the decisions not challenged here to fill the office of supreme court justice by partisan election, to have a supreme court of seven justices, and to elect the justices from districts rather than have statewide at-large elections, the State of Illinois faced the difficult question of how to district the state for this electoral purpose. If it divided it into seven districts on the basis of population, Cook County, with a little more than half the state’s population, would be entitled to three and a half districts. And this meant that if, in recognition that a county is a community of interest, no district were to cross the county line, Cook County would have to have either three or four districts. Either way, each supreme court justice- from Cook County would represent a tiny (relative to the other districts in the state) though densely populated geographical area. This might be thought to encourage too parochial an outlook or to limit the field of selection too tightly. The alternative chosen was to give Cook County as many justices as it would be entitled to if it had three districts, but to make the candidates for these positions run at large, so they would have a county-wide perspective and so the field of selection of candidates would be county-wide. This alternative to splitting the County into three or four districts, or creating districts that crossed the county line, was not so irrational from a nonpartisan “good government” standpoint as to prove without more that the actual motive was partisan.
We are not saying that the justification that we have just sketched was the actual reason for the adoption of the challenged system. We have no idea whether it was. But because such a justification is possible, the plaintiffs could not ask the district court to infer from the system itself, with no (other) evidence of motive, that the motive was a bad one. Since the suit was dismissed on the pleadings, the plaintiffs had no opportunity to present such evidence. But in their reply brief in this court, responding to the defendants’ alternative contention that the complaint, although detailed, does not allege facts that if proved would support an inference of intentional denial of equal protection of the laws, so that the suit should have been dismissed on the merits even if justiciable, the plaintiffs argue that the facts alleged in the complaint are enough to support an inference of intentional discrimination, so that the dismissal of the complaint was premature. They allude vaguely to a desire to prove additional facts but do not indicate what facts they have in mind.
A plaintiff is not held to the factual allegations of his complaint when he is faced with a motion to dismiss it for failure to state a claim. He can oppose the motion, either in the district court, or, if the motion is granted, on appeal, with any factual allegations that are consistent with the allegations of the complaint. American Inter-Fidelity Exchange v. American Re-Insurance Co., 17 F.3d 1018, 1022 (7th Cir.1994); Harrell v. United States, 13 F.3d 232, 236 (7th Cir. [1065]*10651993); Orthmann v. Apple River Campground, Inc., 757 F.2d 909, 914-15 (7th Cir.1985); Sunn Resource Systems, Inc. v. Lycoming County, 883 F.2d 245, 247 (3d Cir.1989). And he can of course argue in the alternative—that the facts alleged in the complaint are sufficient by themselves to state a claim, but if not then they are at least consistent with stating a claim. But the plaintiffs have pointed to no facts that they want to prove other than those alleged in the complaint. They have no other facts because they waited too long to bring this suit. The discrimination of which they complain occurred almost thirty years ago, when the state constitution was amended to create the at-large system for judicial elections. It is too late to conduct discovery to learn the motives of the proponents of the amendment. The plaintiffs are stuck with the published legislative history, which in some cases would suffice to establish an intentional violation of the Constitution, see Edwards v. Aguillard, 482 U.S. 578, 587-93, 107 S.Ct. 2573, 96 L.Ed.2d 510 (1987); id. at 599-602 (concurring opinion); Wallace v. Jaffree, 472 U.S. 38, 56-60, 105 S.Ct. 2479, 86 L.Ed.2d 29 (1985); Village of Arlington Heights v. Metropolitan Housing Development Corp., supra, 429 U.S. at 268, 97 S.Ct. 555; see also Bown v. Gwinnett County School District, 112 F.3d 1464, 1471-72 (11th Cir.1997)—but not in this case.
All that the plaintiffs can hope to show at this late date, with the evidence they have or can obtain, is that the Democrats favored the at-large approach to electing supreme court justices in Cook County in the 1970 constitutional convention and that the approach has indeed favored Democratic candidates for those positions. The Republicans opposed the approach, but mainly because they wanted to substitute gubernatorial appointment of supreme court justices for election. There is a great deal to be said for appointing judges rather than for electing them, Steven P. Croley, “The Majoritarian Difficulty: Elective Judiciaries and the Rule of Law,” 62 U. Chi. L.Rev. 689 (1995), let alone for electing them in partisan elections. But a preference for the electoral method cannot be read automatically as a desire to dilute the voting power of one’s political opponents, given the neutral reasons that might be offered for not making election districts for state supreme court justices smaller than a county, any more than the Republicans’ preference for appointed justices can be read automatically as motivated by the fact that the Republican Party has dominated the Illinois governorship for the last two decades.
As we said earlier, it may well be infeasible to obtain better evidence of intent when one is dealing with a constitutional amendment adopted many years ago—and by a referendum, expressing the views of the electorate and not just of some backroom schemers—and when the real vice may be the failure to change the method of election now that experience has shown that it really does discriminate against Republican candidates. But inability to obtain evidence of an essential element of a claim is not a good reason for dispensing with the element, and anyway that is something that only the Supreme Court, which is responsible for the rule that denials of equal protection must be intentional to be actionable, can do. We cannot accept the suggestion that Davis v. Bandemer fixes a lower threshold for proof of a denial of equal protection in a political vote dilution case than' in a racial one. An intentional denial must be proved, and cannot be inferred just from the fact that the challenged method of election favors one party. There would no end to litigation, since a method of election is bound to favor one party over another and thus be constitutionally suspicious if Davis allows intention to be inferred from consequences.
Republican Party v. Martin, supra, is consistent with these principles, even though the court found a prima facie violation. At issue was a North Carolina law that provided for the election of superior court judges in statewide elections. As a result of this system, only one Republican had been elected to a superior court judgeship in literally hundreds of elections, 980 F.2d at 948, even though Republicans constituted almost half the state’s voters, id. at 949; and since the superior court judges had a local jurisdiction, no neutral, “good government” explanation for the statewide at-large method of electing them was discernible. Id. at 955. In these circumstances, taken against the background of North Carolina’s history as a “one-party state,” an inference of deliberate vote dilution fairly leapt out at the reader of the [1066]*1066complaint. No such inference is possible here.
All that the plaintiffs would have had to do to update the record and present some evidence of intentional discrimination was to press for a constitutional amendment changing the at-large system and to make a careful record of the reaction of the Democratic Party to the proposal. The plaintiffs have offered to prove that after the oral argument of this appeal the Illinois senate voted on a resolution to amend the state constitution to allow the division of Cook County into three districts for the election of supreme court justices—and the Democrats voted in a bloc against the resolution and as a result it failed for want of the required supermajority for constitutional amendments. But in the absence of legislative history, to which the plaintiffs have not directed us, the Democrats’ preference for the system that favors them cannot be equated to intentional discrimination, given the neutral objections noted earlier to districting the county for purposes of judicial elections.
To summarize, the suit is justiciable, contrary to what the district judge thought, but it fails to state a claim because the plaintiffs have not alleged and do not seek an opportunity to prove facts essential to establish that the discrimination of which they complain is intentional. So far as appears, the electoral pi’aetice that they challenge was adopted and is maintained because the alternatives are even worse from a good-government standpoint. We therefore modify the judgment of the district court to base dismissal on lack of merit rather than on lack of jurisdiction and we affirm the judgment as modified.
MODIFIED AND AFFIRMED.