dissenting.

I join Justice Souter’s powerfully persuasive dissenting opinion, and would make concrete what is implicit in his exposition. The plurality’s interpretation of §2 of the Voting Rights Act of 1965 is difficult to fathom and severely undermines the statute’s estimable aim. Today’s decision returns the ball to Congress’ court. The Legislature has just cause to clarify beyond debate the appropriate reading of § 2.

dissenting.

I join Justice Souter’s opinion in full. I write separately in light of the plurality’s claim that a bright-line 50% rule (used as a Thornburg v. Gingles, 478 U. S. 30 (1986), gateway) serves administrative objectives. In the plurality’s view, that rule amounts to a relatively simple administrative device that will help separate at the outset those cases that are more likely meritorious from those that are not. Even were that objective as critically important as the plurality believes, however, it is not difficult to find other numerical gateway rules that would work better.

Assume that a basic purpose of a gateway number is to separate (1) districts where a minority group can “elect representatives of their choice,” from (2) districts where the minority, because of the need to obtain majority crossover votes, can only “elect representatives” that are consensus candidates. 42 U. S. C. § 1973(b) (2000 ed.); League of *45United Latin American Citizens v. Perry, 548 U. S. 399, 445 (2006) (plurality opinion). At first blush, one might think that a 50% rule will work in this respect. After all, if a 50% minority population votes as a bloc, can it not always elect the candidate of its choice? And if a minority population constitutes less than 50% of a district, is not any candidate elected from that district always a consensus choice of minority and majority voters? The realities of voting behavior, however, make clear that the answer to both these questions is “no.” See, e. g., Brief for Nathaniel Persily et al. as Amici Curiae 5-6 (“Fifty percent is seen as a magic number by some because under conditions of complete racial polarization and equal rates of voting eligibility, registration, and turnout, the minority community will be able to elect its candidate of choice. In practice, such extreme conditions are never present. . . . [S]ome districts must be more than 50% minority, while others can be less than 50% minority, in order for the minority community to have an equal opportunity to elect its candidate of choice” (emphasis added)); see also ante, at 32-33 (Souter, J., dissenting).

No voting group is 100% cohesive. Except in districts with overwhelming minority populations, some crossover votes are often necessary. The question is how likely it is that the need for crossover votes will force a minority to reject its “preferred choice” in favor of a “consensus candidate.” A 50% number does not even try to answer that question. To the contrary, it includes, say, 51% minority districts, where imperfect cohesion may, in context, prevent election of the “minority-preferred” candidate, while it excludes, say, 45% districts where a smaller but more cohesive minority can, with the help of a small and reliable majority crossover vote, elect its preferred candidate.

Why not use a numerical gateway rule that looks more directly at the relevant question: Is the minority bloc large enough, is it cohesive enough, is the necessary majority crossover vote small enough, so that the minority (tending *46to vote cohesively) can likely vote its preferred candidate (rather than a consensus candidate) into office? See ante, at 33 (Souter, J., dissenting) (“[Ejmpirical studies confir[m] that. . . minority groups” constituting less than 50% of the voting population “regularly elect their preferred candidates with the help of modest crossover by members of the majority”); see also Pildes, Is Voting-Rights Law Now at War With Itself? Social Science and Voting Rights in the 2000s, 80 N. C. L. Rev. 1517, 1529-1535 (2002) (reviewing studies showing small but reliable crossover voting by whites in districts where minority voters have demonstrated the ability to elect their preferred candidates without constituting 50% of the population in that district). We can likely find a reasonably administrable mathematical formula more directly tied to the factors in question.

To take a possible example: Suppose we pick a numerical ratio that requires the minority voting age population to be twice as large as the percentage of majority crossover votes needed to elect the minority’s preferred candidate. We would calculate the latter (the percentage of majority crossover votes the minority voters need) to take account of both the percentage of minority voting age population in the district and the cohesiveness with which they vote. Thus, if minority voters account for 45% of the voters in a district and 89% of those voters tend to vote cohesively as a group, then the minority needs a crossover vote of about 20% of the majority voters to elect its preferred candidate. (Such a district with 100 voters would have 45 minority voters and 55 majority voters; 40 minority voters would vote for the minority group’s preferred candidate at election time; the minority voters would need 11 more votes to elect their preferred candidate; and 11 is about 20% of the majority’s 55.) The larger the minority population, the greater its cohesiveness, and thus the smaller the crossover vote needed to assure success, the greater the likelihood that the minority can *47elect its preferred candidate and the smaller the likelihood that the cohesive minority, in order to find the needed majority crossover vote, must support a consensus, rather than its preferred, candidate.

In reflecting the reality that minority voters can elect the candidate of their choice when they constitute less than 50% of a district by relying on a small majority crossover vote, this approach is in no way contradictory to, or even in tension with, the third Gingles requirement. Since Gingles itself, we have acknowledged that the requirement of majority-bloc voting can be satisfied even when some small number of majority voters cross over to support a minority-preferred candidate. See 478 U. S., at 59 (finding majority-bloc voting where the majority group supported African-American candidates in the general election at a rate of between 26% and 49%, with an average support of one-third). Given the difficulty of obtaining totally accurate statistics about cohesion, or even voting age population, the district courts should administer the numerical ratio flexibly, opening (or closing) the Gingles gate (in light of the probable merits of a case) where only small variances are at issue (e. g., where the minority group is 39% instead of 40% of a district). But the same is true with a 50% number (e. g., where the minority group is 49% instead of 50% of a district). See, e. g., Brief for United States as Amicus Curiae 15.

I do not claim that the 2-to-l ratio is a perfect rule; I claim only that it is better than the plurality’s 50% rule. After all, unlike 50%, a 2-to-l ratio (of voting age minority population to necessary nonminority crossover votes) focuses directly upon the problem at hand, better reflects voting realities, and consequently far better separates at the gateway likely sheep from likely goats. See Gingles, supra, at 45 (The §2 inquiry depends on a “ ‘functional’ view of the political process” and “‘a searching practical evaluation of the past and present reality’ ” (quoting S. Rep. No. 97-417, p. 30, and n. 120 (1982))); Gingles, supra, at 94-95 (O’Connor, J., *48concurring in judgment) (“[Tjhere is no indication that Congress intended to mandate a single, universally applicable standard for measuring undiluted minority voting strength, regardless of local conditions ... ”). In most cases, the 50% rule and the 2-to-l rule would have roughly similar effects. Most districts where the minority voting age population is greater than 50% will almost always satisfy the 2-to-l rule; and most districts where the minority population is below 40% will almost never satisfy the 2-to-l rule. But in districts with minority voting age populations that range from 40% to 50%, the divergent approaches of the two standards can make a critical difference — as well they should.

In a word, Justice Soutek well explains why the majority’s test is ill suited to the statute’s objectives. I add that the test the majority adopts is ill suited to its own administrative ends. Better gateway tests, if needed, can be found.

With respect, I dissent.

announced the judgment of the Court and delivered an opinion, in which The Chief Justice and Justice Alito join.

This case requires us to interpret §2 of the Voting Rights Act of 1965,79 Stat. 487, as amended, 42 U. S. C. § 1973 (2000 ed.). The question is whether the statute can be invoked to require state officials to draw election-district lines to allow a racial minority to join with other voters to elect the minority’s candidate of choice, even where the racial minority is less than 50 percent of the voting-age population in the district to be drawn. To use election-law terminology: In a district that is not a majority-minority district, if a racial minority could elect its candidate of choice with support from crossover majority voters, can § 2 require the district to be drawn to accommodate this potential?

I

The ease arises in a somewhat unusual posture. State authorities who created a district now invoke the Voting Rights *7Act as a defense. They argue that §2 required them to draw the district in question in a particular way, despite state laws to the contrary. The state laws are provisions of the North Carolina Constitution that prohibit the General Assembly from dividing counties when drawing legislative districts for the State House and Senate. Art. II, §§3, 5. We will adopt the term used by the state courts and refer to both sections of the State Constitution as the Whole County Provision. See Pender County v. Bartlett, 361 N. C. 491, 493, 649 S. E. 2d 364, 366 (2007) (case below).

It is common ground that state election-law requirements like the Whole County Provision may be superseded by federal law — for instance, the one-person, one-vote principle of the Equal Protection Clause of the United States Constitution. See Reynolds v. Sims, 377 U. S. 533 (1964). Here the question is whether §2 of the Voting Rights Act requires district lines to be drawn that otherwise would violate the Whole County Provision. That, in turn, depends on how the statute is interpreted.

We begin with the election district. The North Carolina House of Representatives is the larger of the two chambers in the State’s General Assembly. District 18 of that body lies in the southeastern part of North Carolina. Starting in 1991, the General Assembly drew District 18 to include portions of four counties, including Pender County, in order to create a district with a majority African-American voting-age population and to satisfy the Voting Rights Act. Following the 2000 census, the North Carolina Supreme Court, to comply with the Whole County Provision, rejected the General Assembly’s first two statewide redistricting plans. See Stephenson v. Bartlett, 355 N. C. 354, 375, 562 S. E. 2d 377, 392, stay denied, 535 U. S. 1301 (2002) (Rehnquist, C. J., in chambers); Stephenson v. Bartlett, 357 N. C. 301, 314, 582 S. E. 2d 247, 254 (2003).

District 18 in its present form emerged from the General Assembly’s third redistricting attempt, in 2003. By that *8time the African-American voting-age population had fallen below 50 percent in the district as then drawn, and the General Assembly no longer could draw a geographically compact majority-minority district. Rather than draw District 18 to keep Pender County whole, however, the General Assembly drew it by splitting portions of Pender and New Hanover counties. District 18 has an African-American voting-age population of 39.36 percent. App. 139. Had it left Pender County whole, the General Assembly could have drawn District 18 with an African-American voting-age population of 35.33 percent. Id., at 73. The General Assembly’s reason for splitting Pender County was to give African-American voters the potential to join with majority voters to elect the minority group’s candidate of its choice. Ibid. Failure to do so, state officials now submit, would have diluted the minority group’s voting strength in violation of § 2.

In May 2004, Pender County and the five members of its board of commissioners filed the instant suit in North Carolina state court against the Governor of North Carolina, the Director of the State Board of Elections, and other state officials. The plaintiffs alleged that the 2003 plan violated the Whole County Provision by splitting Pender County into two House districts. Id., at 5-14. The state-official defendants answered that dividing Pender County was required by § 2. Id., at 25. As the trial court recognized, the procedural posture of this case differs from most §2 cases. Here the defendants raise § 2 as a defense. As a result, the trial court stated, they are “in the unusual position” of bearing the burden of proving that a §2 violation would have occurred absent splitting Pender County to draw District 18. App. to Pet. for Cert, 90a.

The trial court first considered whether the defendant state officials had established the three threshold requirements for § 2 liability under Thornburg v. Gingles, 478 U. S. 30, 50-51 (1986) — namely, (1) that the minority group “is suf*9ficiently large and geographically compact to constitute a majority in a single-member district,” (2) that the minority group is “politically cohesive,” and (3) “that the white majority votes sufficiently as a bloc to enable it . . . usually to defeat the minority’s preferred candidate.”

As to the first Gingles requirement, the trial court concluded that, although African-Americans were not a majority of the voting-age population in District 18, the district was a “de facto” majority-minority district because African-Americans could get enough support from crossover majority voters to elect the African-Americans’ preferred candidate. The court ruled that African-Americans in District 18 were politically cohesive, thus satisfying the second requirement. And later, the plaintiffs stipulated that the third Gingles requirement was met. App. to Pet. for Cert. 102a-103a, 130a. The court then determined, based on the totality of the circumstances, that §2 required the General Assembly to split Pender County. The court sustained the lines for District 18 on that rationale. Id., at 116a-118a.

Three of the Pender County Commissioners appealed the trial court’s ruling that the defendants had established the first Gingles requirement. The Supreme Court of North Carolina reversed. It held that a “minority group must constitute a numerical majority of the voting population in the area under consideration before Section 2 . . . requires the creation of a legislative district to prevent dilution of the votes of that minority group.” 361 N. C., at 502, 649 S. E. 2d, at 371. On that premise the State Supreme Court determined District 18 was not mandated by § 2 because African-Americans do not “constitute a numerical majority of citizens of voting age.” Id., at 507, 649 S. E. 2d, at 374. It ordered the General Assembly to redraw District 18. Id,., at 510, 649 S. E. 2d, at 376.

We granted certiorari, 552 U. S. 1256 (2008), and now affirm.

*10II

Passage of the Voting Rights Act of 1965 was an important step in the struggle to end discriminatory treatment of minorities who seek to exercise one of the most fundamental rights of our citizens: the right to vote. Though the Act as a whole was the subject of debate and controversy, §2 prompted little criticism. The likely explanation for its general acceptance is that, as first enacted, §2 tracked, in part, the text of the Fifteenth Amendment. It prohibited practices “imposed or applied by any State or political subdivision to deny or abridge the right of any citizen of the United States to vote on account of race or color.” 79 Stat. 437; cf. U. S. Const., Arndt. 15 (“The right of citizens of the United States to vote shall not be denied or abridged by the United States or by any State on account of race, color, or previous condition of servitude”); see also S. Rep. No. 162, 89th Cong., 1st Sess., pt. 3, pp. 19-20 (1965). In Mobile v. Bolden, 446 U. S. 55, 60-61 (1980), this Court held that §2, as it then read, “no more than elaborates upon . . . the Fifteenth Amendment” and was “intended to have an effect no different from that of the Fifteenth Amendment itself.”

In 1982, after the Mobile ruling, Congress amended §2, giving the statute its current form. The original Act had employed an intent requirement, prohibiting only those practices “imposed or applied ... to deny or abridge” the right to vote. 79 Stat. 437. The amended version of § 2 requires consideration of effects, as it prohibits practices “imposed or applied ... in a manner which results in a denial or abridgment” of the right to vote. 96 Stat. 134, 42 U. S. C. § 1973(a) (2000 ed.). The 1982 amendments also added a subsection, §2(b), providing a test for determining whether a §2 violation has occurred. The relevant text of the statute now states:

“(a) No voting qualification or prerequisite to voting or standard, practice, or procedure shall be imposed or *11applied by any State or political subdivision in a manner which results in a denial or abridgement of the right of any citizen of the United States to vote on account of race or color [or membership in a language minority group], as provided in subsection (b) of this section.

“(b) A violation of subsection (a) of this section is established if, based on the totality of circumstances, it is shown that the political processes leading to nomination or election in the State or political subdivision are not equally open to participation by members of a class of citizens protected by subsection (a) of this section in that its members have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice.” 42 U.S. C. §1973.

This Court first construed the amended version of §2 in Thornburg v. Gingles, 478 U. S. 30 (1986). In Gingles, the plaintiffs were African-American residents of North Carolina who alleged that multimember districts diluted minority voting strength by submerging black voters into the white majority, denying them an opportunity to elect a candidate of their choice. The Court identified three “necessary preconditions” for a claim that the use of multimember districts constituted actionable vote dilution under § 2: (1) The minority group must be “sufficiently large and geographically compact to constitute a majority in a single-member district,” (2) the minority group must be “politically cohesive,” and (3) the majority must vote “sufficiently as a bloc to enable it. . . usually to defeat the minority’s preferred candidate.” Id., at 50-51.

The Court later held that the three Gingles requirements apply equally in § 2 cases involving single-member districts, such as a claim alleging vote dilution because a geographically compact minority group has been split between two or more single-member districts. Growe v. Emison, 507 U. S. 25, 40-41 (1993). In a §2 case, only when a party has estab*12lished the Gingles requirements does a court proceed to analyze whether a violation has occurred based on the totality of the circumstances. Gingles, supra, at 79; see also Johnson v. De Grandy, 512 U. S. 997, 1013 (1994).

Ill

A

This case turns on whether the first Gingles requirement can be satisfied when the minority group makes up less than 50 percent of the voting-age population in the potential election district. The parties agree on all other parts of the Gingles analysis, so the dispositive question is: What size minority group is sufficient to satisfy the first Gingles requirement?

At the outset the answer might not appear difficult to reach, for the Gingles Court said the minority group must “demonstrate that it is sufficiently large and geographically compact to constitute a majority in a single-member district.” 478 U. S., at 50. This would seem to end the matter, as it indicates the minority group must demonstrate it can constitute “a majority.” But in Gingles and again in Growe the Court reserved what it considered to be a separate question — whether, “when a plaintiff alleges that a voting practice or procedure impairs a minority’s ability to influence, rather than alter, election results, a showing of geographical compactness of a minority group not sufficiently large to constitute a majority will suffice.” Growe, supra, at 41, n. 5; see also Gingles, supra, at 46-47, n. 12. The Court has since applied the Gingles requirements in §2 eases but has declined to decide the minimum size minority group necessary to satisfy the first requirement. See Voinovich v. Quilter, 507 U. S. 146, 154 (1993); De Grandy, supra, at 1009; League of United Latin American Citizens v. Perry, 548 U. S. 399, 443 (2006) (LULAC) (opinion of Kennedy, J.). We must consider the minimum-size question in this case.

*13It is appropriate to review the terminology often used to describe various features of election districts in relation to the requirements of the Voting Rights Act. In majority-minority districts, a minority group composes a numerical, working majority of the voting-age population. Under present doctrine, §2 can require the creation of these districts. See, e. g., Voinovich, supra, at 154 (“Placing black voters in a district in which they constitute a sizeable and therefore ‘safe’ majority ensures that they are able to elect their candidate of choice”); but see Holder v. Hall, 512 U. S. 874, 922-923 (1994) (Thomas, J., concurring in judgment). At the other end of the spectrum are influence districts, in which a minority group can influence the outcome of an election even if its preferred candidate cannot be elected. This Court has held that § 2 does not require the creation of influénee districts. LULAC, supra, at 445 (opinion of Kennedy, J.).

The present case involves an intermediate type of district — a so-called crossover district. Like an influence district, a crossover district is one in which minority voters make up less than a majority of the voting-age population. But in a crossover district, the minority population, at least potentially, is large enough to elect the candidate of its choice with help from voters who are members of the majority and who cross over to support the minority’s preferred candidate. 361 N. C., at 501-502, 649 S. E. 2d, at 371 (case below). This Court has referred sometimes to crossover districts as “coalitional” districts, in recognition of the necessary coalition between minority and crossover majority voters. See Georgia v. Ashcroft, 539 U. S. 461, 483 (2003); see also Pildes, Is Voting Rights Law Now at War With Itself? Social Science and Voting Rights in the 2000s, 80 N. C. L. Rev. 1517, 1539 (2002) (hereinafter Pildes). But that term risks confusion with coalition-district claims in which two minority groups form a coalition to elect the candidate of the coalition’s choice. See, e. g., Nixon v. Kent County, 76 F. 3d 1381, 1393 (CA6 1996) (en bane). We do not address that type of coali*14tion district here. The petitioners in the present case (the state officials who were the defendants in the trial court) argue that §2 requires a crossover district, in which minority voters might be able to persuade some members of the majority to cross over and join with them.

Petitioners argue that although crossover districts do not include a numerical majority of minority voters, they still satisfy the first Gingles requirement because they are “effective minority districts.” Under petitioners’ theory keeping Pender County whole would have violated §2 by cracking the potential crossover district that they drew as District 18. See Gingles, supra, at 46, n. 11 (vote dilution “may be caused by the dispersal of blacks into districts in which they constitute an ineffective minority of voters”). So, petitioners contend, §2 required them to override state law and split Pender County, drawing District 18 with an African-American voting-age population of 39.36 percent rather than keeping Pender County whole and leaving District 18 with an African-American voting-age population of 35.33 percent. We reject that claim.

First, we conclude, petitioners’ theory is contrary to the mandate of § 2. The statute requires a showing that minorities “have less opportunity than other members of the electorate to . . . elect representatives of their choice.” 42 U. S. C. § 1973(b) (2000 ed.). But because they form only 39 percent of the voting-age population in District 18, African-Americans standing alone have no better or worse opportunity to elect a candidate than does any other group of voters with the same relative voting strength. That is, African-Americans in District 18 have the opportunity to join other voters — including other racial minorities, or whites, or both — to reach a majority and elect their preferred candidate. They cannot, however, elect that candidate based on their own votes and without assistance from others. Recognizing a § 2 claim in this circumstance would grant minority voters “a right to preserve their strength for the purposes *15of forging an advantageous political alliance.” Hall v. Virginia, 385 F. 3d 421, 431 (CA4 2004); see also Voinovich, 507 U. S., at 154 (minorities in crossover districts “could not dictate electoral outcomes independently”). Nothing in §2 grants special protection to a minority group’s right to form political coalitions. “[Minority voters are not immune from the obligation to pull, haul, and trade to find common political ground.” De Grandy, 512 U. S., at 1020.

Although the Court has reserved the question we confront today and has cautioned that the Gingles requirements “cannot be applied mechanically,” Voinovich, supra, at 158, the reasoning of our cases does not support petitioners’ claims. Section 2 does not impose on those who draw election districts a duty to give minority voters the most potential, or the best potential, to elect a candidate by attracting crossover voters. In setting out the first requirement for §2 claims, the Gingles Court explained that “[ujnless minority voters possess the potential to elect representatives in the absence of the challenged structure or practice, they cannot claim to have been injured by that structure or practice.” 478 U. S., at 50, n. 17. The Growe Court stated that the first Gingles requirement is “needed to establish that the minority has the potential to elect a representative of its own choice in some single-member district.” 507 U. S., at 40. Without such a showing, “there neither has been a wrong nor can be a remedy.” Id., at 41. There is a difference between a racial minority group’s “own choice” and the choice made by a coalition. In Voinovich, the Court stated that the first Gingles requirement “would have to be modified or eliminated” to allow crossover-district claims. 507 U. S., at 158. Only once, in dicta, has this Court framed the first Gingles requirement as anything other than a majority-minority rule. See De Grandy, 512 U. S., at 1008 (requiring “a sufficiently large minority population to elect candidates of its choice”). And in the same case, the Court rejected the proposition, inherent in petitioners’ claim here, that § 2 enti*16ties minority groups to the maximum possible voting strength:

“[Rjeading §2 to define dilution as any failure to maximize tends to obscure the very object of the statute and to run counter to its textually stated purpose. One may suspect vote dilution from political famine, but one is not entitled to suspect (much less infer) dilution from mere failure to guarantee a political feast.” Id., at 1016-1017.

Allowing crossover-district claims would require us to revise and reformulate the Gingles threshold inquiry that has been the baseline of our §2 jurisprudence. Mandatory recognition of claims in which success for a minority depends upon crossover majority voters would create serious tension with the third Gingles requirement that the majority votes as a bloc to defeat minority-preferred candidates. It is difficult to see how the majority-bloc-voting requirement could be met in a district where, by definition, white voters join in sufficient numbers with minority voters to elect the minority’s preferred candidate. (We are skeptical that the bloc-voting test could be satisfied here, for example, where minority voters in District 18 cannot elect their candidate of choice without support from almost 20 percent of white voters. We do not confront that issue, however, because for some reason respondents conceded the third Gingles requirement in state court.)

As the Gingles Court explained, “in the absence of significant white bloc voting it cannot be said that the ability of minority voters to elect their chosen representatives is inferior to that of white voters.” 478 U. S., at 49, n. 15. Were the Court to adopt petitioners’ theory and dispense with the majority-minority requirement, the ruling would call in question the Gingles framework the Court has applied under §2. See LULAC, 548 U. S., at 490, n. 8. (Souter, X, concurring in part and dissenting in part) (“All aspects of our established analysis for majority-minority districts in Gingles and *17its progeny may have to be rethought in analyzing ostensible coalition districts”); cf. Metts v. Murphy, 363 F. 3d 8, 12 (CA1 2004) (en banc) (per curiam) (allowing influence-district claim to survive motion to dismiss but noting “there is tension in this case for plaintiffs in any effort to satisfy both the first and third prong of Gingles”).

We find support for the majority-minority requirement in the need for workable standards and sound judicial and legislative administration. The rule draws clear lines for courts and legislatures alike. The same cannot be said of a less exacting standard that would mandate crossover districts under § 2. Determining whether a § 2 claim would lie — i. e,, determining whether potential districts could function as crossover districts — would place courts in the untenable position of predicting many political variables and tying them to race-based assumptions. The Judiciary would be directed to make predictions or adopt premises that even experienced polling analysts and political experts could not assess with certainty, particularly over the long term. For example, courts would be required to pursue these inquiries: What percentage of white voters supported minority-preferred candidates in the past? How reliable would the crossover votes be in future elections? What types of candidates have white and minority voters supported together in the past and will those trends continue? Were past crossover votes based on incumbency and did that depend on race? What are the historical turnout rates among white and minority voters and will they stay the same? Those questions are speculative, and the answers (if they could be supposed) would prove elusive. A requirement to draw election districts on answers to these and like inquiries ought not to be inferred from the text or purpose of § 2. Though courts are capable of making refined and exacting factual inquiries, they “are inherently ill-equipped” to “make decisions based on highly political judgments” of the sort that crossover-district claims would require. Holder, 512 U. S., at 894 *18(Thomas, J., concurring in judgment). There is an underlying principle of fundamental importance: We must be most cautious before interpreting a statute to require courts to make inquiries based on racial classifications and race-based predictions. The statutory mandate petitioners urge us to find in § 2 raises serious constitutional questions. See infra, at 21-23.

Heightening these concerns even further is the fact that §2 applies nationwide to every jurisdiction that must draw lines for election districts required by state or local law. Crossover-district claims would require courts to make predictive political judgments not only about familiar, two-party contests in large districts but also about regional and local jurisdictions that often feature more than two parties or candidates. Under petitioners’ view courts would face the difficult task of discerning crossover patterns in nonpartisan contests for a city commission, a school board, or a local water authority. The political data necessary to make such determinations are nonexistent for elections in most of those jurisdictions. And predictions would be speculative at best given that, especially in the context of local elections, voters’ personal affiliations with candidates and views on particular issues can play a large role.

Unlike any of the standards proposed to allow crossover-district claims, the majority-minority rule relies on an objective, numerical test: Do minorities make up more than 50 percent of the voting-age population in the relevant geographic area? That rule provides straightforward guidance to courts and to those officials charged with drawing district lines to comply with §2. See LULAC, supra, at 485 (opinion of Souter, J.) (recognizing need for “clear-edged rule”). Where an election district could be drawn in which minority voters form a majority but such a district is not drawn, or where a majority-minority district is cracked by assigning some voters elsewhere, then — assuming the other Gingles factors are also satisfied — denial of the opportunity to elect *19a candidate of choice is a present and discernible wrong that is not subject to the high degree of speculation and prediction attendant upon the analysis of crossover claims. Not an arbitrary invention, the majority-minority rule has its foundation in principles of democratic governance. The special significance, in the democratic process, of a majority means it is a special wrong when a minority group has 50 percent or more of the voting population and could constitute a compact voting majority but, despite racially polarized bloc voting, that group is not put into a district.

Given the text of § 2, our cases interpreting that provision, and the many difficulties in assessing §2 claims without the restraint and guidance provided by the majority-minority rule, no federal court of appeals has held that §2 requires creation of coalition districts. Instead, all to consider the question have interpreted the first Gingles factor to require a majority-minority standard. See Hall, 385 F. 3d, at 427-430 (CA4 2004), cert. denied, 544 U. S. 961 (2005); Valdespino v. Alamo Heights Independent School Dist., 168 F. 3d 848, 852-853 (CA5 1999), cert. denied, 528 U.S. 1114 (2000); Cousin v. Sundquist, 145 F. 3d 818, 828-829 (CA6 1998), cert. denied, 525 U. S. 1138 (1999); Sanchez v. Colorado, 97 F. 3d 1303, 1311-1312 (CA10 1996), cert. denied, 520 U. S. 1229 (1997); Romero v. Pomona, 883 F. 2d 1418, 1424, n. 7, 1425-1426 (CA9 1989), overruled on other grounds, 914 F. 2d 1136, 1141 (CA9 1990); McNeil v. Springfield Park Dist., 851 F. 2d 937, 947 (CA7 1988), cert. denied, 490 U. S. 1031 (1989). Cf. Metts, supra, at 11 (expressing unwillingness “at the complaint stage to foreclose the possibility” of influence-district claims). We decline to depart from the uniform interpretation of § 2 that has guided federal courts and state and local officials for more than 20 years.

To be sure, the Gingles requirements “cannot be applied mechanically and without regard to the nature of the claim.” Voinovich, 507 U. S., at 158. It remains the rule, however, that a party asserting § 2 liability must show by a preponder*20anee of the evidence that the minority population in the potential election district is greater than 50 percent. No one contends that the African-American voting-age population in District 18 exceeds that threshold. Nor does this case involve allegations of intentional and wrongful conduct. We therefore need not consider whether intentional discrimination affects the Gingles analysis. Cf. Brief for United States as Amicus Curiae 14 (evidence of discriminatory intent “tends to suggest that the jurisdiction is not providing an equal opportunity to minority voters to elect the representative of their choice, and it is therefore unnecessary to consider the majority-minority requirement before proceeding to the ultimate totality-of-the-circumstanees analysis”); see also Garza v. County of Los Angeles, 918 F. 2d 763, 771 (CA9 1990). Our holding does not apply to cases in which there is intentional discrimination against a racial minority.

B

In arguing for a less restrictive interpretation of the first Gingles requirement petitioners point to the text of §2 and its guarantee that political processes be “equally open to participation” to protect minority voters’ “opportunity... to elect representatives of their choice.” 42 U. S. C. § 1973(b) (2000 ed.). An “opportunity,” petitioners argue, occurs in crossover districts as well as majority-minority districts; and these extended opportunities, they say, require §2 protection.

But petitioners put emphasis on the word “opportunity” at the expense of the word “equally.” The statute does not protect any possible opportunity or mechanism through which minority voters could work with other constituencies to elect their candidate of choice. Section 2 does not guarantee minority voters an electoral advantage. Minority groups in crossover districts cannot form a voting majority without crossover voters. In those districts minority voters have the same opportunity to elect their candidate as any other political group with the same relative voting strength.

*21The majority-minority rule, furthermore, is not at odds with §2’s totality-of-the-circumstances test. The Court in De Grandy confirmed “the error of treating the three Gingles conditions as exhausting the enquiry required by §2.” 512 U. S., at 1018. Instead the Gingles requirements are preconditions, consistent with the text and purpose of § 2, to help courts determine which claims could meet the totality-of-the-circumstances standard for a § 2 violation. See Growe, 507 U. S., at 40 (describing the “Gingles threshold factors”).

To the extent there is any doubt whether § 2 calls for the majority-minority rule, we resolve that doubt by avoiding serious constitutional concerns under the Equal Protection Clause. See Clark v. Martinez, 543 U. S. 371, 381-382 (2005) (canon of constitutional avoidance is “a tool for choosing between competing plausible interpretations of a statutory text, resting on the reasonable presumption that Congress did not intend the alternative which raises serious constitutional doubts”). Of course, the “moral imperative of racial neutrality is the driving force of the Equal Protection Clause,” and racial classifications are permitted only “as a last resort.” Richmond v. J. A. Croson Co., 488 U. S. 469, 518, 519 (1989) (Kennedy, J., concurring in part and concurring in judgment). “Racial classifications with respect to voting earry particular dangers. Racial gerrymandering, even for remedial purposes, may balkanize us into competing racial factions; it threatens to carry us further from the goal of a political system in which race no longer matters — a goal that the Fourteenth and Fifteenth Amendments embody, and to which the Nation continues to aspire.” Shaw v. Reno, 509 U. S. 630, 657 (1993). If §2 were interpreted to require crossover districts throughout the Nation, “it would unnecessarily infuse race into virtually every redistricting, raising serious constitutional questions.” LULAC, 548 U. S., at 446 (opinion of Kennedy, J.); see also Ashcroft, 539 U. S., at 491 (Kennedy, J., concurring). That interpretation would result in a substantial increase in the number of mandatory *22districts drawn with race as “the predominant factor motivating the legislature’s decision.” Miller v. Johnson, 515 U. S. 900, 916 (1995).

On petitioners’ view of the case courts and legislatures would need to scrutinize every factor that enters into districting to gauge its effect on crossover voting. Injecting this racial measure into the nationwide districting process would be of particular concern with respect to consideration of party registration or party influence. The easiest and most likely alliance for a group of minority voters is one with a political party, and some have suggested using minority voters’ strength within a particular party as the proper yardstick under the first Gingles requirement. See, e.g., LULAC, supra, at 485-486 (opinion of Souter, J.) (requiring only “that minority voters ... constitute a majority of those voting in the primary of... the party tending to win in the general election”). That approach would replace an objective, administrable rule with a difficult “judicial inquiry into party rules and local politics” to determine whether a minority group truly “controls” the dominant party’s primary process. McLoughlin, Gingles in Limbo: Coalitional Districts, Party Primaries and Manageable Vote Dilution Claims, 80 N. Y. U. L. Rev. 312, 349 (2005). More troubling still is the inquiry's fusion of race and party affiliation as a determinant when partisan considerations themselves may be suspect in the drawing of district lines. See Vieth v. Jubelirer, 541 U. S. 267, 317 (2004) (STEVENS, J., dissenting); id., at 316 (Kennedy, J., concurring in judgment); see also Pildes 1565 (crossover-district requirement would essentially result in political party “entitlement to ... a certain number of seats”). Disregarding the majority-minority rule and relying on a combination of race and party to presume an effective majority would involve the law and courts in a perilous enterprise. It would rest on judicial predictions, as a matter of law, that race and party would hold together as an effective majority over time — at least for the decennial apportion*23ment cycles and likely beyond. And thus would the relationship between race and party further distort and frustrate the search for neutral factors and principled rationales for districting.

Petitioners’ approach would reverse the canon of avoidance. It invites the divisive constitutional questions that are both unnecessary and contrary to the purposes of our precedents under the Voting Rights Act. Given the consequences of extending racial considerations even further into the districting process, we must not interpret § 2 to require crossover districts.

C

Our holding that §2 does not require crossover districts does not consider the permissibility of such districts as a matter of legislative choice or discretion. Assuming a majority-minority district with a substantial minority population, a legislative determination, based on proper factors, to create two crossover districts may serve to diminish the significance and influence of race by encouraging minority and majority voters to work together toward a common goal. The option to draw such districts gives legislatures a choice that can lead to less racial isolation, not more. And as the Court has noted in the context of §5 of the Voting Rights Act, “various studies have suggested that the most effective way to maximize minority voting strength may be to create more influence or [crossover] districts.” Ashcroft, 539 U. S., at 482. Much like § 5, § 2 allows States to choose their own method of complying with the Voting Rights Act, and we have said that may include drawing crossover districts. See id., at 480-483. When we address the mandate of § 2, however, we must note it is not concerned with maximizing minority voting strength, De Grandy, supra, at 1022; and, as a statutory matter, § 2 does not mandate creating or preserving crossover districts.

Our holding also should not be interpreted to entrench majority-minority districts by statutory command, for that, *24too, could pose constitutional concerns. See Miller v. Johnson, supra; Shaw v. Reno, 509 U. S. 630. States that wish to draw crossover districts are free to do so where no other prohibition exists. Majority-minority districts are only required if all three Gingles factors are met and if § 2 applies based on a totality of the circumstances. In areas with substantial crossover voting it is unlikely that the plaintiffs would be able to establish the third Gingles precondition— bloc voting by majority voters. See supra, at 16. In those areas majority-minority districts would not be required in the first place; and in the exercise of lawful discretion States could draw crossover districts as they deemed appropriate. See Pildes 1567 (“Districts could still be designed in such places that encouraged coalitions across racial lines, but these districts would result from legislative choice, not. . . obligation”). States can — and in proper cases should — defend against alleged §2 violations by pointing to crossover voting patterns and to effective crossover districts. Those can be evidence, for example, of diminished bloc voting under the third Gingles factor or of equal political opportunity under the §2 totality-of-the-cireumstances analysis. And if there were a showing that a State intentionally drew district lines in order to destroy otherwise effective crossover districts, that would raise serious questions under both the Fourteenth and Fifteenth Amendments. See Reno v. Bossier Parish School Bd., 520 U. S. 471, 481-482 (1997); Brief for United States as Amicus Curiae 13-14. There is no evidence of discriminatory intent in this case, however. Our holding recognizes only that there is no support for the claim that §2 can require the creation of crossover districts in the first instance.

Petitioners claim the majority-minority rule is inconsistent with § 5, but we rejected a similar argument in LULAC, 548 U. S., at 446 (opinion of Kennedy, J.). The inquiries under §§2 and 5 are different. Section 2 concerns minority *25groups’ opportunity “to elect representatives of their choice,” 42 U. S. C. § 1973(b) (2000 ed.), while the more stringent § 5 asks whether a change has the purpose or effect of “denying or abridging the right to vote,” § 1973c. See LULAC, supra, at 446; Bossier Parish, supra, at 476-480. In LULAC, we held that although the presence of influence districts is relevant for the §5 retrogression analysis, “the lack of such districts cannot establish a §2 violation.” 548 U. S., at 446 (opinion of Kennedy, J.); see also Ashcroft, 539 U. S., at 482-483. The same analysis applies for crossover districts: Section 5 “leaves room” for States to employ crossover districts, id., at 483, but §2 does not require them.

IV

Some commentators suggest that racially polarized voting is waning — as evidenced by, for example, the election of minority candidates where a majority of voters are white. See Note, The Future of Majority-Minority Districts in Light of Declining Racially Polarized Voting, 116 Harv. L. Rev. 2208, 2209 (2003); see also id., at 2216-2222; Pildes 1529-1539; Bullock & Dunn, The Demise of Racial Districting and the Future of Black Representation, 48 Emory L. J. 1209 (1999). Still, racial discrimination and racially polarized voting are not ancient history. Much remains to be done to ensure that citizens of all races have equal opportunity to share and participate in our democratic processes and traditions; and §2 must be interpreted to ensure that continued progress.

It would be an irony, however, if §2 were interpreted to entrench racial differences by expanding a “statute meant to hasten the waning of racism in American politics.” De Grandy, 512 U. S., at 1020. Crossover districts are, by definition, the result of white voters joining forces with minority voters to elect their preferred candidate. The Voting Rights Act was passed to foster this cooperation. We decline now to expand the reaches of § 2 to require, by force of *26law, the voluntary cooperation our society has achieved. Only when a geographically compact group of minority voters could form a majority in a single-member district has the first Gingles requirement been met.

The judgment of the Supreme Court of North Carolina is affirmed.

It is so ordered.

I continue to adhere to the views expressed in my opinion in Holder v. Hall, 512 U. S. 874, 891 (1994) (opinion concurring in judgment). The text of § 2 of the Voting Rights Act of 1965 does not authorize any vote dilution claim, regardless of the size of the minority population in a given district. See 42 U. S. C. § 1973(a) (2000 ed.) (permitting only a challenge to a “voting qualification or prerequisite to voting or standard, practice, or procedure”); see also Holder, supra, at 893 (stating that the terms “‘standard, practice, or procedure’” “reach only state enactments that limit citizens’ access to the ballot”). I continue to disagree, therefore, with the framework set forth in Thornburg v. Gingles, 478 U. S. 30 (1986), for analyzing vote dilution claims because it has no basis in the text of §2. I would not evaluate any Voting Rights Act claim under a test that “has produced such a disastrous misadventure in judicial policymaking.” Holder, supra, at 893. For these reasons, I concur only in the judgment.

The question in this case is whether a minority with under 50% of the voting population of a proposed voting district can ever qualify under §2 of the Voting Rights Act of 1965 (VRA) as residents of a putative district whose minority vot*27ers would have an opportunity “to elect representatives of their choice.” 42 U. S. C. § 1973(b) (2000 ed.). If the answer is no, minority voters in such a district will have no right to claim relief under §2 from a statewide districting scheme that dilutes minority voting rights. I would hold that the answer in law as well as in fact is sometimes yes: a district may be a minority-opportunity district so long as a cohesive minority population is large enough to elect its chosen candidate when combined with a reliable number of crossover voters from an otherwise polarized majority.

In the plurality’s view, only a district with a minority population making up 50% or more of the citizen voting age population (CVAP) can provide a remedy to minority voters lacking an opportunity “to elect representatives of their choice.” This is incorrect as a factual matter if the statutory phrase is given its natural meaning; minority voters in districts with minority populations under 50% routinely “elect representatives of their choice.” The effects of the plurality’s unwillingness to face this fact are disturbing by any measure and flatly at odds with the obvious purpose of the VRA. If districts with minority populations under 50% can never count as minority-opportunity districts to remedy a violation of the States’ obligation to provide equal electoral opportunity under § 2, States will be required under the plurality’s rule to pack black voters into additional majority-minority districts, contracting the number of districts where racial minorities are having success in transcending racial divisions in securing their preferred representation. The object of the VRA will now be promoting racial blocs, and the role of race in districting decisions as a proxy for political identification will be heightened by any measure.

I

Recalling the basic premises of vote-dilution claims under §2 will show just how far astray the plurality has gone. *28Section 2 of the VR A prohibits districting practices that “result] in a denial or abridgement of the right of any citizen of the United States to vote on account of race.” 42 U. S. C. § 1973(a). A denial or abridgment is established if, “based on the totality of circumstances,” it is shown that members of a racial minority “have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice.” § 1973(b).

Since § 2 was amended in 1982, 96 Stat. 134, we have read it to prohibit practices that result in “vote dilution,” see Thornburg v. Gingles, 478 U. S. 30 (1986), understood as distributing politically cohesive minority voters through voting districts in ways that reduce their potential strength. See id., at 47-48. There are two classic patterns. Where voting is racially polarized, a districting plan can systemically discount the minority vote either “by the dispersal of blacks into districts in which they constitute an ineffective minority of voters” or from “the concentration of blacks into districts where they constitute an excessive majority,” so as to eliminate their influence in neighboring districts. Id., at 46, n. 11. Treating dilution as a remediable harm recognizes that §2 protects not merely the right of minority voters to put ballots in a box, but to claim a fair number of districts in which their votes can be effective. See id., at 47.

Three points follow. First, to speak of a fair chance to get the representation desired, there must be an identifiable baseline for measuring a group’s voting strength. Id., at 88 (O’Connor, J., concurring in judgment) (“In order to evaluate a claim that a particular multimember district or single-member district has diluted the minority group’s voting strength to a degree that violates §2, . . . it is . . . necessary to construct a measure of 'undiluted' minority voting strength”). Several baselines can be imagined; one could, for example, compare a minority’s voting strength under a particular districting plan with the maximum strength possi*29ble under any alternative.1 Not surprisingly, we have conclusively rejected this approach; the VRA was passed to guarantee minority voters a fair game, not a killing. See Johnson v. De Grandy, 512 U. S. 997, 1016-1017 (1994). We have held that the better baseline for measuring opportunity to elect under § 2, although not dispositive, is the minority’s rough proportion of the relevant population. Id., at 1013-1023. Thus, in assessing §2 claims under a totality of the circumstances, including the facts of history and geography, the starting point is a comparison of the number of districts where minority voters can elect their chosen candidate with the group’s population percentage. Ibid.; see also League of United Latin American Citizens v. Perry, 548 U. S. 399, 436 (2006) (LULAC) (“We proceed now to the totality of the circumstances, and first to the proportionality inquiry, comparing the percentage of total districts that are [minority] opportunity districts with the [minority] share of the citizen voting-age population”).2

*30Second, the significance of proportionality means that a § 2 claim must be assessed by looking at the overall effect of a multidistrict plan. A State with one congressional seat cannot dilute a minority’s congressional vote, and only the systemic submergence of minority votes where a number of single-member districts could be drawn can be treated as harm under §2. So a §2 complaint must look to an entire districting plan (normally, statewide), alleging that the challenged plan creates an insufficient number of minority-opportunity districts in the territory as a whole. See id., at 436-437.

Third, while a § 2 violation ultimately results from the dilutive effect of a districting plan as a whole, a §2 plaintiff must also be able to place himself in a reasonably compact district that could have been drawn to improve upon the plan actually selected. See, e.g., De Grandy, supra, at 1001-1002. That is, a plaintiff must show both an overall deficiency and a personal injury open to redress.

Our first essay at understanding these features of statutory vote dilution was Thornburg v. Gingles, which asked whether a multimember district plan for choosing representatives by at-large voting deprived minority voters of an equal opportunity to elect their preferred candidates. In answering, we set three now-familiar conditions that a §2 claim must meet at the threshold before a court will analyze it under the totality of circumstances:

“First, the minority group must be able to demonstrate that it is sufficiently large and geographically compact to constitute a majority in a single-member district.... Second, the minority group must be able to show that it is politically cohesive. . . . Third, the minority must be able to demonstrate that the white majority votes sufficiently as a bloc to enable it . . . usually to defeat the minority’s preferred candidate.” 478 U. S., at 50-51.

*31As we have emphasized over and over, the Gingles conditions do not state the ultimate standard under §2, nor could they, since the totality of the circumstances standard has been set explicitly by Congress. See LULAC, supra, at 425-426; De Grandy, supra, at 1011. Instead, each condition serves as a gatekeeper, ensuring that a plaintiff who proceeds to plenary review has a real chance to show a redressable violation of the ultimate §2 standard. The third condition, majority racial bloc voting, is necessary to establish the premise of vote-dilution claims: that the minority as a whole is placed at a disadvantage owing to race, not the happenstance of independent politics. Gingles, 478 U. S., at 51. The second, minority cohesion, is there to show that minority voters will vote together to elect a distinct representative of choice. Ibid. And the first, a large and geographically compact minority population, is the condition for demonstrating that a dilutive plan injures the §2 plaintiffs by failing to draw an available remedial district that would give them a chance to elect their chosen candidate. Growe v. Emison, 507 U. S. 25, 40-41 (1993); Gingles, supra, at 50.

II

Though this case arose under the Constitution of North Carolina, the dispositive issue is one of federal statutory law: whether a district with a minority population under 50%, but large enough to elect its chosen candidate with the help of majority voters disposed to support the minority favorite, can ever count as a district where minority voters have the opportunity “to elect representatives of their choice” for purposes of §2. I think it clear from the nature of a vote-dilution claim and the text of §2 that the answer must be yes. There is nothing in the statutory text to suggest that Congress meant to protect minority opportunity to elect solely by the creation of majority-minority districts. See Voinovich v. Quilter, 507 U. S. 146, 155 (1993) (“[Section 2] *32says nothing about majority-minority districts”). On the contrary, §2 “focuses exclusively on the consequences of apportionment,” ibid., as Congress made clear when it explicitly prescribed the ultimate functional approach: a totality of the circumstances test. See 42 U. S. C. § 1973(b) (“[a] violation ... is established if, based on the totality of circumstances, it is shown ... ”). And a functional analysis leaves no doubt that crossover districts vindicate the interest expressly protected by §2: the opportunity to elect a desired representative.

It has been apparent from the moment the Court first took up §2 that no reason exists in the statute to treat a crossover district as a less legitimate remedy for dilution than a majority-minority one (let alone to rule it out). See Gingles, supra, at 90, n. 1 (O’Connor, J., concurring in judgment) (“[I]f a minority group that is not large enough to constitute a voting majority in a single-member district can show that white support would probably... enable the election of the candidates its members prefer, that minority group would appear to have demonstrated that, at least under this measure of its voting strength, it would be able to elect some candidates of its choice”); see also Pildes, Is Voting-Rights Law Now at War With Itself? Social Science and Voting Rights in the 2000s, 80 N. C. L. Rev. 1517, 1553 (2002) (hereinafter Pildes) (“What should be so magical, then, about whether there are enough black voters to become a formal majority so that a conventional ‘safe’ district can be created? If a safe and a coalitional district have the same probability of electing a black candidate, are they not functionally identical, by definition, with respect to electing such candidates?”).

As these earlier comments as much as say, whether a district with a minority population under 50% of the CVAP may redress a violation of § 2 is a question of fact with an obvious answer: of course minority voters constituting less than 50% of the voting population can have an opportunity to elect the *33candidates of their choice, as amply shown by empirical studies confirming that such minority groups regularly elect their preferred candidates with the help of modest crossover by members of the majority. See, e.g., id., at 1531-1534, 1538. The North Carolina Supreme Court, for example, determined that voting districts with a black voting age population of as little as 38.37% have an opportunity to elect black candidates, Pender Cty. v. Bartlett, 361 N. C. 491, 494-495, 649 S. E. 2d 364, 366-367 (2007), a factual finding that has gone unchallenged and is well supported by electoral results in North Carolina. Of the nine House districts in which blacks make up more than 50% of the voting age population (VAP), all but two elected a black representative in the 2004 election. See App. 109. Of the 12 additional House districts in which blacks are over 39% of the VAP, all but one elected a black representative in the 2004 election. Ibid. It would surely surprise legislators in North Carolina to suggest that black voters in these 12 districts cannot possibly have an opportunity to “elect [the] representatives of their choice.”

It is of course true that the threshold population sufficient to provide minority voters with an opportunity to elect their candidates of choice is elastic, and the proportions 'will likely shift in the future, as they have in the past. See Pildes 1527-1532 (explaining that blacks in the 1980s required well over 50% of the population in a district to elect the candidates of their choice, but that this number has gradually fallen to well below 50%); id., at 1527, n. 26 (stating that some courts went so far as to refer to 65% “as a ‘rule of thumb’ for the black population required to constitute a safe district”). That is, racial polarization has declined, and if it continues downward the first Gingles condition will get easier to satisfy.

But this is no reason to create an arbitrary threshold; the functional approach will continue to allow dismissal of claims for districts with minority populations too small to demon*34strate an ability to elect, and with “crossovers” too numerous to allow an inference of vote dilution in the first place. No one, for example, would argue based on the record of experience in this case that a district with a 25% black population would meet the first Gingles condition. And the third Gingles requirement, majority-bloc voting, may well provide an analytical limit to claims based on crossover districts. See LULAC, 548 U. S., at 490, n. 8 (Souter, J., concurring in part and dissenting in part) (noting the interrelationship of the first and third Gingles factors); see also post, at 44-48 (Breyer, J., dissenting) (looking to the third Gingles condition to suggest a mathematical limit to the minority population necessary for a cognizable crossover district). But whatever this limit may be, we have no need to set it here, since the respondent state officials have stipulated to majority-bloc voting, App. to Pet. for Cert. 130a. In sum, §2 addresses voting realities, and for practical purposes a 89%-minority district in which we know minorities have the potential to elect their preferred candidate is every bit as good as a 50%-minority district.

In fact, a crossover district is better. Recognizing crossover districts has the value of giving States greater flexibility to draw districting plans with a fair number of minority-opportunity districts, and this in turn allows for a beneficent reduction in the number of majority-minority districts with their “quintessentially race-conscious calculus,” De Grandy, 512 U. S., at 1020, thereby moderating reliance on race as an exclusive determinant in districting decisions, cf. Shaw v. Reno, 509 U. S. 630 (1993). See also Pildes 1547-1548 (“In contrast to the Court’s concerns with bizarrely designed safe districts, it is hard to see how coalitional districts could ‘convey the message that political identity is, or should be, predominantly racial.’ . . . Coalitional districts would seem to encourage and require a kind of integrative, cross-racial political alliance that might be thought consistent with, even the very ideal of, both the VRA and the U. S. Constitution” (quoting Bush v. Vera, 517 U. S. 952, 980 (1996))). A cross*35over is thus superior to a majority-minority district precisely because it requires polarized factions to break out of the mold and form the coalitions that discourage racial divisions.

Ill

A

The plurality’s contrary conclusion that § 2 does not recognize a crossover claim is based on a fundamental misunderstanding of vote-dilution claims, a mistake epitomized in the following assessment of the crossover district in question:

“[Bjecause they form only 39 percent of the voting-age population in District 18, African-Americans standing alone have no better or worse opportunity to elect a candidate than does any other group of voters with the same relative voting strength [in District 18].” Ante, at 14.

See also ante, at 20 (“[In crossover districts,] minority voters have the same opportunity to elect their candidate as any other political group with the same relative voting strength”).

The claim that another political group in a particular district might have the same relative voting strength as the minority if it had the same share of the population takes the form of a tautology: the plurality simply looks to one district and says that a 39% group of blacks is no worse off than a 39% group of whites would be. This statement might be true, or it might not be, and standing alone it demonstrates nothing.

Even if the two 39% groups were assumed to be comparable in fact because they will attract sufficient crossover (and so should be credited with satisfying the first Gingles condition), neither of them could prove a §2 violation without looking beyond the 39% district and showing a disproportionately small potential for success in the State’s overall configuration of districts. As this Court has explained before, the ultimate question in a § 2 case (that is, whether the *36minority group in question is being denied an equal opportunity to participate and elect) can be answered only by examining the broader pattern of districts to see whether the minority is being denied a roughly proportionate opportunity. See LULAC, supra, at 436-437. Hence, saying one group’s 39% equals another’s, even if true in particular districts where facts are known, does not mean that either, both, or neither group could show a § 2 violation. The plurality simply fails to grasp that an alleged § 2 violation can only be proved or disproved by looking statewide.

B

The plurality’s more specific justifications for its counter-factual position are no more supportable than its 39% tautology.

1

The plurality seems to suggest that our prior cases somehow require its conclusion that a minority population under 50% will never support a § 2 remedy, emphasizing that Gingles spoke of a majority and referred to the requirement that minority voters have “ ‘the potential to elect’ ” their chosen representatives. Ante, at 15 (quoting Gingles, 478 U. S., at 50, n. 17). It is hard to know what to make of this point since the plurality also concedes that we have explicitly and repeatedly reserved decision on today’s question. See LULAC, supra, at 443 (plurality opinion); De Grandy, supra, at 1009; Voinovich, 507 U. S., at 154; Growe, 507 U. S., at 41, n. 5; Gingles, supra, at 46-47, n. 12. In fact, in our more recent cases applying § 2, Court majorities have formulated the first Gingles prong in a way more consistent with a functional approach. See LULAC, supra, at 430 (“[I]n the context of a challenge to the drawing of district lines, ‘the first Gingles condition requires the possibility of creating more than the existing number of reasonably compact districts with a sufficiently large minority population to elect candidates of its choice’ ” (quoting De Grandy, supra, at *371008)). These Court majorities get short shrift from today’s plurality.

In any event, even if we ignored Gingles’s reservation of today’s question and looked to Gingles’s “potential to elect” as if it were statutory text, I fail to see how that phrase dictates that a minority’s ability to compete must be single-handed in order to count under §2. As explained already, a crossover district serves the same interest in obtaining representation as a majority-minority district; the potential of 45% with a 6% crossover promises the same result as 51% with no crossover, and there is nothing in the logic of §2 to allow a distinction between the two types of district.

In fact, the plurality’s distinction is artificial on its own terms. In the past, when black voter registration and black voter turnout were relatively low, even black voters with 55% of a district’s CVAP would have had to rely on crossover voters to elect their candidate of choice. See Pildes 1527-1528. But no one on this Court (and, so far as I am aware, any other court addressing it) ever suggested that reliance on crossover voting in such a district rendered minority success any less significant under §2, or meant that the district failed to satisfy the first Gingles factor. Nor would it be any answer to say that black voters in such a district, assuming unrealistic voter turnout, theoretically had the “potential” to elect their candidate without crossover support; that would be about as relevant as arguing in the abstract that a black CVAP of 45% is potentially successful, on the assumption that black voters could turn out en masse to elect the candidate of their choice without reliance on crossovers if enough majority voters stay home.

2

The plurality is also concerned that recognizing the “potential” of anything under 50% would entail an exponential expansion of special minority districting; the plurality goes so far as to suggest that recognizing crossover districts as possible minority-opportunity districts would inherently “en*38titl[e] minority groups to the maximum possible voting strength.” Ante, at 15-16. But this conclusion again reflects a confusion of the gatekeeping function of the Gingles conditions with the ultimate test for relief under §2. See ante, at 14 (“African-Americans standing alone have no better or worse opportunity to elect a candidate than does any other group of voters with the same relative voting strength”).

As already explained, swpra, at 31, the mere fact that all threshold Gingles conditions could be met and a district could be drawn with a minority population sufficiently large to elect the candidate of its choice does not require drawing such a district. This case simply is about the first Gingles condition, not about the number of minority-opportunity districts needed under §2, and accepting Bartlett’s position would in no way imply an obligation to maximize districts with minority voter potential. Under any interpretation of the first Gingles factor, the State must draw districts in a way that provides minority voters with a fair number of districts in which they have an opportunity to elect candidates of their choice; the only question here is which districts will count toward that total.

3

The plurality’s fear of maximization finds a parallel in the concern that treating crossover districts as minority-opportunity districts would “create serious tension” with the third Gingles prerequisite of majority-bloc voting. Ante, at 16. The plurality finds “[i]t . . . difficult to see how the majority-bloc-voting requirement could be met in a district where, by definition, white voters join in sufficient numbers with minority voters to elect the minority’s preferred candidate.” Ibid.

It is not difficult to see. If a minority population with 49% of the CVAP can elect the candidate of its choice with crossover by 2% of white voters, the minority “by definition” relies on white support to elect its preferred candidate. But this fact alone would raise no doubt, as a matter of definition *39or otherwise, that the majority-bloc-voting requirement could be met, since as much as 98% of the majority may have voted against the minority’s candidate of choice. As explained above, supra, at 34, the third Gingles condition may well impose an analytical floor to the minority population and a ceiling on the degree of crossover allowed in a crossover district; that is, the concept of majority-bloc voting requires that majority voters tend to stick together in a relatively high degree. The precise standard for determining majority-bloc voting is not at issue in this case, however; to reflate the plurality’s 50% rule, one need only recognize that racial cohesion of 98% would be bloc voting by any standard.3

4

The plurality argues that qualifying crossover districts as minority-opportunity districts would be less administrable than demanding 50%, forcing courts to engage with the various factual and predictive questions that would come up in determining what percentage of majority voters would provide the voting minority with a chance at electoral success. Ante, at 17. But claims based on a State's failure to draw majority-minority districts raise the same issues of judicial judgment; even when the 50% threshold is satisfied, a court will still have to engage in factually messy enquiries about *40the “potential” such a district may afford, the degree of minority cohesion and majority-bloc voting, and the existence of vote dilution under a totality of the circumstances. See supra, at 30-31, 33-34. The plurality’s rule, therefore, conserves an uncertain amount of judicial resources, and only at the expense of ignoring a class of § 2 claims that this Court has no authority to strike from the statute’s coverage.

5

The plurality again misunderstands the nature of §2 in suggesting that its rule does not conflict with what the Court said in Georgia v. Ashcroft, 539 U. S. 461, 480-482 (2003): that crossover districts count as minority-opportunity districts for the purpose of assessing whether minorities have the opportunity “to elect their preferred candidates of choice” under § 5 of the VRA, 42 U. S. C. § 1973c(b) (2006 ed.). While the plurality is, of course, correct that there are differences between the enquiries under §§2 and 5, ante, at 24, those differences do not save today’s decision from inconsistency with the prior pronouncement. A districting plan violates § 5 if it diminishes the ability of minority voters to “elect their preferred candidates of choice,” §1973c(b), as measured against the minority’s previous electoral opportunity, Ashcroft, supra, at 477. A districting plan violates §2 if it diminishes the ability of minority voters to “elect representatives of their choice,” 42 U. S. C. § 1973(b) (2000 ed.), as measured under a totality of the circumstances against a baseline of rough proportionality. It makes no sense to say that a crossover district counts as a minority-opportunity district when comparing the past and the present under § 5, but not when comparing the present and the possible under § 2.

6

Finally, the plurality tries to support its insistence on a 50% threshold by invoking the policy of constitutional avoidance, which calls for construing a statute so as to avoid a *41possibly unconstitutional result. The plurality suggests that allowing a lower threshold would “require crossover districts throughout the Nation,” ante, at 21, thereby implicating the principle of Shaw v. Reno that districting with an excessive reliance on race is unconstitutional (“excessive” now being equated by the plurality with the frequency of creating opportunity districts). But the plurality has it precisely backwards. A State will inevitably draw some crossover districts as the natural byproduct of districting based on traditional factors. If these crossover districts count as minority-opportunity districts, the State will be much closer to meeting its §2 obligation without any reference to race, and fewer minority-opportunity districts will, therefore, need to be created purposefully. But if, as a matter of law, only majority-minority districts provide a minority seeking equality with the opportunity to eleet its preferred candidates, the State will have much further to go to create a sufficient number of minority-opportunity districts, will be required to bridge this gap by creating exclusively majority-minority districts, and will inevitably produce a districting plan that reflects a greater focus on race. The plurality, however, seems to believe that any reference to race in districting poses a constitutional concern, even a State’s decision to reduce racial blocs in favor of crossover districts. A judicial position with these consequences is not constitutional avoidance.

IV

More serious than the plurality opinion’s inconsistency with prior cases construing §2 is the perversity of the results it portends. Consider the effect of the plurality’s rule on North Carolina’s districting scheme. Black voters make up approximately 20% of North Carolina’s VAP4 and are dis*42tributed throughout 120 State House districts, App. to Pet. for Cert. 58a. As noted before, black voters constitute more than 50% of the VAP in 9 of these districts and over 39% of the VAP in an additional 12. Supra, at 33. Under a functional approach to §2, black voters in North Carolina have an opportunity to elect (and regularly do elect) the representative of their choice in as many as 21 House districts, or 17.5% of North Carolina’s total districts. See App. 109-110. North Carolina’s districting plan is therefore close to providing black voters with proportionate electoral opportunity. According to the plurality, however, the remedy of a crossover district cannot provide opportunity to minority voters who lack it, and the requisite opportunity must therefore be lacking for minority voters already living in districts where they must rely on crossover. By the plurality’s reckoning, then, black voters have an opportunity to elect representatives of their choice in, at most, nine North Carolina House districts. See ibid. In the plurality’s view, North Carolina must have a long way to go before it satisfies the § 2 requirement of equal electoral opportunity.5

*43A State like North Carolina faced with the plurality’s opinion, whether it wants to comply with § 2 or simply to avoid litigation, will, therefore, have no reason to create crossover districts. Section 2 recognizes no need for such districts, from which it follows that they can neither be required nor be created to help the State meet its obligation of equal electoral opportunity under §2. And if a legislature were induced to draw a crossover district by the plurality’s encouragement to create them voluntarily, ante, at 24, it would open itself to attack by the plurality based on the pointed suggestion that a policy favoring crossover districts runs counter to Shaw. The plurality has thus boiled §2 down to one option: the best way to avoid suit under § 2, and the only way to comply with §2, is by drawing district lines in a way that packs minority voters into majority-minority districts, probably eradicating crossover districts in the process.

Perhaps the plurality recognizes this aberrant implication, for it eventually attempts to disavow it. It asserts that “§ 2 allows States to choose their own method of complying with the Voting Rights Act, and we have said that may include drawing crossover districts. . . . [But] §2 does not mandate creating or preserving crossover districts.” Ante, at 23. See also ante, at 24 (crossover districts “can be evidence ... of equal political opportunity ... ”). But this is judicial fiat, not legal reasoning; the plurality does not even attempt to explain how a crossover district can be a minority-opportunity district when assessing the compliance of a districting plan with §2, but cannot be one when sought as a remedy to a §2 violation. The plurality cannot have it both ways. If voluntarily drawing a crossover district brings a State into compliance with §2, then requiring creation of a crossover district must be a way to remedy a violation of § 2, and eliminating a crossover district must in some cases take a State out of compliance with the statute. And when the elimination of a crossover district does cause a violation of *44§ 2, I cannot fathom why a voter in that district should not be able to bring a claim to remedy it.

In short, to the extent the plurality’s holding is taken to control future results, the plurality has eliminated the protection of § 2 for the districts that best vindicate the goals of the statute, and has done all it can to force the States to perpetuate racially concentrated districts, the quintessential manifestations of race consciousness in American politics.

I respectfully dissent.