with whom Justice Ginsburg joins, and with whom Justice Souter joins as to Parts I, II, and III, dissenting.
The Court’s decision today diminishes the value of two important rights protected by the First Amendment: the individual citizen’s right to vote for the candidate of her choice and a political party’s right to define its own mission. No one would contend that a citizen’s membership in either the Republican or the Democratic Party could disqualify her from attending political functions sponsored by another party, or from voting for a third party’s candidate in a general election. If a third party invites her to participate in its primary election, her right to support the candidate of her choice merits constitutional protection, whether she elects to make a speech, to donate funds, or to cast a ballot. The importance of vindicating that individual right far outweighs any public interest in punishing registered Republicans or Democrats for acts of disloyalty. The balance becomes even more lopsided when the individual right is *609reinforced by the right of the Libertarian Party of Óklahoma (LPO) to associate with willing voters.
In concluding that the State’s interests override those important values, the Court focuses on interests that are not legitimate. States do not have a valid interest in manipulating the outcome of elections, in protecting the major parties from competition, or in stunting the growth of new parties. While States do have a valid interest in conducting orderly elections and in encouraging the maximum participation of voters, neither of these interests overrides (or, indeed, even conflicts with) the valid interests of both the LPO and the voters who wish to participate in its primary.
In the final analysis, this case is simple. Occasionally, a political party’s interest in defining its platform and its procedures for selecting and supporting its candidates conflicts with the voters’ interest in participating in the selection of their elected representatives. If those values do conflict, we may be faced with difficult choices. But when, as in this case, those values reinforce one another a decision should be easy. Oklahoma has enacted a statute that impairs both; it denies a party the right to invite willing voters to participate in its primary elections. I would therefore affirm the Court of Appeals’ judgment.
I
In rejecting the individual respondents’ claims, the majority focuses on their associational interests. While the voters in this case certainly have an interest in associating with the LPO, they are primarily interested in voting for a particular candidate, who happens to be in the LPO. Indeed, I think we have lost sight of the principal purpose of a primary: to nominate a candidate for office. Cf. Burdick v. Takushi, 504 U. S. 428, 445 (1992) (Kennedy, J., dissenting) (“[T]he purpose of casting, counting, and recording votes is to elect public officials, not to serve as a general forum for political expression”).
*610Because our recent cases have focused on the associational interest of voters, rather than the right to vote itself, it is important to identify three basic precepts. First, it is clear that the right to vote includes the right to vote in a primary election. See United States v. Classic, 313 U. S. 299, 318 (1941); Terry v. Adams, 345 U. S. 461 (1953). When the State makes the primary an “integral part of the procedure of choice,” every eligible citizen’s right to vote should receive the same protection as in the general election. Classic, 313 U. S., at 318; see also, e. g., Gray v. Sanders, 372 U. S. 368 (1963) (invalidating primary system that diluted individual’s vote in a primary). Second, the right to vote, whether in the primary or the general election, is the right to vote “for the candidate of one’s choice.” Reynolds v. Sims, 377 U. S. 533, 555 (1964). Finally, in assessing burdens on that right — burdens that are not limited to absolute denial of the right — we should focus on the realities of the situation, not on empty formalism. See Classic, 313 U. S., at 313 (identifying “the practical operation of the primary law”); Terry, 345 U. S., at 469-470 (noting that the Jaybird primary is “the only effective part” of the election process and examining “[t]he effect of the whole procedure” in determining whether the scheme violated the Fifteenth Amendment).
Here, the impact of the Oklahoma statute on the voters’ right to vote for the candidate of their choosing is not a mere “burden”; it is a prohibition.1 By virtue of the fact that their preferred candidate is a member of a different party, respondents are absolutely precluded from voting for him or her in the primary election. It is not an answer that the *611voters could participate in another primary (1 e., the primary for the party with which they are registered) since the individual for whom they wish to vote is not a candidate in that primary. If the so-called “white primary” cases make anything clear, it is that the denial of the right to vote cannot be cured by the ability to participate in a subsequent or different election. Just as the “only election that has counted” in Terry, 345 U. S., at 469, was the Jaybird primary, since it was there that the public official was selected in any meaningful sense, the only primary that counts here is the one in which the candidate respondents want to vote for is actually running. See Burdick, 504 U. S., at 442 (Kennedy, J., dissenting) (“Because [petitioner] could not write in the name of a candidate he preferred, he had no way to cast a meaningful vote”). '
This is not to say . that voters have an absolute right to participate in whatever primary they desire. For instance, the parties themselves have a strong associational interest in determining which individuals may vote in their primaries, and that interest will normally outweigh the interest of the uninvited voter.2 But in the ordinary case the State simply has no interest in classifying voters by their political party and in limiting the elections in which voters may participate as a result of that classification. Just as we held in Reynolds that all voters of a State stand in the same relation to the State regardless of where they live, and that the State must thus not make their vote count more or less depending *612upon that factor, 377 U. S., at 565, so too do citizens stand in the same relation to the State regardless of the political party to which they belong. The State may thus not deny them participation in a primary of a party that seeks their participation absent a state interest of overriding importance.
II
In addition to burdening the individual respondent’s right to vote, the Oklahoma scheme places a heavy burden on the LPO’s associational rights. While Oklahoma permits independent voters to participate in the LPO’s primary elections, it refuses to allow registered Republicans or Democrats to do so. That refusal has a direct impact on the LPO’s selection of candidates for public office, the importance of which cannot be overstated. A primary election plays a critical role in enabling a party to disseminate its message to the public. California Democratic Party v. Jones, 530 U. S. 567, 575 (2000). It is through its candidates that a party is able to give voice to its political views, to engage other candidates on important issues of the day, and to affect change in the government of our society. Our cases “vigorously affirm the special place the First Amendment reserves for, and the special protection it accords, the process by which a political party ‘seleet[s] a standard bearer who best represents the party’s ideologies and preferences.’” Ibid, (quoting Eu v. San Francisco County Democratic Central Comm., 489 U. S. 214, 224 (1989)).
The Oklahoma statute prohibits the LPO from associating with all of the voters it believes will best enable it to select a viable candidate. The ability to select those individuals with whom to associate is, of course, at the core of the First Amendment and goes to the heart of the associational interest itself. “Freedom of association means not only that an individual voter has the right to associate with the political party of her choice, but also that a political party has a right to identify the people who constitute the association . . . .” *613Ibid, (internal quotation marks and citations omitted). See also Democratic Party of United States v. Wisconsin ex rel. La Follette, 450 U. S. 107, 122 (1981). While Libertarians can undoubtedly associate with Democrats and Republicans in other ways and at other times, the Oklahoma statute “limits the Party’s associational opportunities at the crucial juncture at which the appeal to common principles may be translated into concerted action, and hence to political power in the community.” Tashjian v. Republican Party of Conn., 479 U. S. 208, 216 (1986).
In concluding that the Oklahoma statute is constitutional, the majority argues that associational interests between the LPO and registered members of other parties are either nonexistent or not heavily burdened by the Oklahoma scheme. The plurality relies on a single footnote in Jones to show that there are no associational interests between the LPO and registered Republicans and Democrats. See ante, at 588-589 (citing 530 U. S., at 573-574, n. 5). In Jones, of course, the political parties did not want voters of other parties participating in their primaries; the putative associational inter- ■ est in this case, in which the LPO is actively courting voters of other parties, simply did not exist. More importantly, our decision in Tashjian rejected these arguments.
In Tashjian we held that the State could not prohibit Republicans from inviting voters who were not registered with a political party to participate in the Republican primary. We recognized that “[t]he Party’s attempt to broaden the base of public participation in and support for its activities is conduct undeniably central to the exercise of the right of association.” 479 U. S., at 214. Importantly, we rejected the notion that the associational interest was somehow diminished because the voters the party sought to include were not formally registered as Republicans. Id., at 215 (“[Considered from the standpoint of the Party itself, the act of formal enrollment or public affiliation with the Party is merely one element in the continuum of participation in *614Party affairs, and need not be in any sense the most important”). We reasoned that a State could not prohibit independents from contributing financial support to a Republican candidate or from participating in the party’s events; it would be anomalous if it were able to prohibit participation by independents in the “‘basic function’” of the party. Id., at 216. Because of the importance of those interests, we carefully examined the interests asserted by the State, and finding them lacking, struck down the prohibition on independents’ participation in the Republican primary.
Virtually identical interests are at stake in this case. It is the LPO’s belief that attracting a more diverse group of voters in its primary would enable it to select a more mainstream candidate who would be more viable in the general election. Like the Republicans in Tashjian, the LPO is cognizant of the fact that in order to enjoy success at the voting booth it must have support from voters who identify themselves as independents, Republicans, or Democrats.
The LPO’s desire to include Democrats and Republicans is undoubtedly informed by the fact that, given the stringent requirements of Oklahoma law, the LPO ceases to become a formally recognized party after each election cycle, and its members automatically revert to being independents.3 Because the LPO routinely loses its status as a recognized party, many voters who might otherwise register as Libertarians instead register as Democrats or Republicans.4 Thus, the LPO’s interest in inviting registered Republicans *615and Democrats to participate in the selection of its standard bearer has even greater force than did the Republican Party’s desire to invite independents to associate with it in Tashjian.
Ill
As justification for the State’s abridgment of the constitutionally protected interests asserted by the LPO and the voters, the majority relies on countervailing state interests that are either irrelevant or insignificant. Neither separately nor in the aggregate do these interests support the Court’s decision.
First, the Court makes the remarkable suggestion that by opening up its primary to Democrats and Republicans, the LPO will be saddled with so many nonlibertarian voters that the ultimate candidate will not be, in any sense, “libertarian.” See ante, at 594.5 But the LPO is seeking the crossover voting of Republicans and Democrats. Rightly or wrongly, the LPO feels that the best way to produce a viable candidate is to invite voters from other parties to participate in its primary. That may dilute what the Court believes to be the core of the Libertarian philosophy, but it is no business of the State to tell a political party what its message should be, how it should select its candidates, or how it should form coalitions to ensure electoral success. See Jones, 530 U. S., at 581-582 (rejecting state interests in producing candidates that are more centrist than the nominee the party would have selected absent the blanket primary).6
*616Second, the majority expresses concern that crossover voting may create voter confusion. This paternalistic concern is belied by the District Court’s finding that no significant voter confusion would occur. App. to Pet. for Cert. A-43 (noting that “very simple rules for voting eligibility can be posted at polling places when the primary and runoff elections are conducted”).
Third, the majority suggests that crossover voting will impair the State’s interest in properly classifying candidates and voters. As an empirical matter, a crossover voter may have a lesser commitment to the party with which he is registered if he votes in another party’s primary. Nevertheless, the State does not have a valid interest in defining what it means to be a Republican or a Democrat, or in attempting to ensure the political orthodoxy of party members simply for the convenience of those parties. Cf. West Virginia Bd. of Ed. v. Barnette, 319 U. S. 624, 642 (1943) (“If there is any fixed star in our constitutional constellation, it is that no official, high or petty, can prescribe what shall be orthodox in politics, nationalism, religion, or other matters of opinion ... ”). Even if participation in the LPO’s primary causes a voter to be a less committed “Democrat” or “Republican” (a proposition I reject7), the dilution of that commitment does not justify abridgment of the fundamental rights at issue in this case. While party identity is important in *617our political system, it should not be immunized from the risk of change.8
Fourth, the majority argues that opening up the LPO primary to members of the Republican and Democratic Parties might interfere with electioneering and party-building efforts. It is clear, of course, that the majority here is concerned only with the Democratic and Republican Parties, since party building is precisely what the LPO is attempting to accomplish. Nevertheless, that concern is misplaced. Even if, as the majority claims, the Republican and Democratic voter rolls, mailing lists, and phone banks are not as accurate as they would otherwise be,9 the administrative inconvenience of the major parties does not outweigh the right to vote or the associational interests of those voters and the LPO. At its core, this argument is based on a fear that the LPO might be successful in convincing Democratic or Republican voters to participate more fully in the LPO. Far from being a compelling interest, it is an impermissible one. Timmons v. Twin Cities Area New Party, 520 U. S. 351, 367 *618(1997) (State may not “completely insulate the two-party system from minor parties’ or independent candidates’ competition and influence”).
Finally, the majority warns against the possibility of raiding, ante, at 596, by which voters of another party maliciously vote in a primary in order to change the outcome of the primary, either to nominate a particularly weak candidate, a “sore-loser” candidate, or a candidate who would siphon votes from another party. The District Court, whose factual findings are entitled to substantial deference, found as a factual and legal matter that the State’s argument concerning raiding was “unpersuasive.” App. to Pet. for Cert. A-61.
Even if raiding were a possibility, however, the state interests are remote. The possibility of harm to the LPO itself is insufficient to overcome the LPO’s associational rights. See Eu, 489 U. S., at 227-228 (“[E]ven if a ban on endorsements saves a political party from pursuing self-destructive acts, that would not justify a State substituting its judgment for that of the party”). If the LPO is willing to take the risk that its party may be “hijacked” by individuals who hold views opposite to their own, the State has little interest in second-guessing the LPO’s decision.
With respect to the possibility that Democratic or Republican voters might raid the LPO to the detriment of their own or another party, neither the State nor the majority has identified any evidence that voters are sufficiently organized to achieve such a targeted result.10 Such speculation is not, in *619my view, sufficient to override the real and acknowledged interest of the LPO and the voters who wish to participate in its primary. See Timmons, 520 U. S., at 375 (Stevens, J., dissenting) (citing Eu, 489 U. S., at 226; Anderson v. Celebrezze, 460 U. S. 780, 789 (1983); and Norman v. Reed, 502 U. S. 279, 288-289 (1992)).11
In the end, the balance of interests clearly favors the LPO and those voters who wish to participate in its primary. The associational interests asserted — the right to select a standard bearer that the party thinks has the best chance of success, the ability to associate at the crucial juncture of selecting a candidate, and the desire to reach out to voters of other parties — are substantial and undoubtedly burdened by Oklahoma’s statutory scheme. Any doubt about that fact is clearly answered by Tashjian. On the other side, the interests asserted by the State are either entirely speculative or simply protectionist measures that benefit the parties in power. No matter what the standard, they simply do not outweigh the interests of the LPO and its voters.
IV
The Libertarian Party of Oklahoma is not the only loser in this litigation. Other minor parties and voters who have primary allegiance to one party but sometimes switch their support to rival candidates are also harmed by this decision. In my judgment, however, the real losers include all participants in the political market. Decisions that give undue def*620erence to the interest in preserving the two-party system,12 like decisions that encourage partisan gerrymandering,13 enhance the likelihood that so-called “safe districts” will play an increasingly predominant role in the electoral process. Primary elections are already replacing general elections as the most common method of actually determining the composition of our legislative bodies. The trend can only increase the bitter partisanship that has already poisoned some of those bodies that once provided inspiring examples of courteous adversary debate and deliberation.
The decision in this case, like the misguided decisions in Timmons, 520 U. S. 351, and Jones, 530 U. S. 567, attaches overriding importance to the interest in preserving the two-party system. In my view, there is over a century of experience demonstrating that the two major parties are fully capable of maintaining their own positions of dominance in the political marketplace without any special assistance from the state governments that they dominate or from this Court. Whenever they receive special advantages, the offsetting harm to independent voters may be far more significant than the majority recognizes.
In Anderson, 460 U. S. 780, we considered the impact of' early filing dates on small political parties and independent candidates. Commenting on election laws that disadvantage independents, we noted:
“By limiting the opportunities of independent-minded voters to associate in the electoral arena to enhance their political effectiveness as a group, such restrictions threaten to reduce diversity and competition in the marketplace of ideas. Historically political figures outside *621the two major parties have been fertile sources of new ideas and new programs; many of their challenges to the status quo have in time made their way into the political mainstream. In short, the primary values protected by the First Amendment — ‘a profound national commitment to the principle that debate on public issues should be uninhibited, robust, and wide-open,’ — are served when election campaigns are not monopolized by the existing political parties.” Id., at 794 (citations omitted).
Because the Court’s holding today has little to support it other than a naked interest in protecting the two major parties, I respectfully dissent.
It is not enough that registered members of other parties may simply change their registration. See ante, at 590-591 (plurality opinion). Changing one’s political party is not simply a matter of filing a form with the State; for many individuals it can be a significant decision. A view that' party membership is merely a label demeans for many the personal significance of party identification and illustrates what little weight the majority actually gives to the associational interests in this case.
The voters’ interest may still prevail if, as was the case in Terry v. Adams, 345 U. S. 461 (1953), and Smith v. Allwright, 321 U. S. 649 (1944), the party primary is the defacto election. In part because of this Court’s refusal to intervene in political gerrymandering eases, Davis v. Bandemer, 478 U. S. 109 (1986), an increasing number of districts are becoming “safe districts” in which one party effectively controls the outcome of the election. See, e. g., Courtney, Redistricting: What the United States Can Learn from Canada, 3 Election L. J. 488 (2004) (concluding that 400 of the 435 Members of the House of Representatives were elected in safe districts in the 2002 election, 81 of whom ran unopposed).
See Okla. Stat. Ann., Tit. 26, § 1-109 (West Supp. 2005) (requiring that a party’s nominee for Governor, President, or Vice President receive 10% of the vote in a general election for the party to maintain its status).
See App. to Pet. for Cert. A-48 (District Court recognizing that “it is highly likely that the ranks of independents and, indeed, of registered Republicans and Democrats, contain numerous voters who sympathize with the LPO but who simply do not wish to go through the motions of re-registering every time they are purged from the rolls”).
Of course, as the majority recognizes, ante, at 594, since the number of independent voters overwhelms the number of registered-LPO voters, that is already the ease.
See also Democratic Party of United States v. Wisconsin ex rel. La Follette, 450 U. S. 107, 123-124 (1981) (State may not substitute its own judgment for that of the party); Jones, 530 U. S., at 587 (Kennedy, J., concurring) (“A political party might be better served by allowing blanket primaries as a means of nominating candidates with broader appeal. Under the First Amendment’s guarantee of speech through free associa*616tion, however, this is an issue for the party to resolve, not for the State” (emphasis added)). Such coalition building, and reaching out to other groups to ensure a candidate gets elected, is a vital part of the political process. Cf. Colorado Republican Federal Campaign Comm. v. Federal Election Comm’n, 518 U. S. 604, 622-623 (1996) (citing W. Keefe, Parties, Politics, and Public Policy in America 59-74 (5th ed. 1988)).
Allowing a potential crossover voter to vote in the LPO primary would not change the level of commitment he has toward his party of registration; it would simply give him an outlet to express the views he already holds.
If, of course, States were able to protect the incumbent parties in the name of protecting the stability of the two-party system in general, we might still have the Federalists, the Anti-federalists, or the Whigs. See generally J. Aldrich, Why Parties? The Origin and Transformation of Political Parties in America (1995). In any event, we would not have the evolution of thought or policies that are occasioned through the change of political parties. While no such change has occurred in recent memory, that is no reason to ossify the status quo.
The majority’s argument is that voters who would otherwise vote in the Republican or Democratic primaries would vote in the LPO primary, and that the Democratic and Republican lists would not be an accurate indicator of who is likely to vote in those primaries, and of which voters to spend party resources on. First, I find it doubtful that those voters who vote in the LPO primary would have voted in the Democratic or Republican primary; rather, they probably would not have been sufficiently motivated to vote at all. Further, this would actually give Republicans and Democrats additional information as to which of their voters have Libertarian leanings.
To change the outcome of an election in a way that would benefit their own party, voters would have to be relatively certain that their preferred candidate in their own primary would win that primary and to vote in the LPO primary for a previously agreed-on candidate who is opposed to their own ideological preferences. Given that voters typically do not focus on an election until several days or weeks before an election, this prospect is unlikely. See California Democratic Party v. Jones, 530 U. S. 567, 586 (2000) (Kennedy, J., concurring). Further, one would have expected to see some evidence of this in States where it is relatively easy to switch parties close to a primary.
The flimsy character of the state interests in this case confirms my view that today’s decision rests primarily on a desire to protect the two-party system. In Jones, the Court concluded that the associational interests of the parties trumped state interests that were much more compelling than those asserted in this case. Here, by contrast, where the associational interests are being asserted by a minor party rather than by one of the dominant parties, the Court has reversed course and rejected those associational interests as insubstantial compared to the interests asserted by the State.
Examples are cases permitting lengthy registration periods, Rosario v. Rockefeller, 410 U. S. 752 (1973), and cases approving bans on fusion candidates, Timmons v. Twin Cities Area New Party, 520 U. S. 351 (1997), and write-in candidates, Burdick v. Takushi, 504 U. S. 428 (1992).
See, e. g., Vieth v. Jubelirer, 541 U. S. 267 (2004); Davis v. Bandemer, 478 U. S. 109 (1986).