delivered the opinion of the Court.

Most States prohibit multiple-party, or “fusion,” candidacies for elected office.1 The Minnesota laws challenged in *354this case prohibit a candidate from appearing on the ballot as the candidate of more than one party. Minn. Stat. §§204B.06, subd. 1(b), and 204B.04, subd. 2 (1994). We hold that such a prohibition does not violate the First and Fourteenth Amendments to the United States Constitution.

Respondent is a chartered chapter of the national New Party. Petitioners are Minnesota election officials. In April 1994, Minnesota State Representative Andy Dawkins was running unopposed in the Minnesota Democratic-Farmer-Labor Party’s (DFL) primary.2 That same month, New Party members chose Dawkins as their candidate for the same office in the November 1994 general election. Neither Dawkins nor the DFL objected, and Dawkins signed the required affidavit of candidacy for the New Party. Minn. Stat. § 204B.06 (1994). Minnesota, however, prohibits fusion candidacies.3 Because Dawkins had already filed as a candidate for the DFL’s nomination, local election officials refused to accept the New Party’s nominating petition.4

*355The New Party filed suit in United States District Court, contending that Minnesota’s antifusion laws violated the party’s associational rights under the First -and Fourteenth Amendments. The District Court granted summary judgment for the state defendants, concluding that Minnesota’s fusion ban was “a valid and non-discriminatory regulation of the election process,” and noting that “issues concerning the mechanics of choosing candidates . . . are, in large part, matters of policy best left to the deliberative bodies themselves.” Twin Cities Area New Party v. McKenna, 863 F. Supp. 988, 994 (D. Minn. 1994).

The Court of Appeals reversed. Twin Cities Area New Party v. McKenna, 73 F. 3d 196, 198 (CA8 1996). First, the court determined that Minnesota’s fusion ban “unquestionably” and “severe[ly]” burdened the New Party’s “freedom to select a standard bearer who best represents the party’s ideologies and preferences” and its right to “broaden the base of public participation in and support for [its] activities.” Ibid, (citations and internal quotation marks omitted). The court then decided that Minnesota’s absolute ban on multiple-party nominations was “broader than necessary to serve the State’s asserted interests” in avoiding intra-party discord and party splintering, maintaining a stable political system, and avoiding voter confusion, and that the State’s remaining concerns about multiple-party nomination were “simply unjustified in this case.” Id., at 199-200. The court noted, however, that the Court of Appeals for the Seventh Circuit had upheld Wisconsin’s similar fusion ban in Swamp v. Kennedy, 960 F. 2d 383, 386 (1991) (fusion ban did not burden associational rights and, even if it did, the State’s interests justified the burden), cert. denied, 505 U. S. 1204 (1992). Nonetheless, the court concluded that Minnesota’s fusion-ban provisions, Minn. Stat. §§204B.06, subd. 1(b), and *356204B.04, subd. 2 (1994), were unconstitutional because they severely burdened the New Party’s associational rights and were not narrowly tailored to advance Minnesota’s valid interests. We granted certiorari, 517 U. S. 1219 (1996), and now reverse.

Fusion was a regular feature of Gilded Age American politics. Particularly in the West and Midwest, candidates of issue-oriented parties like the Grangers, Independents, Greenbackers, and Populists often succeeded through fusion with the Democrats, and vice versa. Republicans, for their part, sometimes arranged fusion candidacies in the South, as part of a general strategy of encouraging and exploiting divisions within the dominant Democratic Party. See generally Argersinger, “A Place on the Ballot”: Fusion Politics and Antifusion Laws, 85 Am. Hist. Rev. 287, 288-290 (1980).

Fusion was common in part because political parties, rather than local or state governments, printed and distributed their own ballots. These ballots contained only the names of a particular party’s candidates, and so a voter could drop his party’s ticket in the ballot box without even knowing that his party’s candidates were supported by other parties as well. But after the 1888 presidential election, which was widely regarded as having been plagued by fraud, many States moved to the “Australian ballot system.” Under that system, an official ballot, containing the names of all the candidates legally nominated by all the parties, was printed at public expense and distributed by public officials at polling places. Id., at 290-292; Burdick v. Takushi, 504 U. S. 428, 446-447 (1992) (Kennedy, J., dissenting) (States’ move to the Australian ballot system was a “progressive reform to reduce fraudulent election practices”). By 1896, use of the Australian ballot was widespread. During the same period, many States enacted other election-related reforms, including bans on fusion candidacies. See Argersinger, supra, at *357288, 295-298. Minnesota banned fusion in 1901.5 This trend has continued and, in this century, fusion has become the exception, not the rule. Today, multiple-party candidacies are permitted in just a few States,6 and fusion plays a significant role only in New York.7

The First Amendment protects the right of citizens to associate and to form political parties for the advancement of common political goals and ideas. Colorado Republican Federal Campaign Comm. v. Federal Election Comm’n, 518 U. S. 604, 616 (1996) (“The independent expression of a political party’s views is ‘core’ First Amendment activity no less than is the independent expression of individuals, candidates, or other political committees”); Norman v. Reed, 502 U. S. 279, 288 (1992) (“constitutional right of citizens to create and develop new political parties . . . advances the constitutional interest of like-minded voters to gather in pursuit of common political ends”); Tashjian v. Republican Party of Conn., 479 *358U. S. 208, 214 (1986). As a result, political parties’ government, structure, and activities enjoy constitutional protection. Eu v. San Francisco County Democratic Central Comm., 489 U. S. 214, 230 (1989) (noting political party’s “discretion in how to organize itself, conduct its affairs, and select its leaders”); Tashjian, supra, at 224 (Constitution protects a party’s “determination ... of the structure which best allows it to pursue its political goals”).

On the other hand, it is also clear that States may, and inevitably must, enact reasonable regulations of parties, elections, and ballots to reduce election- and campaign-related disorder. Burdick, supra, at 433 (“ ‘[A]s a practical matter, there must be a substantial regulation of elections if they are to be fair and honest and if some sort of order, rather than chaos, is to accompany the democratic process’ ”) (quoting Storer v. Brown, 415 U. S. 724, 730 (1974)); Tashjian, supra, at 217 (The Constitution grants States “broad power to prescribe the ‘Time, Places and Manner of holding Elections for Senators and Representatives,’ Art. I, §4, cl. 1,-which power is matched by state control over the election process for state offices”).

When deciding whether a state election law violates First and Fourteenth Amendment associational rights, we weigh the “‘character and magnitude’” of the burden the State’s rule imposes on those rights against the interests the State contends justify that burden, and consider the extent to which the State’s concerns make the burden necessary. Burdick, supra, at 434 (quoting Anderson v. Celebrezze, 460 U. S. 780, 789 (1983)). Regulations imposing severe burdens on plaintiffs' rights must be narrowly tailored and advance a compelling state interest. Lesser burdens, however, trigger less exacting review, and a State’s “‘important regulatory interests’” will usually be enough to justify “‘reasonable, nondiscriminatory restrictions.’” Burdick, supra, at 434 (quoting Anderson, supra, at 788); Norman, supra, at 288-289 (requiring “corresponding interest sufficiently weighty *359to justify the limitation”). No bright line separates permissible election-related regulation from unconstitutional infringements on First Amendment freedoms. Storer, supra, at 730 (“[N]o litmus-paper test. . . separates] those restrictions that are valid from those that are invidious .... The rule is not self-executing and is no substitute for the hard judgments that must be made”).

The New Party’s claim that it has a right to select its own candidate is uncontroversial, so far as it goes. See, e. g., Cousins v. Wigoda, 419 U. S. 477 (1975) (party, not State, has right to decide who will be State’s delegates at party convention). That is, the New Party, and not someone else, has the right to select the New Party’s “standard bearer.” It does not follow, though, that a party is absolutely entitled to have its nominee appear on the ballot as that party’s candidate. A particular candidate might be ineligible for office,8 unwilling to serve, or, as here, another party’s candidate. That a particular individual may not appear on the ballot as a particular party’s candidate does not severely burden that party’s associational rights. See Burdick, 504 U. S., at 440, n. 10 (“It seems to us that limiting the choice of candidates to those who have complied with state election law requirements is the prototypical example of a regulation that, while it affects the right to vote, is eminently reasonable”); Anderson, 460 U. S., at 792, n. 12 (“Although a disaffiliation provision may preclude . . . voters from supporting a particular ineligible candidate, they remain free to support and promote other candidates who satisfy the State’s disaffiliation requirements”); id., at 793, n. 15.

The New Party relies on Eu v. San Francisco County Democratic Central Comm., supra, and Tashjian v. Republican Party of Conn., supra. In Eu, we struck down Califor*360nia election provisions that prohibited political parties from endorsing candidates in party primaries and regulated parties’ internal affairs and structure. And in Tashjian, we held that Connecticut’s closed-primary statute, which required voters in a party primary to be registered party members, interfered with a party’s associational rights by limiting “the group of registered voters whom the Party may invite to participate in the basic function of selecting the Party’s candidates.” 479 U. S., at 215-216 (internal quotation marks and citations omitted). But while Tashjian and Eu involved regulation of political parties’ internal affairs and core associational activities, Minnesota’s fusion ban does not. The ban, which applies to major and minor parties alike, simply precludes one party’s candidate from appearing on the ballot, as that party’s candidate, if already nominated by another party. Respondent is free to try to convince Representative Dawkins to be the New Party’s, not the DFL’s, candidate. See Swamp, 950 F. 2d, at 385 (“[A] party may nominate any candidate that the party can convince to be its candidate”). Whether the party still wants to endorse a candidate who, because of the fusion ban, will not appear on the ballot as the party’s candidate, is up to the party.

The Court of Appeals also held that Minnesota’s laws “keep the New Party from developing consensual political alliances and thus broadening the base of public participation in and support for its activities.” McKenna, 73 F. 3d, at 199. The burden on the party was, the court held, severe because “[h]istory shows that minor parties have played a significant role in the electoral system where multiple party nomination is legal, but have no meaningful influence where multiple party nomination is banned.” Ibid. In the view of the Court of Appeals, Minnesota’s fusion ban forces members of the New Party to make a “no-win choice” between voting for “candidates with no realistic chance of winning, defecting] from their party and vot[ing] for a major party candidate who does, or deelin[ing] to vote at all.” Ibid.

*361But Minnesota has not directly precluded minor political parties from developing and organizing. Cf. Norman, 502 U. S., at 289 (statute “foreclose^] the development of any political party lacking the resources to run a statewide campaign”). Nor has Minnesota excluded a particular group of citizens, or a political party, from participation in the election process. Cf. Anderson, supra, at 792-793 (filing deadline “places a particular burden on an identifiable segment of Ohio’s independent-minded voters”); Bullock v. Carter, 405 U. S. 134 (1972) (striking down Texas statute requiring candidates to pay filing fees as a condition to having their names placed on primary-election ballots). The New Party remains free to endorse whom it likes, to ally itself with others, to nominate candidates for office, and to spread its message to all who will listen. Cf. Eu, 489 U. S., at 223 (California law curtailed right to “[f]ree discussion about candidates for public office”); Colorado Republican Federal Campaign Comm’n, 518 U. S., at 615 (restrictions on party’s spending impair its ability to “engage in direct political advocacy”).

The Court of Appeals emphasized its belief that, without fusion-based alliances, minor parties cannot thrive. This is a predictive judgment which is by no means self-evident.9 *362But, more importantly, the supposed benefits of fusion to minor parties do not require that Minnesota permit it. See Tashjian, supra, at 222 (refusing to weigh merits of closed and open primaries). Many features of our political system — e. g., single-member districts, “first past the post” elections, and the high costs of campaigning — make it difficult for third parties to succeed in American politics. Burnham Declaration, App. 12-13. But the Constitution does not require States to permit fusion any more than it requires them to move to proportional-representation elections or public financing of campaigns. See Mobile v. Bolden, 446 U. S. 55, 75 (1980) (plurality opinion) (“Whatever appeal the dissenting opinion’s view may have as a matter of political theory, it is not the law”).

The New Party contends that the fusion ban burdens its “right... to communicate its choice of nominees on the ballot on terms equal to those offered other parties, and the right of the party’s supporters and other voters to receive that information,” and insists that communication on the ballot of a party’s candidate choice is a “critical source of information for the great majority of voters . . . who . .. rely upon party ‘labels’ as a voting guide.” Brief for Respondent 22-23.

It is true that Minnesota’s fusion ban prevents the New Party from using the ballot to communicate to the public that it supports a particular candidate who is already another party’s candidate. In addition, the ban shuts off one possible avenue a party might use to send a message to its preferred candidate because, with fusion, a candidate who wins an election on the basis of two parties’ votes will likely know more — if the parties’ votes are counted separately — about the particular wishes and ideals of his constituency. We are *363unpersuaded, however, by the party’s contention that it has a right to use the ballot itself to send a particularized message, to its candidate and to the voters, about the nature of its support for the candidate. Ballots serve primarily to elect candidates, not as forums for political expression. See Burdick, 504 U. S., at 438; id., at 445 (Kennedy, J., dissenting). Like all parties in Minnesota, the New Party is able to use the ballot to communicate information about itself and its candidate to the voters, so long as that candidate is not already someone else’s candidate. The party retains great latitude in its ability to communicate ideas to voters and candidates through its participation in the campaign, and party members may campaign for, endorse, and vote for their preferred candidate' even if he is listed on the ballot as another party’s candidate. See Anderson, 460 U. S., at 788 (“[A]n election campaign is an effective platform for the expression of views on the issues of the day”); Illinois Bd. of Elections v. Socialist Workers Party, 440 U. S. 173, 186 (1979) (“[A]n election campaign is a means of disseminating ideas”).

In sum, Minnesota’s laws do not restrict the ability of the New Party and its members to endorse, support, or vote for anyone they like. The laws do not directly limit the party’s access to the ballot. They are silent on parties’ internal structure, governance, and policymaking. Instead, these provisions reduce the universe of potential candidates who may appear on the ballot as the party’s nominee only by ruling out those few individuals who both have already agreed to be another party’s candidate and also, if forced to choose, themselves prefer that other party. They also limit, slightly, the party’s ability to send a message to the voters and to its preferred candidates. We conclude that the burdens Minnesota imposes on the party’s First and Fourteenth Amendment associational rights — though not trivial — are not severe.

The Court of Appeals determined that Minnesota’s fusion ban imposed “severe” burdens on the New Party’s associa*364tional rights, and so it required the State to show that the ban was narrowly tailored to serve compelling state interests. McKenna, 73 F. 3d, at 198. We disagree; given the burdens imposed, the bar is not so high. Instead, the State’s asserted regulatory interests need only be “sufficiently weighty to justify the limitation” imposed on the party’s rights. Norman, 502 U. S., at 288-289; Burdick, supra, at 434 (quoting Anderson, supra, at 788). Nor do we require elaborate, empirical verification of the weightiness of the State’s asserted justifications. See Munro v. Socialist Workers Party, 479 U. S. 189, 195-196 (1986) (“Legislatures . . . should be permitted to respond to potential deficiencies in the electoral process with foresight rather than reactively, provided that the response is reasonable and does not significantly impinge on constitutionally protected rights”).

The Court of Appeals acknowledged Minnesota’s interests in avoiding voter confusion and overcrowded ballots, preventing party splintering and disruptions of the two-party system, and being able to clearly identify the election winner. McKenna, supra, at 199-200. Similarly, the Seventh Circuit, in Swamp, noted Wisconsin’s “compelling” interests in avoiding voter confusion, preserving the integrity of the election process, and maintaining a stable political system. 950 F. 2d, at 386; cf. id., at 387-388 (Fairchild, J., concurring) (State has a compelling interest in “maintaining the distinct identity of parties”). Minnesota argues here that its fusion ban is justified by its interests in avoiding voter confusion, promoting candidate competition (by reserving limited ballot space for opposing candidates), preventing electoral distortions and ballot manipulations, and discouraging party splintering and “unrestrained factionalism.” Brief for Petitioners 41-50.

States certainly have an interest in protecting the integrity, fairness, and efficiency of their ballots and election processes as means for electing public officials. Bullock, 405 U. S., at 145 (State may prevent “frivolous or fraudulent *365candidacies”) (citing Jenness v. Fortson, 403 U. S. 431, 442 (1971)); Eu, 489 U. S., at 231; Norman, supra, at 290 (States have an interest in preventing “misrepresentation”); Rosario v. Rockefeller, 410 U. S. 752, 761 (1973). Petitioners contend that a candidate or party could easily exploit fusion as a way of associating his or its name with popular slogans and catchphrases. For example, members of a major party could decide that a powerful way of “sending a message” via the ballot would be for various factions of that party to nominate the major party’s candidate as the candidate for the newly formed “No New Taxes,” “Conserve Our Environment,” and “Stop Crime Now” parties. In response, an opposing major party would likely instruct its factions to nominate that party’s candidate as the “Fiscal Responsibility,” “Healthy Planet,” and “Safe Streets” parties’ candidate.

Whether or not the putative “fusion” candidates’ names appeared on one or four ballot lines, such maneuvering would undermine the ballot’s purpose by transforming it from a means of choosing candidates to a billboard for political advertising. The New Party responds to this concern, ironically enough, by insisting that the State could avoid such manipulation by adopting more demanding ballot-access standards rather than prohibiting multiple-party nomination. Brief for Respondent 38. However, as we stated above, because the burdens the fusion ban imposes on the party’s associational rights are not severe, the State need not narrowly tailor the means it chooses to promote ballot integrity. The Constitution does not require that Minnesota compromise the policy choices embodied in its ballot-access requirements to accommodate the New Party’s fusion strategy. See Minn. Stat. §204B.08, subd. 3 (1994) (signature requirements for nominating petitions); Rosario, supra, at 761-762 (New York’s time limitation for enrollment in a political party was part of an overall scheme aimed at the preservation of the integrity of the State’s electoral process).

*366Relatedly, petitioners urge that permitting fusion would undercut Minnesota’s ballot-access regime by allowing minor parties to capitalize on the popularity of another party’s candidate, rather than on their own appeal to the voters, in order to secure access to the ballot. Brief for Petitioners 45-46. That is, voters who might not sign a minor party’s nominating petition based on the party’s own views and candidates might do so if they viewed the minor party as just another way of nominating the same person nominated by one of the major parties. Thus, Minnesota fears that fusion would enable minor parties, by nominating a major party’s candidate, to bootstrap their way to major-party status in the next election and circumvent the State’s nominating-petition requirement for minor parties. See Minn. Stat. §§ 200.02, subd. 7 (defining “major party”), and 204D.13 (1994) (describing ballot order for major and other parties). The State surely has a valid interest in making sure that minor and third parties who are granted access to the ballot are bona fide and actually supported, on their own merits, by those who have provided the statutorily required petition or ballot support. Anderson, 460 U. S., at 788, n. 9; Storer, 415 U. S., at 733, 746.

States also have a strong interest in the stability of their political systems.10 Eu, supra, at 226; Storer, supra, at 736. *367This interest does not permit a State to completely insulate the two-party system from minor parties’ or independent candidates’ competition and influence, Anderson, supra, at 802; Williams v. Rhodes, 393 U. S. 23 (1968), nor is it a paternalistic license for States to protect political parties from the consequences of their own internal disagreements. Eu, supra, at 227; Tashjian, 479 U. S., at 224. That said, the States’ interest permits them to enact reasonable election regulations that may, in practice, favor the traditional two-party system, see Burnham Declaration, App. 12 (American politics has been, for the most part, organized around two parties since the time of Andrew Jackson), and that temper the destabilizing effects of party splintering and excessive factionalism. The Constitution permits the Minnesota Legislature to decide that political stability is best served through a healthy two-party system. See Rutan v. Republican Party of Ill., 497 U. S. 62, 107 (1990) (Scalia, J., dissenting) (“The stabilizing effects of such a [two-party] system are obvious”); Davis v. Bandemer, 478 U. S. 109, 144-145 (1986) (O’Connor, J., concurring) (“There can be little doubt that the emergence of a strong and stable two-party system in this country has contributed enormously to sound and effective government”); Branti v. Finkel, 445 U. S. 507, 532 (1980) (Powell, J., dissenting) (“Broad-based political parties supply an essential coherence and flexibility to the American political scene”). And while an interest in securing the perceived benefits of a stable two-party system will not justify unreasonably exclusionary restrictions, see Williams, supra, at 31-32, States need not remove all of the many hurdles third parties face in the American political arena today.

In Storer, we upheld a California statute that denied ballot positions to independent candidates who had voted in the immediately preceding primary elections or had a registered party affiliation at any time during the year before the same *368primary elections. 415 U. S., at 728.11 After surveying the relevant case law, we “ha[d] no hesitation in sustaining” the party-disaffiliation provisions. Id., at 733. We recognized that the provisions were part of a “general state policy aimed at maintaining the integrity of . . . the ballot,” and noted that the provision did not discriminate against independent candidates. Ibid. We concluded that while a “State need not take the course California has, . . . California apparently believes with the Founding Fathers that splintered parties and unrestrained factionalism may do significant damage to the fabric of government. See The Federalist No. 10 (Madison). It appears obvious to us that the one-year disaffiliation provision furthers the State’s interest in the stability of its political system.” 415 U. S., at 736; see also Lippitt v. Cipollone, 404 U. S. 1032 (1972) (affirming, without opinion, district-court decision upholding statute banning party-primary candidacies of those who had voted in another party’s primary within last four years).12

*369Our decision in Burdick v. Takushi, supra, is also relevant. There, we upheld Hawaii’s ban on write-in voting against a claim that the ban unreasonably infringed on citizens’ First and Fourteenth Amendment rights. In so holding, we rejected the petitioner’s argument that the ban “deprive[d] him of the opportunity to cast a meaningful ballot,” emphasizing that the function of elections is to elect candidates and that “we have repeatedly upheld reasonable, politically neutral regulations that have the effect of channeling expressive activities] at the polls.” 504 U. S., at 437-438.

Minnesota’s fusion ban is far less burdensome than the disaffiliation rule upheld in Storer, and is justified by similarly weighty state interests. By reading Storer as dealing only with “sore-loser candidates,” Justice Stevens, in our view, fails to appreciate the case’s teaching. Post, at 377 (dissenting opinion). Under the California disaffiliation statute at issue in Storer, any person affiliated with a party at any time during the year leading up to the primary election was absolutely precluded from appearing on the ballot as an independent or as the candidate of another party. Minnesota’s fusion ban is not nearly so restrictive; the challenged provisions say nothing about the previous party affiliation of would-be candidates but only require that, in order to appear on the ballot, a candidate not be the nominee of more than one party. California’s disaffiliation rule limited the field of candidates by thousands; Minnesota’s precludes only a handful who freely choose to be so limited. It is also worth noting that while California’s disaffiliation statute absolutely banned many candidacies, Minnesota’s fusion ban only prohibits a candidate from being named twice.

We conclude that the burdens Minnesota’s fusion ban imposes on the New Party’s associational rights are justified by “correspondingly weighty” valid state interests in ballot *370integrity and political stability.13 In deciding that Minnesota’s fusion ban does not unconstitutionally burden the New Party’s First and Fourteenth Amendment rights, we express no views on the New Party’s policy-based arguments concerning the wisdom of fusion. It may well be that, as support for new political parties increases, these arguments will carry the day in some States’ legislatures. But the Constitution does not require Minnesota, and the approximately 40 other States that do not permit fusion, to allow it. The judgment of the Court of Appeals is reversed.

It is so ordered.

dissenting.

In Minnesota, the Twin Cities Area New Party (Party) is a recognized minor political party entitled by state law to have the names of its candidates for public office appear on the state ballots. In April 1994, Andy Dawkins was qualified to be a candidate for election to the Minnesota Legislature as the representative of House District 65A. With Dawkins’ consent, the Party nominated him as its candidate for that office. In my opinion the Party and its members had a constitutional right to have their candidate’s name appear on the ballot despite the fact that he was also the nominee of another party.

The Court’s conclusion that the Minnesota statute prohibiting multiple-party candidacies is constitutional rests on three dubious premises: (1) that the statute imposes only a *371minor burden on the Party’s right to choose and to support the candidate of its choice; (2) that the statute significantly serves the State’s asserted interests in avoiding ballot manipulation and factionalism; and (3) that, in any event, the interest in preserving the two-party system justifies the imposition of the burden at issue in this case. I disagree with each of these premises.

I

The members of a recognized political party unquestionably have a constitutional right to select their nominees for public office and to communicate the identity of their nominees to the voting public. Both the right to choose and the right to advise voters of that choice are entitled to the highest respect.

The Minnesota statutes place a significant burden on both of those rights. The Court’s recital of burdens that the statute does not inflict on the Party, ante, at 363, does nothing to minimize the severity of the burdens that it does impose. The fact that the Party may nominate its second choice surely does not diminish the significance of a restriction that denies it the right to have the name of its first choice appear on the ballot. Nor does the point that it may use some of its limited resources to publicize the fact that its first choice is the nominee of some other party provide an adequate substitute for the message that is conveyed to every person who actually votes when a party’s nominees appear on the ballot.

As to the first point, the State contends that the fusion ban in fact limits by only a few candidates the range of individuals a party may nominate, and that the burden is therefore quite small. But the number of candidates removed from the Party’s reach cannot be the determinative factor. The ban leaves the Party free to nominate any eligible candidate except the particular “ ‘standard bearer who best represents the party’s ideologies and preferences.’” Eu v. San Francisco County Democratic Central Comm., 489 U. S. 214, 224 (1989).

*372The Party could perhaps choose to expend its resources supporting a candidate who was not in fact the best representative of its members’ views. But a party’s choice of a candidate is the most effective way in which that party can communicate to the voters what the party represents and, thereby, attract voter interest and support.1 Political parties “exist to advance their members’ shared political beliefs,” and “in the context of particular elections, candidates are necessary to make the party’s message known and effective, and vice versa.” Colorado Republican Federal Campaign Comm. v. Federal Election Comm’n, 518 U. S. 604, 629 (1996) (Kennedy, J., dissenting). See also Anderson v. Celebrezze, 460 U. S. 780, 821 (1983) (Rehnquist, J., dissenting) (“Political parties have, or at least hope to have, a continuing existence, representing particular philosophies. Each party has an interest in finding the best candidate to advance its philosophy in each election”).

The State next argues that — instead of nominating a second-choice candidate — the Party could remove itself from *373the ballot altogether, and publicly endorse the candidate of another party. But the right to be on the election ballot is precisely what separates a political party from any other interest group.2 The Court relies on the fact that the Party remains free “to spread its message to all who will listen,” ante, at 361, through forums other than the ballot. Given the limited resources available to most minor parties, and the less-than-universal interest in the messages of third parties, it is apparent that the Party’s message will, in this manner, reach a much smaller audience than that composed of all voters who can read the ballot in the polling booth.

The majority rejects as unimportant the limits that the fusion ban may impose on the Party’s ability to express its political views, ante, at 362-363, relying on our decision in Burdick v. Takushi, 504 U. S. 428, 445 (1992), in which we noted that “the purpose of casting, counting, and recording votes is to elect public officials, not to serve as a general forum for political expression.” But in Burdick we concluded simply that an individual voter’s interest in expressing his disapproval of the single candidate running for office in a particular election did not require the State to finance and provide a mechanism for tabulating write-in votes. Our conclusion that the ballot is not principally a forum for the individual expression of political sentiment through the casting of a vote does not justify the conclusion that the ballot serves no expressive purpose for the parties who place candidates on the ballot. Indeed, the long-recognized right to choose a “ ‘standard bearer who best represents the party’s ideologies and preferences,’ ” Eu, 489 U. S., at 224, is inescapably an expressive right. “To the extent that party labels *374provide a shorthand designation of the views of party candidates on matters of public concern, the identification of candidates with particular parties plays a role in the process by which voters inform themselves for the exercise of the franchise.” Tashjian v. Republican Party of Conn., 479 U. S. 208, 220 (1986).

In this case, and presumably in most cases, the burden of a statute of this kind is imposed upon the members of a minor party, but its potential impact is much broader. Popular candidates like Andy Dawkins sometimes receive nationwide recognition. Fiorello LaGuardia, .Earl Warren, Ronald Reagan, and Franklin D. Roosevelt are names that come readily to mind as candidates whose reputations and political careers were enhanced because they appeared on election ballots as fusion candidates. See Note, Fusion and the Associational Rights of Minor Political Parties, 95 Colum. L. Rev. 683 (1995). A statute that denied a political party the right to nominate any of those individuals for high office simply because he had already been nominated by another party would, in my opinion, place an intolerable burden on political expression and association.

II

Minnesota argues that the statutory restriction on the Party’s right to nominate the candidate of its choice is justified by the State’s interests in avoiding voter confusion, preventing ballot clutter and manipulation, encouraging candidate competition, and minimizing intraparty factionalism. None of these rationales can support the fusion ban because the State has failed to explain how the ban actually serves the asserted interests.

I believe that the law significantly abridges First Amendment freedoms and that the State therefore must shoulder a correspondingly heavy burden of justification if the law is to survive judicial scrutiny. But even accepting the majority’s view that the burdens imposed by the law are not weighty, *375the State’s asserted interests must at least bear some plausible relationship to the burdens it places on political parties. See Anderson, 460 U. S., at 789. Although the Court today suggests that the State does not have to support its asserted justifications for the fusion ban with evidence that they have any empirical validity, ante, at 364, we have previously required more than a bare assertion that some particular state interest is served by a burdensome election requirement. See, e. g., Eu, 489 U. S., at 226 (rejecting California’s argument that the State’s endorsement ban protected political stability because the State “never adequately explained] how banning parties from endorsing or opposing primary candidates advances that interest”); Anderson, 460 U. S., at 789 (evaluating a State’s interests, we examine “the extent to which those interests make it necessary to burden the plaintiff’s rights”); Norman v. Reed, 502 U. S. 279, 288-289 (1992) (“corresponding interest” must be “sufficiently weighty to justify the limitation”).3

While the State' describes some imaginative theoretical sources of voter confusion that could result from fusion candidacies, in my judgment the argument that the burden on First Amendment interests is justified by this concern is meritless and severely underestimates the intelligence of the *376typical voter.4 We have noted more than once that “[a] State’s claim that it is enhancing the ability of its citizenry to make wise decisions by restricting the flow of information to them must be viewed with some skepticism.” Eu, 489 U. S., at 228; Tashjian, 479 U. S., at 221; Anderson, 460 U. S., at 798.

The State’s concern about ballot manipulation, readily accepted by the majority, is similarly farfetched. The possibility that members of the major parties will begin to create dozens of minor parties with detailed, issue-oriented titles for the sole purpose of nominating candidates under those titles, see ante, at 365, is entirely hypothetical. The majority dismisses out-of-hand the Party’s argument that the risk of this type of ballot manipulation and crowding is more easily averted by maintaining reasonably stringent requirements for the creation of minor parties. Ibid. In fact, though, the Party’s point merely illustrates the idea that a State can place some kinds — but not every kind — of limitation on the abilities of small parties to thrive. If the State wants to make it more difficult for any group to achieve the legal status of being a political party, it can do so within reason and still not run up against the First Amendment. “The State has the undoubted right to require candidates to make a preliminary showing of substantial support in order to qualify for a place on the ballot, because it is both wasteful and confusing to encumber the ballot with the names of frivolous candidates.” Anderson, 460 U. S., at 788-789, n. 9. See also Jenness v. Fortson, 403 U. S. 431, 442 (1971). But once the State has established a standard for achieving party status, forbidding an acknowledged party to put on the ballot its chosen candidate clearly frustrates core associational rights.5

*377The State argues that the fusion ban promotes political stability by preventing intraparty factionalism and party raiding. States do certainly have an interest in maintaining a stable political system. Eu, 489 U. S., at 226. But the State has not convincingly articulated how the fusion ban will prevent the factionalism it fears. Unlike the law at issue in Storer v. Brown, 415 U. S. 724 (1974), for example, this law would not prevent sore-loser candidates from defecting with a disaffected segment of a major party and running as an opposition candidate for a newly formed minor party. Nor does this law, like those aimed at requiring parties to show a modicum of support in order to secure a place on the election ballot, prevent the formation of numerous small parties. Indeed, the activity banned by Minnesota’s law is the formation of coalitions, not the division and dissension of “splintered parties and unrestrained factionalism.” Id., at 736.

As for the State’s argument that the fusion ban encourages candidate competition, this claim treats “candidates” as fungible goods, ignoring entirely each party’s interest in nominating not just any candidate, but the candidate who best represents the party’s views. Minnesota’s fusion ban simply cannot be justified with reference to this or any of the above-mentioned rationales. I turn, therefore, to what appears to be the true basis for the Court’s holding — the interest in preserving the two-party system.

III

Before addressing the merits of preserving the two-party-system as a justification for Minnesota’s fusion ban, I should note that, in my view, it is impermissible for the Court to consider this rationale. Minnesota did not argue in its *378briefs that the preservation of the two-party system supported the fusion ban, and indeed, when pressed at oral argument on the matter, the State expressly rejected this rationale. Tr. of Oral Arg. 26. Our opinions have been explicit in their willingness to consider only the particular interests put forward by a State to support laws that impose any sort of burden on First Amendment rights. See Anderson, 460 U. S., at 789 (the Court will “identify and evaluate the precise interests put forward by the State as justifications for the burden imposed by its rule”); id., at 817 (Rehnquist, J., dissenting) (state laws that burden First Amendment rights are upheld when they are “‘tied to a particularized legitimate purpose’ ”) (quoting Rosario v. Rockefeller, 410 U. S. 752, 762 (1973)); Burdick, 504 U. S., at 434.

Even if the State had put forward this interest to support its laws, it would not be sufficient to justify the fusion ban. In most States, perhaps in all, there are two and only two major political parties. It is not surprising, therefore, that most States have enacted election laws that impose burdens on the development and growth of third parties. The law at issue in this case is undeniably such a law. The fact that the law was both intended to disadvantage minor parties and has had that effect is a matter that should weigh against, rather than in favor of, its constitutionality.6

*379Our jurisprudence in this area reflects a certain tension: On the one hand, we have been clear that political stability is an important state interest and that incidental burdens on the formation of minor parties are reasonable to protect that interest, see Storer, 415 U. S., at 736; on the other, we have struck down state elections laws specifically because they give “the two old, established parties a decided advantage over any new parties struggling for existence,” Williams v. Rhodes, 393 U. S. 23, 31 (1968).7 Between these boundaries, we have acknowledged that there is “no litmus-paper test for separating those restrictions that are valid from those that are invidious .... The rule is not self-executing and is no substitute for the hard judgments that must be made.” Storer, 415 U. S., at 730.

Nothing in the Constitution prohibits the States from maintaining single-member districts with winner-take-all voting arrangements. And these elements of an election system do make it significantly more difficult for third parties to thrive. But these laws are different in two respects from the fusion bans at issue here. First, the method by which they hamper third-party development is not one that impinges on the associational rights of those third parties; minor parties remain free to nominate candidates of their choice, and to rally support for those candidates. The small parties’ relatively limited likelihood of ultimate success on election day does not deprive them of the right to try. Second, the establishment of single-member districts correlates *380directly with the States’ interests in political stability. Systems of proportional representation, for example, may tend toward factionalism and fragile coalitions that diminish legislative effectiveness. In the context of fusion candidacies, the risks to political stability are extremely attenuated.8 Of course, the reason minor parties so ardently support fusion politics is because it allows the parties to build up a greater base of support, as potential minor party members realize that a vote for the smaller party candidate is not necessarily a “wasted” vote. Eventually, a minor party might gather sufficient strength that — were its members so inclined — it could successfully run a candidate not endorsed by any major party, and legislative coalition building will be made more difficult by the presence of third-party legislators. But the risks to political stability in that scenario are speculative at best.9

In some respects, the fusion candidacy is the best marriage of the virtues of the minor party challenge to entrenched viewpoints10 and the political stability that the two-party *381system provides. The fusion candidacy does not threaten to divide the legislature and create significant risks of factionalism, which is the principal risk proponents of the two-party system point to. But it does provide a means by which voters with viewpoints not adequately represented by the platforms of the two major parties can indicate to a particular candidate that — in addition to his support for the major party views — he should be responsive to the views of the minor party whose support for him was demonstrated where political parties demonstrate support — on the ballot.

The strength of the two-party system — and of each of its major components — is the product of the power of the ideas, the traditions, the candidates, and the voters that constitute the parties.11 It demeans the strength of the two-party system to assume that the major parties need to rely on laws that discriminate against independent voters and minor parties in order to preserve their positions of power.12 Indeed, *382it is a central theme of our jurisprudence that the entire electorate, which necessarily includes the members of the major parties, will benefit from robust competition in ideas and governmental policies that “ ‘is at the core of our electoral process and of the First Amendment freedoms.’ ” Anderson, 460 U. S., at 802, quoting Williams v. Rhodes, 393 U. S., at 32.

In my opinion legislation that would otherwise be unconstitutional because it burdens First Amendment interests and discriminates against minor political parties cannot survive simply because it benefits the two major parties. Accordingly, I respectfully dissent.

dissenting.

I join Parts I and II of Justice Stevens’s dissent, agreeing as I do that none of the concerns advanced by the State suffices to justify the burden of the challenged statutes on respondent’s First Amendment interests. I also agree with Justice Stevens’s view, set out in the first paragraph of Part III, that the State does not assert the interest in preserving “the traditional two-party system” upon which the majority repeatedly relies in upholding Minnesota’s statutes, see, e. g., ante, at 367 (“The Constitution permits the Minnesota Legislature to decide that political stability is best served through a healthy two-party system”). Actually, Minnesota’s statement of the “important regulatory concerns advanced by the State’s ban on ballot fusion,” Brief for Petitioners 40, contains no reference whatsoever to the “two-party system,” nor even any explicit reference to “political stability” generally. See id., at 40-50.

To be sure, the State does assert its intention to prevent “party splintering,” id., at 46-50, which may not be separable in the abstract from a desire to preserve political stabil*383ity.1 But in fact the State has less comprehensive concerns; the primary dangers posed by what it calls “major-party splintering and factionalism,” id., at 47, are said to be those of “turning] the general election ballot into a forum for venting intraparty squabbles,” ibid., and reducing elections to “a thinly disguised ballot-issue campaign,” id., at 49. Nowhere does the State even intimate that the splintering it wishes to avert might cause or hasten the demise of the two-party system. In these circumstances, neither the State’s point about “splintering,” nor its tentative reference to “political stability” at oral argument, n. 1, infra, may fairly be assimilated to the interest, posited by the Court of preserving the “two-party system.” Accordingly, because I agree with Justice Stevens, ante, at 378, that our election cases restrict our consideration to “the precise interests put forward by the State as justifications for the burden imposed by its rule,” Anderson v. Celebrezze, 460 U. S. 780, 789 (1983),2 1 would judge the challenged statutes only on the interests the State has raised in their defense and would hold them unconstitutional.

I am, however, unwilling to go the further distance of considering and rejecting the majority’s “preservation of the two-party system” rationale. For while Minnesota has made no such argument before us, I cannot discount the possibility of a forceful one. There is considerable consensus that party loyalty among American voters has declined significantly in the past four decades, see, e. g., W. Crotty, American Parties in Decline 26-34 (2d ed. 1984); Jensen, *384The Last Party System: Decay of Consensus, 1932-1980, in The Evolution of American Electoral Systems 219-225, (P. Kleppner et al. eds. 1981), and that the overall influence of the parties in the political process has decreased considerably, see, e. g., Cutler, Party Government Under the American Constitution, 134 U. Penn. L. Rev. 25 (1987); Sundquist, Party Decay and the Capacity to Govern, in The Future of American Political Parties: The Challenge of Governance 42-69 (J. Fleishman ed. 1982). In the wake of such studies, it may not be unreasonable to infer that the two-party system is in some jeopardy. See, e. g., Lowi, N. Y. Times, Aug. 23, 1992, Magazine, p. 28 (“ [Historians will undoubtably focus on 1992 as the beginning of the end of America’s two-party system”).

Surely the majority is right that States “have a strong interest in the stability of their political systems,” ante, at 366, that is, in preserving a political system capable of governing effectively. If it could be shown that the disappearance of the two-party system would undermine that interest, and that permitting fusion candidacies poses a substantial threat to the two-party scheme, there might well be a sufficient predicate for recognizing the constitutionality of the state action presented by this case. Right now, however, no State has attempted even to make this argument, and I would therefore leave its consideration for another day.