dissenting in No. 543.*
If it were the function of this Court to impose upon the States our own ideas of wise policy, I might be inclined to join my Brethren in compelling the Ohio election authorities to disregard the laws enacted by the legislature of that State. We deal, however, not with a question of policy, but with a problem of constitutional power. And to me it is clear that,, under the Constitution as it is written, the Ohio Legislature has the power to do what it has done.
I.
The Constitution does not provide for popular election of a President or Vice President of the United States, either nationally or on a state-by-state basis. On the contrary, the Constitution explicitly specifies:
“Each State shall appoint, in such Manner as the Legislature thereof may direct, a Number of Electors, equal to the whole Number of Senators and Representatives to which the State may be entitled in the Congress . ...”1 (Emphasis supplied.)
*49“The Electors shall meet in their respective states and vote by ballot for President and Vice-President . 2
Chief Justice Fuller, therefore, was stating no more than the obvious when he wrote for a unanimous Court in McPherson v. Blacker, 146 U. S. 1, more than 75 years ago:
“The Constitution does not provide that the appointment of electors shall be by popular vote, nor that the electors shall be voted for upon a general ticket, nor that the majority of those who exercise the elective franchise can alone choose the electors. It recognizes that the people act through their representatives in the legislature, and leaves it to the legislature exclusively to define the method of effecting the object.
“In short, the appointment and mode of appointment of electors belong exclusively to the States under the Constitution of the United States. . . Id., at 27, 35.
A State is perfectly free under the Constitution to provide for the selection of its presidential electors by the legislature itself. Such a process of appointment was in fact utilized by several States throughout our early history, and by one State, Colorado, as late as 1-876.3 Or a state legislature might nominate two slates of electors, and allow all eligible voters of the State to choose between them. Indeed, many of the States formerly provided for the appointment of presidential electors by *50various kinds of just such cooperative action of their legislatures and their electorates.4
Here, the Ohio Legislature has gone further, and has provided for a choice by the State’s eligible voters among slates of electors put forward by all political parties that meet the even-handed requirements of longstanding state laws. We are told today, however, that, despite the power explicitly granted to the state legislatures under Art. II, § 1, the Legislature of Ohio nonetheless violated the Constitution in providing for the selection of electors in this way. I can perceive no such constitutional violation.
I agree with my Brethren that, in spite of the broad language of Art. II, § 1, a state legislature is not completely unfettered in choosing whatever process it may wish for the appointment of electors. Three separate constitutional amendments explicitly limit a legislature’s power. The Fifteenth Amendment makes clear that if voters are to be included in the process, no voter may be excluded “on account of race, color, or previous condition of servitude.” The Nineteenth Amendment makes equally clear that no voter may be excluded “on account of sex.” And the Twenty-fourth Amendment prohibits exclusion of any voter “by reason of failure to pay any poll tax or other tax.” But no claim has been or could be made in this case that any one of these Amendments has been violated by Ohio.
*51Rather, it is said that Ohio has violated the provisions of the Fourteenth Amendment. The Court holds that the State has violated that Clause of the Amendment which prohibits it from denying “to any person within its jurisdiction the equal protection of the laws.” And two concurring opinions emphasize First Amendment principles, made applicable to the States through the Fourteenth Amendment’s guarantees, in summarily concluding that Ohio’s statutory scheme is invalid. I concede that the Fourteenth Amendment imposes some limitations upon a state legislature’s freedom to choose a method for the appointment of electors. A State may not, for example, adopt a system that discriminates on grounds of religious or political belief. But I cannot agree that Ohio’s system violates the Fourteenth Amendment in any way.
II.
In view of the broad leeway specifically given the States by Art. II, § 1, of the Constitution, it seems clear to me that the basic standard of constitutional adjudication under the Equal Protection Clause — a standard under which only “invidious discrimination” is forbidden — is the most stringent test that properly can be held applicable here. A single quotation should suffice to summarize that standard of equal protection:
“The constitutional safeguard is offended only if the classification rests on grounds wholly irrelevant to the achievement of the State’s objective. State legislatures are presumed to have acted within their constitutional power despite the fact that, in practice, their laws result in some inequality. A statutory discrimination will not be set aside if any state of facts reasonably may be conceived to justify it.” McGowan v. Maryland, 366 U. S. 420, 425-426.
*52The provisions enacted by the Ohio Legislature fully meet that standard.5
The laws of Ohio classify political parties, for purposes of access to that State’s ballot, according to size and strength.6 Those that timely demonstrate widespread support in the State may submit a slate of presidential electors to Ohio’s voters, while those that neither have participated in past elections nor can show the support of 15% of the voting public 90 days before a primary election may not.7 The appellants claim that these provisions discriminate against them. They assert that although Ohio may establish “reasonable” qualifying standards so that ballots do not become unwieldy, the *53strength of the American Independent Party is so substantial that no such requirement could possibly suffice to keep the Party’s candidates off the presidential ballot. Ohio’s requirements are so high, they contend, that the legislative purpose behind those requirements can be only to keep new parties — even those that, like the American Independent Party, have gained considerably more than “splinter” support — off the ballot. And such requirements, they conclude, thus deny .persons in their position equal protection of the laws.
Ohio for its part concedes that the legislative objective underlying the statutes in question is to prevent the appearance on its ballot of slates of presidential electors whose substantial party support has not been timely demonstrated. That the basic classification drawn by the provisions is not “irrelevant to the achievement of the State’s objective” — the traditional standard for judging the validity of a legislative classification under the Equal Protection Clause — is clear. The Court seems to concede as much, but nonetheless holds that the Ohio provisions are invalid — a result which may rest in part, I believe, upon possible doubts regarding the permissibility of the legislative objective itself. The propriety of that objective is, then, a critical issue for determination.
III.
I can discern no basis for the position that Ohio’s objective is in any way an illegitimate one. Surely a State may justifiably assert an interest in seeing that its presidential electors vote for the candidate best able to draw the support of a majority of voters within the State. By preventing parties that have not demonstrated timely and widespread support from gaining places on its ballot, Ohio’s provisions tend to guard against the possibility that small-party candidates will draw enough support to prevent either of the major contenders from obtaining *54an absolute majority of votes — and against the consequent possibility that election may be secured by candidates who gain a plurality but who are, vis-á-vis their principal opponents, preferred by less than half of those voting.8 Surely the attainment of these objectives is well within the scope of a State’s authority under our Constitution. One may perhaps disagree with the political theory on which the objectives are based, but it is inconceivable to me that the Constitution imposes on the States a political philosophy under which they must be satisfied to award election on the basis of a plurality rather than a majority vote.
In pursuing this interest Ohio has, at the same time, not completely prevented new parties from gaining access to that State’s ballot. It has authorized ballot position for parties that can demonstrate by petition the support of 15% of the voting public 90 days before a primary election is to be held. My Brethren seem to suggest that the percentage figure is set too high, and the date too early. But I cannot join in this kind of second-guessing. While necessarily arbitrary, Ohio’s standards can only be taken to represent reasonable *55attempts at accommodating the conflicting interests involved.9
Although Ohio’s provisions do not freeze the Republican and Democratic Parties into the State’s election structure by specific reference to those parties, it is true that established parties, once they become participants in the electoral process, continue to enjoy ballot position so long as they have polled 10% of the vote in the most recent Ohio gubernatorial election. It is suggested that the disparity between this figure and the 15% requirement applicable to new parties is invidiously discriminatory. But I cannot accept the theory that Ohio is constitutionally compelled to apply precisely the same numerical test in determining whether established parties enjoy widespread support as it applies in determining that question with regard to new parties.
It is by no means clear to me that as an abstract matter there are no differences between parties that have long been on the ballot in a State and those that have not, such as might justify disparate standards for determining in those two classes of cases when widespread support, required for ballot position, has been demonstrated. In any event, I cannot conclude that the disparity involved here denies equal protection of the laws. The difference in figures is a difference between the requirements for getting on and staying on the ballot. It seems to me to be well within the State’s powers to set somewhat different standards for those two requirements, so long as it applies them uniformly to all political parties. The only remaining argument would seem to be that the Republican and Democratic Parties never had to meet the 15% requirement: they were on the ballot in Ohio at the time the statutory scheme was *56enacted, and so have had only to make certain they remain on by meeting the 10% standard. But the Ohio Legislature could well have taken notice at the time the provisions were enacted that the parties which had polled over 10% of the vote in the most recent gubernatorial election — the Republican and Democratic Parties — had both demonstrated strength far beyond the 15% figure specified for ballot entry by new parties. It seems to me totally unrealistic, therefore, to conclude that this minor disparity in standards cannot be justified by “any state of facts [that] reasonably may be conceived.” McGowan v. Maryland, supra, at 426.
IV.
The Court's opinion appears to concede that the State’s interest in attempting to ensure that a minority of voters do not thwart the will of the majority is a legitimate one, but summarily asserts that this legitimate interest cannot constitutionally be vindicated. That assertion seems to echo the claim of my concurring Brethren- — -a claim not made by the appellants — that Ohio’s statutory requirements in some way infringe upon First Amendment rights. I cannot agree.
As the language of Art. II, § 1, and a great deal of history under that section make clear, there is no constitutional right to vote for presidential electors.10 I take it, therefore, that the First Amendment theory of my Brethren rests on the view that, despite the legitimacy of the objective underlying Ohio’s laws, those laws nonetheless have the effect of stifling the activity of persons who disagree with the major political parties now in existence. The concurring opinions cite a series of decisions protecting what has been termed the First *57Amendment right of association. NAACP v. Button, 371 U. S. 415; Bates v. Little Rock, 361 U. S. 516; NAACP v. Alabama, 357 U. S. 449; Thomas v. Collins, 323 U. S. 516; De Jonge v. Oregon, 299 U. S. 353. In my view, however, the principles on which those decisions were based do not call for today’s result.
*56“[T]he Constitution of the United States does not confer the right of suffrage upon any one . .-. .”
*57In Thomas v. Collins and De Jonge v. Oregon, supra, the very design of the statutes in question was to prevent persons from freely meeting together to advance political or social views. Ohio’s laws certainly are not of that nature. In the other three cases cited, all involving the activities of the National Association for the Advancement of Colored People, the statutes challenged were not on their face calculated to affect associational rights. We were able to determine with a good deal of certainty in those cases, however, (1) that application of the statutes to the NAACP would clearly result in a considerable impairment of those rights, and (2) that the interest said to underlie the statutes was insubstantial in the contexts presented. I believe that those conclusions should as a general matter be regarded as prerequisites to any holding that laws such as those involved here, which serve a legitimate state interest but are said to have some impact on First Amendment activity, are invalid. Cf. United States v. O’Brien, 391 U. S. 367.
In NAACP v. Alabama, supra, for instance, where the NAACP was ordered in accord with state law to disclose its membership lists, we outlined the issues as follows:
“We think that the production order, in the respects here drawn in question, must be regarded as entailing the likelihood of a substantial restraint upon the exercise by petitioner’s members of their right to freedom of association. Petitioner has made an uncontroverted showing that on past occasions revelation of the identity of its rank-and-file members has exposed these members to economic reprisal, loss of employment, threat of physical coercion, and *58other manifestations of public hostility. Under these circumstances, we think it apparent that compelled disclosure of petitioner’s Alabama membership is likely to affect adversely the ability of petitioner and its members to pursue their collective effort to foster beliefs which they admittedly have the right to advocate, in that it may induce members to withdraw from the Association and dissuade others from joining it because of fear of exposure of their beliefs shown through their associations and of the consequences of this exposure.
“We turn to the final question whether Alabama has demonstrated an interest in obtaining the disclosures it seeks from petitioner which is sufficient to justify the deterrent effect which we have concluded these disclosures may well have on the free exercise by petitioner’s members of their constitutionally protected right of association. . . .
“. . . The exclusive purpose [of the state authorities] was to determine whether petitioner was conducting intrastate business in violation of the Alabama foreign corporation registration statute, and the membership lists were expected to help resolve this question. The issues in the litigation commenced by Alabama by its bill in equity were whether the character of petitioner and its activities in Alabama had been such as to make petitioner subject to the registration statute, and whether the extent of petitioner’s activities without qualifying suggested its permanent ouster from the State. Without intimating the slightest view upon the merits of these issues, we are unable to perceive that the disclosure of the names of petitioner’s rank-and-file members has a substantial bearing on either of them. . . .” 357 U. S., at 462-464.
*59And in Bates v. Little Rock, supra, where an almost identical requirement was involved, we stated:
“On this record it sufficiently appears that compulsory disclosure of the membership lists of the local branches of the National Association for the Advancement of Colored People would work a significant interference with the freedom of association of their members. There was substantial uncontro-verted evidence that public identification of persons in the community as members of the organizations had been followed by harassment and threats of bodily harm. There was also evidence that fear of community hostility and economic reprisals that would follow public disclosure of the membership fists had discouraged new members from joining the organizations and induced former members to withdraw. This repressive effect, while in part the result of private attitudes and pressures, was brought to bear only after the exercise of governmental power had threatened to force disclosure of the members’ names. . . . Thus, the threat of substantial government encroachment upon important and traditional aspects of individual freedom is neither speculative nor remote.
“Decision in this case must finally turn, therefore, on whether the cities as instrumentalities of the State have demonstrated so cogent an interest in obtaining and making public the membership fists of these organizations as to justify the substantial abridgment of associational freedom which such disclosures will effect. . . .
“In this record we can find no relevant correlation between the power of the municipalities to impose occupational license taxes and the compulsory disclosure and publication of the membership fists of the local branches of the National Association for the *60Advancement of Colored People. . . .” 361 U. S., at 523-525.11
Here, there certainly is no comparable showing that Ohio’s ballot requirements have any substantial impact on the attempts of political dissidents to organize effectively. Such persons are entirely free to assemble, speak, write, and proselytize as they see fit. They are free either to attempt to modify the character of the established major parties or to go their own way and set up separate political organizations. And if they can timely demonstrate that they have substantial support within the State — according to Ohio’s reasonable standards for deciding that question — they may secure ballot position for the candidates they support. Ohio has restricted only their ability to secure ballot position without demonstrating that support. To me the conclusion that that single disability in any way significantly impairs their First Amendment rights is sheer speculation. As my Brethren’s surveys of ballot requirements in the various States suggest, the present two-party system in this country is the product of social and political forces rather than of legal restrictions on minority parties. This Court has been shown neither that in States with minimal ballot restrictions third parties have flourished, nor that in States with more difficult requirements they are moribund. Mere speculation ought not to suffice to strike down a State’s duly enacted laws.
Nor, I think, can we with any confidence conclude that Ohio’s interest in attempting to ensure that the will of the majority shall prevail is an insubstantial one. It requires more insensitivity to constitutional principles of federalism than I possess to tell Ohio that that interest is, ac*61cording to this Court’s scale of values, somehow unworthy of implementation.12 I cannot conclude, therefore, that First Amendment principles call for the result reached today.
y.
It is thought by a great many people that the entire electoral college system of presidential selection set up by the Constitution is an anachronism in need of major overhaul.13 As a citizen, I happen to share that view. But this Court must follow the Constitution as it is written, and Art. II, § 1, vests in the States the broad discretion to select their presidential electors as they see fit. The method Ohio has chosen may be unwise as a matter of policy, but I cannot agree that it violates the Constitution.14
U. S. Const., Art. II, §1. This provision represented a compromise among several conflicting views expressed at the Constitutional Convention regarding the most salutary method for choosing a President, most of which favored some method other than popular election. See McPherson v. Blacker, 146 U. S. 1, 28.
U. S. Const., Arndt. 12. The Twelfth Amendment also specifies the procedures for selecting a President and Vice President in the event that no candidate receives a majority of votes in the electoral college.
See McPherson v. Blacker, supra, at 35.
“[V]arious modes of choosing the electors were pursued, as, by the legislature itself on joint ballot; by the legislature through a concurrent vote of the two houses; by vote of the people for a general ticket; by vote of the people in districts; by choice partly by the people voting in districts and partly by the legislature; by choice by the legislature from candidates voted for by the people in districts; and in other ways . . . .” McPherson v. Blacker, supra, at 29.
For a fuller description of the diverse methods pursued by the States in appointing their electors under Art. II, § 1, during this Country’s first century of constitutional experience, see id., at 26-35.
It is clear that this Court’s decisions in such cases as Baker v. Carr, 369 U. S. 186; Gray v. Sanders, 372 U. S. 368; and Reynolds v. Sims, 377 U. S. 533, all involving the direct popular election of candidates to state or federal office, do not control the issues in this case. Indeed, no opinion today suggests that those cases are apposite. They sustained the right of a voter to cast a ballot whose numerical weight is the equal of that of any other vote cast within the jurisdiction in question. No claim is made in this case that Ohio has in any way violated that right.
The appellants plainly do not object to working through or voting for candidates of partisan political organizations, and I do not understand them to claim discrimination on the basis of Ohio’s failure to allow access to its presidential ballot via an “independent nominating petition.”
Appellants have cited us to a complex group of Ohio statutes which they say are relevant to the participation of political parties in that State’s presidential elections. See Ohio Rev. Code §§ 3505.10, 3513.05, 3513.11, 3513.19, 3513.191, 3517.01-3517.04. It is not entirely clear that all of those provisions are applicable to parties participating in the electoral process for the first time. But we need not examine that question since in any event the appellants clearly failed to file with the Secretary of State of Ohio on February 7 of this year, 90 days before the State’s primary election, a petition signed by a number of voters equal to 15% of the number participating in Ohio’s last gubernatorial election. Ohio Rev. Code §§ 3505.10, 3517.01.
This interest, which several States have chosen to protect in the context of state and local primary contests by providing for runoff elections, may be illustrated by a hypothetical example. Assume a State in which a dissident faction of one of the two major parties — party A — becomes dissatisfied with that party’s nominees and sets itself up as a “third party” — party C — putting forward candidates more to its liking. Still, the members of party C much prefer the candidates of party A to those of party B. A situation is possible in which party B’s candidates poll, for example, 46% of the vote, party A’s candidates 44%, and party C’s candidates 10%. Party B’s candidates would in such a situation be elected by plurality vote. In an election involving only the candidates of parties A and B, however, those persons preferring party C’s candidates might well have voted overwhelmingly for party A’s, thus giving party A’s candidates a substantial majority victory.
The date specified, for instance, is related to Ohio’s requirement that all political parties hold primary elections — another provision that is, it seems to me, well within the State’s power to enact.
Cf. Minor v. Happersett, 21 Wall. 162, 178:
The NAACP eases, furthermore, held invalid only the application of the state laws in question to the parties involved. Here, however, Ohio is told, as I read the opinion of the Court and the concurring opinions, that it cannot in any circumstances validly enforce its ballot requirements.
My Brother Hablan suggests that Ohio’s interest may be protected in “less restrictive” ways. In light of the views I have stated above, I do not see why Ohio should be compelled to utilize one method for achieving its ends rather than another. In any event, each of the methods mentioned by Mb. Justice Hablan appears to me to entail consequences which arguably would frustrate other legitimate state interests. Nor do all of them serve as effectively to promote the interest in question here as does the statutory scheme the Ohio Legislature has in fact enacted. I do not think problems such as those raised in this ease can be solved by means of facile and unelaborated suggestions of “less restrictive alternatives”; issues of legislative policy are too complex for such easy answers to be satisfactory.
Similar suggestions were being made as early as 1804, at the time of the adoption of the Twelfth Amendment. See McPherson v. Blacker, 146 U. S. 1, 33.
For the reasons stated in this opinion, and the further reasons stated in Part IV of the opinion of the Court, I agree with the Court’s denial of equitable relief to the appellants in No. 544, the Socialist Labor Party ease.