McGOWAN, Circuit Judge:
For the third time in four years, this court confronts a challenge from within one of the two major national political parties to the formula fixed by it for the allocation of delegates to its quadrennial national convention. In the earlier instances, divisions of this court held such challenges to be unavailing for want of merit. Today, for the reasons set forth hereinafter, the court en banc reaches the same result with respect to the present appeal.
I
The subject of the appeal is the delegate allocation formula adopted by the National Republican Party for its 1976 convention. The Ripon Society and nine individual plaintiffs1 have secured the judgment of the District Court that parts of that formula are unconstitutional. 369 F.Supp. 368 (D.D.C.1974). The ruling of the District Court is not the first one made in plaintiffs’ favor. In 1971 they sued to enjoin the use of a similar allocation formula at the 1972 Republican National Convention. Partial relief was granted in April of 1972. 343 F.Supp. 168 (D.D.C.1972). That judgment was stayed by Justice Rehnquist, 409 U.S. 1222, 93 S.Ct. 1475 (1972), and the convention was conducted as planned. Thereafter the appeal from the District Court was dismissed, and the complaint amended to state the present challenge to the 1976 formula.
That formula was adopted, on a vote of 910 to 434, by the delegates to the 1972 convention. It provides as follows: 1,605 delegates, or 72 percent of the total, are allocated according to the states’ electoral college votes, each state to receive three delegates per presidential elector; 312 delegates, or 14 percent, are awarded as “victory bonuses” to states voting for the Republican candidate in the last presidential election, each such state to receive a number of additional delegates equal to 60 percent of its electoral college vote, or 20 percent of its electoral college-based delegation (the “proportional victory bonus”); 245 delegates, or 11 percent, are divided equally among the states that voted for the last Republican presidential candidate, each such state to receive five delegates on this basis (the “uniform victory bonus”);2 50 delegates, or 2 percent, are [571]*571awarded to the states for Republican election successes at the state level, one such delegate for each Republican governor, senator, or majority of United States Representatives which the state elects in 1972 or a succeeding year prior to the 1976 convention (this bonus will be considered part of the “uniform victory bonus”);3 and 30 delegates, or 1 percent, are divided among the District of Columbia (14), Puerto Rico (8), Guam and the Virgin Islands (4 each).
Declaratory and injunctive relief was sought on the ground that the formula as a whole, and in particular its various victory bonus features, denied plaintiffs equal protection of the laws. Plaintiffs proposed that the Republican National Committee be permitted to fashion a new formula4 subject to the constraints that (1) a “substantial” number of delegates be allocated according to the Republican vote in one or more recent elections, (2) the remaining delegates be apportioned on the basis of population or electoral college vote, (3) the District of Columbia be treated for allocation purposes as a state, and (4) the territories receive a number of delegates no greater than what they would be entitled to on a population basis.
The district judge granted relief only in part. Ruling on cross-motions for summary judgment, he forbade the use of uniform victory bonuses, but upheld the formula in other respects. 369 F.Supp. at 376. Plaintiffs have appealed the denial of additional relief; defendants have appealed the granting of any relief at all.
II
Defendants assert that there are preliminary issues which, if rightly decided, preclude our reaching the merits. These involve the concepts, respectively, of standing to sue, state action, and justiciability. In this part of the opinion, we address these issues in succession.
A. Standing.
The standing requirement serves many purposes, including that of seeing to it that claims are prosecuted to binding resolution on the merits only by those with a sufficient interest to assure an informed and effective presentation. We would not wish to rule one way or the other in this case without satisfying ourselves that that requirement had been met.5
[572]*572We focus first on the individual plaintiffs. Each of the nine is alleged in the supplementary complaint to be a “citizen of the United States, a Republican, a registered and qualified voter of the District of Columbia or of . California, Illinois, Indiana, Massachusetts, Minnesota, New Jersey [or] New York.” We think that we may fairly take this as an assertion by each plaintiff of an interest in being represented, through the delegation of his or her state or district, at the National Republican Convention.6 We see no reason to differentiate, for purposes of the standing requirement, between that interest, and the interest of one seeking representation in a state or national legislature. There is of course no doubt that in the latter context an individual, claiming that his vote is diluted because his representative represents a greater number of constituents than do other representatives in the same assembly, has standing to challenge the constitutionality of the apportionment scheme. Baker v. Carr, 369 U.S. 186, 204-208, 82 S.Ct. 691, 7 L.Ed.2d 663 (1962). The only remaining question is whether the claims of malapportionment in this case are in fact made by plaintiffs whose representation would be improved if those claims were to prevail.
There appears to be at least one such plaintiff for each claim. It is argued that the formula as a whole deviates too far (in favor of the less populous states) from proportionality to electoral college representation, to total population, and to the Republican vote in past elections. If so, plaintiffs Halliwell, Silverman, White and Vradenburg, residents of California, New York, Illinois and New Jersey, respectively, clearly stand to benefit.7 Victory bonuses in [573]*573general are said to violate the Constitution. If so, there will be a clear benefit to plaintiff Behn, a resident of Massachusetts, the only state which did not cast its electoral vote for the 1972 Republican nominee.8 To the extent the victory bonuses are opposed only for the form they take, i. e., uniform and electoral college-based, the plaintiffs from California, New York, Illinois, and New Jersey once again have concrete reason to complain. Finally, the District of Columbia and the territories are allegedly over-represented. All plaintiffs other than Gillette are residents of states, and thus have a sufficient stake in the matter.
The standing of the Ripon Society is more doubtful. It describes itself as a “nationwide organization of young business, professional and academic men and women organized to engage the talents and energies of thinking young people in the cause of constructive Republicanism.” Its publications are said to provide “a forum for fresh ideas, well-researched proposals and a spirit of criticism within the Republican Party.” Its National Executive Committee authorized this action “to ensure fair and constitutional representation at Republican National Conventions.”
Yet the Society claims no harm to itself,9 nor even to any interest to which it is particularly dedicated.10 Certainly there may be harm to some of its members, but a party may not ordinarily assert the rights of others,11 and the Ripon Society has not made a strong case for being excepted from this rule. It makes no claim that its members are uniquely, or even predominantly, injured.12 It gives no reason to believe that those members who are adversely affected cannot assert their own rights.13 It [574]*574seems, in short, merely to have an Executive Committee whose “spirit of criticism” extends to the allocation formula for convention delegates.
Whether the standing requirement has been so far relaxed as to be satisfied by this coincidence alone is, however, a question we need not decide in this case. We have concluded that the individual plaintiffs had standing to bring this suit. The purpose of the requirement is to ensure the presence of a jurisdictional ease or controversy, and of “that concrete adverseness which sharpens the presentation of issues.”14 We cannot see how the Ripon Society’s participation in this case could lessen the controversy, or blur the presentation of issues, or alter the course of the litigation in any way.15 If the trial judge erred in refusing to strike the Ripon Society as a plaintiff for lack of standing, this error was harmless.
B. State Action.
As their claim is founded entirely on provisions of the United States Constitution which apply only to the federal and state governments, essential to the merits of plaintiffs’ case is the proposition that a national party’s allocation of convention delegates constitutes “state action.”16 We found such action in Georgia v. National Democratic Party, supra note 5 — the first of our prior decisions on delegate allocation. Our reasoning was that since the action of individual state parties in selecting their candidates, and indeed their convention delegates, was state action, it could not be otherwise when those parties acted through their delegates at the national convention. Id. at 1274-75. We also concluded that by placing the nominee of the convention on the ballot, the states “adopted this narrowing process as a necessary adjunct of their election procedures,” and thus subjected it to constitutional scrutiny. Id. at 1275-76. We adhered to our Georgia holding in Bode v. National Democratic Party, 146 U.S.App.D.C. 373, 452 F.2d 1302, 1304 (1971), cert. denied, 404 U.S. 1019, 92 S.Ct. 684, 30 L.Ed.2d 668 (1972). As the question now comes to us a third time, however, its answer is much less clear.
The Supreme Court has in the meantime decided Moose Lodge No. 107 v. Irvis, 407 U.S. 163, 92 S.Ct. 1965, 32 L.Ed.2d 627 (1972), and Jackson v. Metropolitan Edison Co., 419 U.S. 345, 95 S.Ct. 449, 42 L.Ed.2d 477 (1974), taking a different, and arguably narrower, view of what constitutes “state action” than is reflected in our Georgia decision.17 The [575]*575Supreme Court focused in each of those cases on the “nexus between the State and the challenged action” of the ostensibly private entity, rather than on that entity’s relationship to the state in general. 419 U.S. at 351, 95 S.Ct. 449. See also 407 U.S. at 173, 92 S.Ct. 1965. The “nexus” between the states and the delegate allocation formula is open to question, particularly since the Supreme Court has also now held, in Cousins v. Wigoda, 419 U.S. 477, 95 S.Ct. 541, 42 L.Ed.2d 595 (1975), that an individual state is without power to interfere with the delegate selection procedures of a national convention.18
Nor can we take much comfort from the obvious distinctions between these cases and ours. In O’Brien v. Brown, 409 U.S. 1, 92 S.Ct. 2718, 34 L.Ed.2d 1 (1972), the Democratic National Party sought review of an order of this court (152 U.S.App.D.C. 157, 469 F.2d 563) that a number of delegates disqualified by the Party’s Credentials Committee be seated at its 1972 convention. The Supreme Court per curiana stayed our order, professing “grave doubts” as to its propriety.19 Our finding of state action on the part of the Credentials Committee was specifically questioned. To be sure, O’Brien is itself distinguishable as a case not involving the allocation of delegates among the states. Justice Rehnquist at least would apparently not consider this distinction persuasive, since he relied exclusively on O’Brien when he stayed the District Court’s earlier order in this very case.20
If plaintiffs’ prospects for success on the state action issue have been somewhat dimmed by the actions of the Supreme Court, they have conceivably been somewhat brightened by those of the Congress. The Federal Election Campaign Act Amendments of 1974, approved on October 15, 1974, provide for federal financing of the 1976 presidential nominating conventions.21 The national committees of the Republican and Democratic Parties are each to receive two million dollars to defray convention expenses.22 The presidential primaries, in [576]*576which many of the delegates to those conventions will be elected, are to receive federal support in the form of a partial matching of the funds raised by candidates running in those primaries.23 The federal government is also to subsidize the general election campaigns of the party nominees to the tune (potentially) of twenty million dollars for each major party candidate.24
If the parties’ conventions, and their candidates, are to be so far underwritten by the federal government, then perhaps they must share its constitutional obligations. Of course the public financing provisions may never actually be put into effect. The Supreme Court will soon review an en banc decision of this court upholding the constitutionality of these public financing provisions.25 The entire matter is thus in a state of some uncertainty.
Because, as the Supreme Court said in O’Brien, the existence of state action is a difficult and “highly important question,” because it may well be a very different question when the matter of federal financing is settled, and because it cannot in any case affect the outcome of this appeal, we decline to decide it. As elaborated below, it is clear to us that plaintiffs’ case must fail on its merits without regard to whether or not there is state action, a question which we therefore expressly reserve.26
[577]*577C. Justiciability.
We pretermit the question of whether the validity of a national political party’s actions in apportioning convention delegates is justiciable in the federal courts. This also was a matter decided in Georgia, 447 F.2d at 1276—78, regarded as settled in Bode, 452 F.2d at 1305, and subjected to “grave doubts” in O’Brien, 409 U.S. at 4-5, 92 S.Ct. 2718. While it is not, like the state action question, one which rests upon shifting statutory ground, it is nonetheless one of obvious difficulty.
Georgia and Bode were decided in the wake of Baker v. Carr, 369 U.S. 186, 82 S.Ct. 691, 7 L.Ed.2d 663 (1962), and Powell v. McCormack, 395 U.S. 486, 89 S.Ct. 1944, 23 L.Ed.2d 491 (1969), two decisions which seemed to drain the “political question” doctrine of much of its vitality.27 The Court in the former case declared that “the mere fact that the suit seeks protection of a political right does not mean it presents a political question.” 369 U.S. at 209, 82 S.Ct. at 706. Such questions were said to arise from “the relationship between the judiciary and the coordinate branches of the Federal Government.” Id. at 210, 82 S.Ct. at 706. Arguably the question now before us is nonjusticiable under Baker because of a “lack of judicially discoverable and manageable standards for resolving it.” Id. at 217, 82 S.Ct. at 710. Yet the Baker Court considered “manageable” for legislative apportionment schemes a standard quite similar to the one we apply below to delegate apportionment schemes. It held that legislative schemes were to be struck down only if they “reflect ... no policy but simply arbitrary and capricious action.” Id. at 226, 82 S.Ct. at 715.
Powell appeared to curtail the “political question” doctrine even further. Faced with what might have been thought the classically “political question” of whether the House of Representatives — a coordinate branch — had properly excluded one of its members, the Court found the question justiciable. Significantly for our case, it rested that finding in part on the involvement of the right to choose one’s own representatives, which compelled it to “resolve any ambiguity in favor of a narrow construction of the scope of Congress’ power.” 395 U.S. at 547, 89 S.Ct. at 1977. In seemingly sharp contrast with this approach, we have been more recently admonished by the Court to exercise “great caution” before we “interject [ourselves] into the deliberative processes of a national convention, . . . involving as they do, relationships of great delicacy that are essentially political in nature.” O’Brien v. Brown, 409 U.S. at 4, 92 S.Ct. at 2720.
[578]*578In declining to decide the question of justiciability, we note its close relationship to the question we do decide, that is to say, the merits of the constitutional claim. Defendants’ arguments for nonjusticiability are that the “manageable” one person, one vote standard is inapplicable, and that without it a court has no basis for evaluating the political judgments that parties make in choosing among alternative delegate allocation schemes. Our view is not so different. We agree that a strict one person, one vote standard is inapplicable, and, since we consider the party’s choice among allocation schemes to be as much an exercise as an infringement of constitutional rights, we cannot say that it offends the Constitution. What we decline to do, however, is to take the more drastic step of holding that we would never be competent to reach a contrary conclusion.28
Ill
Having assumed arguendo that defendants are subject to justiciable constitutional limitations, we confront the question of whether those limitations have been exceeded in this case. Our discussion falls into two parts, the first dealing with what in general the Constitution requires in the allocation of delegates to a national political convention, and the second inquiring as to whether this particular formula satisfies those requirements.
A. Applicability of One Person, One Vote.
As noted above, plaintiffs rely primarily on the constitutional guarantee of equal protection. They analogize the Republican National Convention to the state legislatures in which that guarantee has been held to require representation on a “one man, one vote” basis. Plaintiffs propose that the constituency whose members are each to have “one vote” be either the entire population of a state, or that part of it that voted Republican in one or more past elections. Their entire argument is couched in terms of the challenged formula’s deviations from proportionality to those constituencies. Although they would apparently accept some such deviations, they would set as an outer limit the deviations present in the electoral college. The disproportionality introduced by the victory bonus system they do not consider justifiable. In short, they argue that, whichever of its permissible constituencies the Republican Party chooses to represent at a national convention, it must represent those constituents as a legislature would, and give them mathematically equal representation or have a [579]*579compelling reason not to. That is the essence of the claim, and also its essential fallacy.29
The fact that the conduct of a national political convention may be subject to the Equal Protection Clause does not in itself establish the applicability of the one person, one vote rule. Manifestly, a given constitutional command may not require of one part of the state what it requires of another. The army and the park commissioner are not equally constrained by the First Amendment; the President is not subject to the same restraints in making appointments as Congress is in passing legislation. And indeed it is clear that the Equal Protection Clause does not impose the same one person, one vote rule upon all elected and decision-making bodies, even if they are formally and indisputably organs of the state.
That rule is not, for example, applicable to the election of state judges. The plaintiffs in Wells v. Edwards, 347 F.Supp. 453 (M.D.La.1972), challenged the federal constitutionality of the provisions in the Louisiana Constitution for the election of the justices of that state’s supreme court from judicial districts of unequal population. It was held that “the rationale behind the one-man, one-vote principle, which evolved out of efforts to preserve a truly representative form of government, is simply not relevant to the makeup of the judiciary.” Id. at 455. The Supreme Court affirmed. 409 U.S. 1095, 93 S.Ct. 904, 34 L.Ed.2d 679 (1973) (memorandum).30
[580]*580At least one federal district court and four state supreme courts have similarly held that the one person, one vote principle is inapplicable to state constitutional conventions, which “can only make proposals which can have no effect unless ratified in another election [with] every vote given the same weight.” Driskell v. Edwards, 374 F.Supp. 1, 3 (W.D.La.), vacated, 419 U.S. 812, 95 S.Ct. 26, 42 L.Ed.2d 38 (1974),31 quoting from West v. Carr, 212 Tenn. 367, 370 S.W.2d 469, 474 (1963), cert. denied, 378 U.S. 557, 84 S.Ct. 1908, 12 L.Ed.2d 1034 (1964). Accord, Bates v. Edwards, 294 So.2d 532, 534 (La.1974); Stander v. Kelley, 433 Pa. 406, 250 A.2d 474, 481, cert. denied sub nom. Lindsay v. Kelley, 395 U.S. 827, 89 S.Ct. 2130, 23 L.Ed.2d 738 (1969); Livingston v. Ogilvie, 43 Ill.2d 9, 250 N.E.2d 138, 145-46 (1969).
We know also that the one person, one vote requirement, though generally applicable to local as well as state and federal assemblies,32 does not apply to certain “special purpose” assemblies whose decisions “disproportionately affect different groups.” Salyer Land Co. v. Tulare Lake Basin Water Storage District, 410 U.S. 719, 727, 93 S.Ct. 1224, 35 L.Ed.2d 659 (1973). Though it had required equal representation on local school and higher education boards,33 the Supreme Court in Salyer permitted the directors of a water storage district to be elected by agricultural land owners only, with the latters’ votes being weighted according to the valuation of their lands. It was reasoned that “the district does not exercise what might be thought of as ‘normal governmental’ authority” and that “all of the costs of district projects are assessed against land by assessors in proportion to the benefits received.” Id. at 729, 93 S.Ct. at 1230.
Obviously, these exceptions to the one person, one vote requirement occur in contexts too far removed from that of a national political convention for them to be dispositive of the case before us. They do demonstrate, however, that the principle of one person, one vote is not an absolute, to be unthinkingly invoked every time two or more persons are selected to make decisions on other people’s behalf, even if the making of those decisions is very plainly “state action.” The constitutional command is not one person, one vote but equal protection of the laws, and what it requires by way of representation in a given assembly must depend on the purposes for which the assembly is convened and the nature of the decisions it makes. The Supreme Court’s inquiry into these matters has led it to the conclusion that where the assembly exercises formal governmental powers one person, one vote is ordinarily required. A similar inquiry in other contexts may well reveal that the public and private interests in making decisions through some other scheme of representation outweigh the interests served by numerically equal apportionment.
We think the national political parties may validly so conclude with respect to [581]*581their presidential nominating conventions. As explained more fully below, these have never had, or been conceived as having, the function of providing a strict one person, one vote representation to a definable national constituency. Weighted representational schemes are only one of the numerous and scarcely distinguishable ways, some of them having judicial sanction, in which the political parties conduct their affairs in something short of a scrupulously democratic fashion. Most significant, the interests they advance by adopting representational schemes of their own choosing seem to us to be of great importance and of clearly constitutional stature.
Political parties are nowhere referred to in the Constitution — not surprisingly in view of their low reputation among our nation’s founders. Washington warned his countrymen “in the most solemn manner against the baneful effects of the spirit of party,”34 and Madison thought it a principal task of the new Constitution to hold the “mischiefs of faction in cheek.35 Nonetheless we have had political parties pretty much from the beginning,36 and with rare and early exceptions we have chosen our Presidents from among the candidates those parties put forward.37 Like the election itself, early political nominations for President were anything but democratic. Throughout the first part of the eighteenth century they were made by party caucuses — usually comprised of that party’s delegation to Congress. Credit for the downfall of “King Caucus” is generally given to Andrew Jackson who successfully fought for, and benefited from, a change to the ostensibly more democratic convention system.38 That system appears to have been firmly in effect in both major parties by 1840, and in substantially the same form that we are familiar with today.
From the start delegate votes at the convention were apportioned according to each State’s electoral college vote.39 Indeed it is somewhat ironic that the convention, the first major reform of the [582]*582nominating process, brought with it an apportionment scheme that bore no relation at all to the relative strength of the parties in the various states, whereas the congressional caucus reflected that strength quite accurately.40
There was recurrent criticism of electoral college-based apportionment, particularly in the Republican Party, where it gave inordinate control to delegates from southern States in which the Party had no hope of electoral success. Reform finally came in 1913, and in a way which presents a second irony for this case: the electoral college basis was in effect retained, but an extra vote was awarded to congressional districts which had voted Republican in past elections.41 Party strength was thus reinstated as a basis for delegate apportionment through a bonus vote system such as the one to which plaintiffs so strenuously object.42 The Party has employed some such mixture of electoral college-based and bonus votes ever since.43
We mention these few historical facts not because we think that what has been true in the past must always remain so, but merely to put the matter in perspective. In requesting that we impose a one person, one vote rule on the Republican Party, plaintiffs are not asking us to correct a historical aberration contrived at the 1972 convention. They are inviting us to take into judicial hands a process of change and adaptation that still continues within the Party.
We have twice declined that invitation in the Georgia and Bode cases. It is particularly significant that in the latter case we expressly upheld the parties’ long-standing practice of apportioning delegates according to electoral college strength.44 The Democratic Party formula challenged in Bode allocated 54 percent of the delegates on that basis. Although it allocated the remaining 46 percent according to Democratic voting strength as measured in past elections, it was not for that reason, or even because of the analogy to the electoral college itself, that we sustained the formula. Electoral college apportionment was perceived to have an “independent rationality for its use,” which was that it reflected “a judgment exercised toward maintaining and enlarging party appeal on a national scale.” 452 F.2d at 1309.
In thus upholding electoral college apportionment we have in effect already discarded the notion that national convention delegates must represent some constituency on a one person, one vote [583]*583basis.45 Electoral college apportionment obviously is not related to any set of Republican Party members or adherents. It bears some relation to total population, but only a very rough one.46
There are any number of other party practices which are seemingly inconsistent with a strict one person, one vote requirement, and which either have survived judicial scrutiny or, we think, would surely do so if challenged. They at least demonstrate the size of the task that the courts would be undertaking were they to impose a one person, one vote rule on the presidential nominating process.
The Republican Party does much of its business through a National Committee, which is malapportioned to the extent of being comprised of two members from each state.47 Moreover, one of the members must be a woman and one a man, a condition of dubious validity in respect of membership in, let us say, a state legislature.48 A committee system of such malapportioned and predetermined membership has historically dominated both parties down to their grass roots.49 They have been excused from equal representation requirements on the ground that they only administer the party’s “internal affairs,”50 but the distinction is not a strong one. In addition to conducting the national convention51 the National Committee dispenses patronage and party funds.52 It makes numerous important political decisions during the periods between national conventions— whose policies are favored in party publications and pronouncements, whose local campaigns are aided by appearances of nationally prominent party members, and so on. The fortunes of presidential hopefuls rise and fall with such decisions.
Turning to the delegates themselves, a party might well wish to impose conditions on delegate selection which are inconsistent with an unconstrained, mathematically equal system of representation. The Democratic Party recently did so by establishing quotas for the membership in state delegations of minorities, wom[584]*584en, and young people.53 Could a national convention take the more drastic step of refusing for some reason perceived to be in the Party’s best interests to seat a State’s delegation at all? Apparently, so. The Supreme Court stated in Cousins v. Wigoda that “[t]he Convention was under no obligation to seat the respondents [whom the Illinois court ordered seated] but was free ... to leave the Chicago seats vacant and thus defeat the objective.” 419 U.S. at 488, 95 S.Ct. at 548.
“Equal” apportionment of delegates among the states is presumably sought in order to insure “equal” representation of the people in those states. Yet the actual selection of delegates at the state level varies from the highly democratic to the opposite extreme. In a number of states the selection is made not in a primary election but through a series of local, county, and state caucuses and conventions. Often these are malapportioned,54 and often voter participation is so slight as to make the selection process one virtually (or even officially) of appointment by party officials.55 A practice that is more defensible perhaps, though scarcely more “democratic,” is the granting of ex officio delegate status to party officials or public office holders, presumably because of their special wisdom and expertise.56
There are a number of respects, then, in which the parties conduct their affairs other than by giving equal attention to the preferences of all voters, or even all party adherents.57 Perhaps this is not [585]*585surprising. A party is after all more than a forum for all its adherents’ views. It is an organized attempt to see the most important of those views put into practice through control of the levers of government.58 One party may think that the best way to do this is through a “strictly democratic” majoritarianism. But another may think it can only be done (let us say) by giving the proven party professional a greater voice than the newcomer. Which of these approaches is the more efficacious we cannot say, but the latter certainly seems a more accurate description of how political parties operate in reality.
What is important for our purposes is that a party’s choice, as among various ways of governing itself, of the one which seems best calculated to strengthen the party and advance its interests, deserves the protection of the Constitution as much if not more than its condemnation. The express constitutional rights of speech and assembly are of slight value indeed if they do not carry with them a concomitant right of political association. Speeches and assemblies are after all not ends in themselves but means to effect change through the political process. If that is so, there must be a right not only to form political associations but to organize and direct them in the way that will make them most effective.
The Supreme Court has frequently stressed the close kinship of the freedoms of speech and of political association. See, e. g., Kusper v. Pontikes, 414 U.S. 51, 56-57, 94 S.Ct. 303, 38 L.Ed.2d 260 (1973); NAACP v. Alabama ex rel. Patterson, 357 U.S. 449, 460, 78 S.Ct. 1163, 2 L.Ed.2d 1488 (1958). It has declared that “[a]ny interference with the freedom of a party is simultaneously an interference with the freedom of its adherents.” Sweezy v. New Hampshire, 354 U.S. 234, 250, 77 S.Ct. 1203, 1212, 1 L.Ed.2d 1311 (1957). It has invoked the First Amendment to strike down state restrictions on access to the general election ballot, stating that “[t]he right to form a party for the advancement of political goals means little if a party can be kept off the election ballot and thus denied an equal opportunity to win [586]*586votes.” Williams v. Rhodes, 393 U.S. 23, 30-31, 89 S.Ct. 5, 11, 21 L.Ed.2d 24 (1968).59
Last term the Court in Cousins v. Wigoda placed the internal workings of a political party squarely within the protection of the First Amendment. The delegates whom the Illinois courts had ordered seated in that case had been selected in accordance with Illinois law, but not the Democratic Party’s Guidelines. Illinois claimed “a compelling interest in protecting the integrity of its electoral process.” The Court held that this interest could not prevail against the “constitutionally protected rights of association” which the Party exercised in seating the delegation of its choice. 419 U.S. at 489, 95 S.Ct. 541. Nor can the case be explained as one in which the Court was preferring the delegates who were elected in the clearly more democratic way: those whom the Illinois courts ordered seated were elected in primaries; those who were seated by the Party were chosen in private caucuses. Id. at 478-80, 95 S.Ct. 541. If First Amendment rights are exercised when a Party determines the make-up, or perhaps even the existence, of state delegations, we think the same is true when it determines their size.
The First Amendment is of course not our only concern. We are keenly aware that “[a]s a practical matter, the ultimate choice of the mass of voters is predetermined when the nominations have been made.”60 If the right to vote is a right to true participation in the elective process, then it is heavily implicated in the nomination process. We do not deny this, but rest our judgment on the view that, as between that right and the right of free political association, the latter is more in need of protection in this case. As is further elaborated below, the right to organize a party in the way that will make it the most effective political organization seems clearly at stake here. The right of one person to one vote is of course preserved in the general election. Theoretically at least, persons dissatisfied with the choice facing them in that election may gain access to the ballot by means other than a major party nomination. They could, of course, form their own party. Of more practical importance is the fact that there are two major parties, which for the time being at least are in intense competition. Persons not heard in one party may be welcomed in the other, and if there are enough such defections, the offending party may lose the general election, as both parties must be well aware.61
We conclude, therefore, that the Equal Protection Clause, assuming it is applicable, does not require the representation in presidential nominating conventions of some defined constituency on a one person, one vote basis. It is satisfied if the representational scheme and each of its elements rationally advance some legitimate interest of the party in [587]*587winning elections or otherwise achieving its political goals. We turn, then, to the question of whether the challenged formula meets this test.
B. Justifications of the Challenged Formula.
By far the greatest number of delegates (72 percent) are' allocated according to the electoral college vote of the States. We upheld this as a basis of representation in Bode; and we pursue the matter further only because the justification for electoral college apportionment carries within it much of the justification for this entire formula. A State’s share of the electoral college vote only roughly approximates its share of the population. It precisely reflects the relative importance of the state to the party in terms of winning the presidential election. As we stated in Bode, “the primary function of a national party convention . . . is to select among a field of available persons Presidential and Vice-Presidential candidates most competent to perform the duties of office, yet capable of attracting a sufficient number of popular votes to carry the requisite number of States in the election.” 452 F.2d at 1309 (emphasis added).
The “requisite number of States” is the number with a majority of electoral college votes. The delegates from those states will presumably know best what kind of candidate is likely to carry them. It may be helpful, or even necessary, to have running mates who actually come from those states. If so, it may be wise, especially if it is thought that delegates from other states may ignore this fact in favor of their “favorite sons,” to “build in” for candidates from large states the advantage of large home-state delegations. Assuming, as we have, the constitutional validity of delegate allocation measures taken to improve the Party’s chances for victory, this one is hardly irrational.
It could stand some improvement, of course. As between two states of equal electoral importance, a party could more profitably focus its attention on the one in which it has a chance of victory. This purpose we think is rationally served by the victory bonus system. A state which has gone Republican in the past may do so again. If electoral college apportionment weights the vote of the states according to the value of the prize, the victory bonus system does the same according to the likelihood of winning it.
The victory bonus system may help to keep a state in the Republican camp not only by orienting party policies to that state’s interests, but also by providing a reward and incentive for the efforts of that state’s party organization. Whether or not it is an effective incentive, it may be the only one that a national party has to offer. In any ease, having accepted the legitimacy of such party-strengthening measures, we can hardly say that it is irrational.
It is most rational, perhaps, with respect to the “proportionate” victory bonuses, since a likelihood of success is more important in a large state than a small one. Not surprisingly, then, it is the “uniform” bonuses which have come in for the severest criticism from plaintiffs, not to mention the district judge. But these, too, seem to us to be rationally conceived. Success in a given state may have a certain value to the party quite out of proportion to the state’s size. When the returns are counted, on a state-by-state basis, carrying Arkansas will in a sense give the party the same psychological boost as carrying Tennessee (more than twice its size). And carrying a state, Arkansas or Tennessee, may also mean electing two Senators and a Governor on the presidential coattails. Holding these offices undoubtedly has a certain “uniform” importance to the party, in terms of the immediate power they represent and also because they are the springboards for future Presidential candidates. These considerations seem to us adequately to justify the uniform victory bonus of six votes for success in a Presidential election. If the bonus for success in state level elections requires a separate justification, we [588]*588think it is present in the need, which plaintiffs so heavily stress, to measure party strength by success in more than one past election.
Objection is also made to the allocations to the District of Columbia (14) and to Guam and the Virgin Islands (4 each). The fourteen delegates from the District of Columbia clearly represent disproportionately few voters, but do they disproportionately represent the importance of this particular governmental unit to the party? No one who has lived in the District can deny the special importance of political ambiance in the city in which the Nation’s business is done. As for the territories, their four votes each will hardly comprise a decisive bloc at the convention; and the party might well have concluded that a delegation of four is the smallest that it makes any sense to bring the considerable distance to the convention site.
There are other justifications for the various elements of this formula, as there are certainly countervailing objections to the ones we have mentioned. The uniform importance of the states might be thought adequately reflected in electoral college vote. Their importance in terms of party strength could perhaps be more accurately reflected by allocating delegates according to the size of the party vote in past elections. And the chances for success in future elections might be maximized by paying more attention to “swing” states than “safe” ones. Yet it is the essence of the First Amendment rights, which the parties exercise, that they may make their own contrary (and rational) judgments without interference from the courts.
It is urged that this formula represents nothing more than an effort by party members from strongly Republican states to perpetuate their control.62 But it seems to us that the First Amendment protects their power to do precisely that. The Party could have chosen a delegate allocation scheme calculated to broaden its base, by giving special influence to delegates from States where the party is weak. Instead it appears to have chosen to consolidate its gains in states where it has been strong. We are not about to hold that this is an irrational way to seek political success. As for those aspects of the formula which treat the states on a uniform basis and thus give disproportionate influence to the smaller states, how could we say that they do not rationally serve the important cause of cohesiveness among the various state parties, when it took precisely such a scheme to bring about the union of the states themselves?63
We therefore hold that the formula does not violate the Equal Protection Clause. To the extent that voting rights are involved, warranting close judicial scrutiny, these rights are offset by the First Amendment rights exercised by the Party in choosing the formula it did. We must emphasize that this is only true because the formula rationally advances legitimate party interests in political effectiveness. The same might not always hold true. There are no racial or other invidious classifications here.64 If there were, the Party’s entitlement to constitutional protection would be as slight as those of the victims would be strong. Similarly, we have said that voting [589]*589rights are not as heavily implicated in a nomination as in an election. It might be otherwise in a case where there is only one party with a realistic chance to win the election, and where a vote in the nominating process is the only effective vote that can be cast.65
These caveats have no significance in the present context other than to suggest that, although courts should be slow to interfere with the internal processes of political parties, circumstances can be conceived of wherein they may grant relief. Where such circumstances do not exist, Georgia, Bode, and this case should serve to discourage resort to this court for the resolution of intra-party differences.66
The judgment of the District Court is reversed, and the case is remanded with directions to dismiss the complaint.
It is so ordered.