BAZELON, Chief Judge:

The Ripon Society is a research and policy organization composed of young business, professional, and academic people who are associated with the National Republican Party. The Ripon Society and nine individual plaintiffs who are registered Republican voters in seven states and the District of Columbia challenge the formula for apportionment of delegates to the 1976 Republican National Convention. This challenge is directed against the National Republican Party and the Republican National Committee. This action was originally filed prior to the 1972 convention and resulted in an order of the District Court granting relief in part to plaintiffs and denying relief in part. The plaintiffs and defendants both appealed with defendants requesting a stay of the District Court order until after the 1972 Republican Convention. This Court denied a stay but Justice Rehnquist granted that relief. After the election, both parties moved to dismiss the appeal and it was so ordered. The plaintiffs then filed a second complaint challenging the formula for apportionment of delegates to the 1976 convention. The District Court again granted relief in part and denied relief in part. Both parties now appeal from that order.

The challenged apportionment formula is set out in Rule 30 adopted by the 1972 Republican National Convention. The District Court summary of this formula is reproduced in the margin. The focus of plaintiffs' concern are the 607 delegates, representing 27% of the total of approximately 2,242 delegates to the 1976 Convention, which are apportioned on the basis of a Republican victory in the 1972 presidential election and on the basis of Republican victories in 1972 and 1974 senatorial, gubernatorial and congressional elections. These so-called "victory bonus" delegates are apportioned in part on a uniform basis of 4.5 delegates for each state that voted Republican in 1972 (hereinafter referred to as a "uniform" victory bonus) and apportioned in part on a proportional basis in which each state that voted Republican in 1972 is given additional delegates totalling 60% of the electoral college vote of that state (hereinafter referred to as a "proportional" victory bonus). These two kinds of "victory bonus" delegates, plaintiffs allege, create an invidious discrimination against certain regions and certain states and are thus apportioned in violation of Article II, Section 1 of the Constitution and either the Fifth Amendment or the Fourteenth Amendment. The District Court in its second consideration of the matter upheld the use of the "victory bonus" concept but forbade further use of "uniform" victory bonuses in which each state received an equal number of additional delegates in return for victory regardless of the population of the state. We find, for reasons expressed below, that the "victory bonus" concept is unconstitutional in toto and, therefore, reverse the order of the District Court in part, affirm it in part and remand the cause with instructions.

I. THE EFFECT OF THE VICTORY BONUS

Plaintiffs have presented a rather detailed, but in part incomplete, statistical showing of the effect of the victory bonus concept. The heart of this showing is that the concept produces very significant deviations among the states as to (1) the number of 1972 Republican votes for President represented by delegates from each state; (2) the number of 1968-71 Republican votes for President and various state-wide offices represented by delegates from each state; and (3) the amount of total population represented by delegates from each state. According to this presentation, which was not challenged before the District Court, the maximum deviation in category (1) is 11.19 to 1. That is, each delegate from Florida, under the victory bonus scheme, would represent 28,148 1972 Republican votes for President while each delegate from the District of Columbia would represent 2,515 such votes. The maximum deviation in category (2) is 12.05 to 1 with each delegate from Pennsylvania representing 25,181 1968-71 Republican votes while each delegate from Alaska represents 2,089 of such votes. The maximum deviation in category (3) is 7.44 to 1 (if the victory bonus is applied in terms of the 1972 election) or 8.67 to 1 (if it applied in terms of the 1968-71 elections). As applied in terms of the 1972 election, each delegate from Massachusetts would represent 132,306 citizens while each delegate from Alaska would represent 17,775 citizens. As applied in terms of the 1968-71 elections, each delegate from New York would represent 147,892 citizens while each delegate from Alaska would represent 16,787 citizens.

The major weakness in plaintiffs' showing is the failure to provide us with data on the average deviation from the mean or ideal number of 1972 Republican voters, 1968-71 Republican voters or total population which each delegate should represent. A summary analysis of plaintiffs' raw data, however, indicates that this average deviation is as significant statistically as the maximum deviations discussed above. For example, the mean number of citizens which should be represented by each delegate (if the victory bonus is applied in terms of the 1972 election) is 91,278; the mean number which should be represented by each delegate (if the victory bonus is applied in terms of the 1968-71 elections) is 99,993. Making the heroic assumption that a 5% deviation should be statistically insignificant, we find that the victory bonus applied in terms of 1972 will produce 12 states whose delegates will each represent more than 91,278 plus 5% and 27 states whose delegates will represent less than 91,278 minus 5%. Even more significant results than these are obtained if the victory bonus is applied in terms of the 1968-71 elections. As so applied, the victory bonus will produce 17 states whose delegates will each represent more than 99,993 plus 5% and 28 states whose delegates will each represent less than 99,993 minus 5%. We have no reason to believe that the application of this analysis to the mean representation of 1972 Republican voters or of 1968-71 Republican voters will not produce similar results. The legal significance of these statistics is, of course, a question wholly distinct from their statistical significance.

II. THRESHOLD ISSUES: STATE OR GOVERNMENTAL ACTION AND JUSTICIABILITY

Defendants initially challenge the jurisdiction of the District Court, alleging that there is no "state action" or "governmental action" involved in either the Republican National Committee or the National Republican Party. They furthermore argue that the plaintiffs' claims are not justiciable and that, in any event, plaintiff Ripon Society has no standing to assert such claims. They finally argue, as a threshold matter, that the Republican National Party is not a juristic entity and has no legal capacity to sue or be sued. All of these claims, with the exception of the jus tertii standing of the Ripon Society, were considered and rejected in Georgia v. National Democratic Party, 145 U.S.App.D.C. 102, 447 F.2d 1271, 1273 n. 2, 1274-78, cert. denied, 404 U.S. 858, 92 S.Ct. 109, 30 L.Ed.2d 101 (1971). These holdings in Georgia were followed without dissent in Bode v. National Democratic Party, 146 U.S.App.D.C. 373, 452 F.2d 1302, 1304-05 (1971), cert. denied, 404 U.S. 1019, 92 S.Ct. 684, 30 L.Ed.2d 668 (1972) and without comment in O'Brien v. Brown, 152 U.S.App.D.C. 157, 469 F.2d 563, 567, stayed 409 U.S. 1, 92 S.Ct. 2718, 34 L.Ed.2d 1, dismissed as moot, 409 U.S. 816, 93 S.Ct. 67, 34 L.Ed.2d 72 (1972). As to the jus tertii standing of the Ripon Society, even though the right to vote is "personal", it is settled that an organization may assert the rights of its members. There is no claim that the members of the Ripon Society do not have standing.

The defendants would have us reconsider these recent decisions rendered by seven different judges of this Court on the basis of the per curiam opinion issuing a stay in O'Brien v. Brown, 409 U.S. 1, 92 S.Ct. 2718, 34 L.Ed.2d 1 (1972) and on the basis of Justice Rehnquist's opinion as Circuit Justice issuing a stay in this case. In those opinions the Court and Justice Rehnquist indicated "grave doubts" about our holdings in O'Brien on the issues of state action and justiciability. The Court did not state the source of its doubts and we agree with the District Court that O'Brien can be distinguished as concerning the internal affairs of a political party and not the right of equal access to decision of those internal affairs. This is a distinction of considerable subtlety which we shall discuss in more detail below. Despite the fact that O'Brien is distinguishable on its facts and involved only a bare statement of "grave doubts", we think it appropriate to reconsider the holdings of Georgia and to reaffirm them in their entirety.

The Georgia court relied on two related concepts to find state action in the apportionment of delegates to the national conventions of the two major political parties. First, the court correctly noted that state action had been found in the conduct of party primaries, conventions and committees for the purpose of proscribing racial discrimination. If this amount of state action be conceded, the court found and we think the logic is inescapable that the actions of the state's delegates acting in concert to apportion their own strength and thereby determine the weight to be given their constituents' previous vote in the state delegate selection procedures must also be state action. These cases upon which the Georgia court relied might be distinguished as involving racial discrimination, an involvement which apparently decreases the "amount" of state action needed to activate the commands of the Fourteenth or Fifth Amendment. However, the Supreme Court in Gray v. Sanders, 372 U.S. 368, 374-75, 83 S.Ct. 801, 9 L.Ed.2d 821 (1963) seemed to take a broader view of these racial discrimination cases, at least in terms of party primaries.

But even if these entirely reasonable arguments are not persuasive, we think the Georgia court also correctly determined that "by placing the nominees' names on the ballot, the states, in effect, have adopted [the] narrowing process [produced by the major party primaries] as a necessary adjunct of their election procedures." Indeed, the Supreme Court has recently determined that the state has a "not only permissible, but compelling" interest in enforcing this narrowing process. Furthermore, the narrowing process may be seen as a "governmental function" and thus be implicitly state action. Finally, the states and the federal government have served to institutionalize the major parties and by implication their narrowing process through protection of incumbents in reapportionment and perhaps in the future through public financing of both primary and general election campaigns of major party candidates. These "state action" arguments, of course, do not apply to the apportionment of delegates or committee people responsible for "the internal management and business" of the party. See Seergy v. Kings County Republican County Committee, 459 F.2d 308, 313-14 (2d Cir. 1972).

Once having determined that state action is implicated in delegate apportionment, the threshold issue of justiciability is answered by Baker v. Carr, 369 U.S. 186, 82 S.Ct. 691, 7 L.Ed.2d 663 (1962) as the Georgia court correctly held. It is difficult to conceive of a different result. However, as the Georgia court was careful to point out, a finding of justiciability does not eliminate consideration of the policies that underlie the doctrine of justiciability or, indeed, the policies that underlie the "state action" doctrine. Of particular relevance here, in our view, is the institutional structure promoted by the First Amendment. See Cousins v. Wigoda, 419 U.S. 744, 95 S.Ct. 541, 547-48, 42 L.Ed.2d 595 (1975). Under that structure, political thought and organization are delegated to the people; the government may not intervene, through any branch including the judiciary, except in the most narrow and compelling circumstances. To be sure, courts have seemingly permitted some intervention into this power reserved for the people to protect the "integrity" of the electoral process and, indeed, to protect the First Amendment or other rights of members of certain associations invested with monopoly or quasi-monopoly powers. The sum total of this discussion is that the policies forwarded by the doctrines of justiciability and state action should not be permitted to foreclose all judicial intervention into the affairs of major political parties. Rather those policies are implicated in the merits of whatever judicial intervention is attempted. In point of fact, the Equal Protection Clause itself could be read to make all disputes justiciable; but in evaluating equal protection claims, particularly as they relate to an evaluation of a legislative purpose, the policies of justiciability — particularly the notion of institutional competence — are a source of the traditional judicial deference to legislative action. This use of policies of justiciability and state action may be seen in recent reapportionment case law. We thus think our inquiry should be directed to specific claims and specific associational activities before the policies of justiciability and state action may be brought into full view. The discussion herein is, of course, limited to the activities of the major political parties, the Republicans and the Democrats, and we intimate no opinion on whether it might also be applicable to any independent or minor party. Such an application would clearly require an extension of the reasoning stated above.

III. DEVIATIONS FROM THE ONE PERSON-ONE VOTE STANDARD AND JUSTIFICATIONS THEREFOR

1. Having found sufficient indicia of state action and justiciability, we address plaintiffs' substantive contention: that the apportionment created by the victory bonus formula violates the principle that each voter is entitled to have his or her vote counted equally with other voters. The findings of state action and justiciability, of course, establish that this principle must apply to some extent but as the Georgia court warned those findings do not establish that the principle will operate equally on political parties as on state and federal legislative bodies. Before turning to possible differences in application, we initially note that the figures discussed in Part I above establish that the deviations from the standard of one person-one vote created by the victory bonus are not sufficiently de minimis to foreclose judicial intervention to reduce that deviation. The deviations mentioned in Part I above are far in excess of the deviations permitted in Gaffney v. Cummings, 412 U.S. 735, 737, 93 S.Ct. 2321, 37 L.Ed.2d 298 (1973), assuming that the more lax standard applied to the states is proper to test apportionment of a national political party. They are more than that permitted in Mahan v. Howell, 410 U.S. 315, 319, 93 S.Ct. 979, 35 L.Ed.2d 320 (1973), assuming that the Mahan court did not rely on the existence of a legitimate justification for the deviation involved therein. We note that the plaintiffs' data is not complete, particularly on the issue of the average deviation from the mean or ideal delegate representation but in the absence of any challenge in the District Court, we think it is sufficient to indicate a prima facie violation of the one person-one vote standard. If the defendants wish to further challenge plaintiffs' data they may do so by proper motion in the District Court.

2. This Court held in the Bode case that a political party may deviate from the one person-one vote standard and apportion delegates on the basis of the electoral college strength of a state and not on the total population of the state alone. The Court's reasoning is sound and we hereby reaffirm its holding. The maximum deviation among the states between the number of 1972 Republican votes or the amount of total population represented by each electoral college vote is 4.4 to 1. The deviations discussed in Part I above are substantially in excess of this figure and are not de minimis under the standards of Mahan and Gaffney. However, we note that those deviations are produced solely by the victory bonus and not by the apportionment formula which allocates the remaining 72% of the delegates. Plaintiffs do not challenge that apportionment.

3. The decisions in both Bode and Georgia establish further that a political party may also deviate from the one person-one vote standard to approximate a different principle of one Republican or one Democrat-one vote. The party is not required to make this approximation and, indeed, if it does, it need not apportion all the delegates on that basis for a reason central to the disposition of this cause which will be discussed in more detail below. But the victory bonus is not directed to this purpose. It would seem highly irrational if it did. We think the true purpose is to approximate party strength not in terms of persons but in terms of states. The victory bonus establishes a one Republican state-many votes standard.

4. The District Court approved the concept of a victory bonus because it held that a measurement of party strength in terms of states instead of in terms of people who have in the past or may in the future vote Republican was permissible under the electoral college system. Without the electoral college system, such a scheme would, of course, be invalid under the principle of Gray v. Sanders, 372 U.S. 368, 83 S.Ct. 801, 9 L.Ed.2d 821 (1963), if it produced, as the victory bonus produces, a substantial deviation from the one person-one vote standard. However, the District Court reasoned that the electoral college sanctions a "unit voting" scheme and thus the Republicans were entitled to use that scheme to measure their strength. The District Court buttressed this reasoning by reference to an asserted purpose of the victory bonus to strengthen state party organization. On the other hand, the District Court held that the use of the "uniform" bonus did not forward these purposes in regard to the large states, thus was not necessary to those purposes and, therefore, implicitly concluded that the uniform bonus was a deviation from the one person-one vote standard which was not justified.

Assuming for the moment the validity of the District Court's conclusion that the victory bonus concept was permissible, we have little doubt that its conclusion in regard to the uniform bonus is correct. It is not necessary to conceptualize the party interest as "compelling" or the one person-one vote standard as "fundamental" to conclude that any legitimately justified deviations from the one person-one vote standard must be reasonably tailored to the justification offered. The Supreme Court so held in White v. Weiser, 412 U.S. 783, 792, 93 S.Ct. 2348, 37 L.Ed.2d 335 (1973). Such a principle is merely a reflection of the increasing tendency to eschew reliance on fictional asserted justifications or purposes in favor of a more searching inquiry into the true purpose as that purpose may reasonably be read from the effect of the classification attacked. If the purpose of the victory bonus is as the District Judge found, there was no further justification of the deviations than that necessary to measure party strength in the electoral college. Since large states have more electoral college votes than small states and since the record indicates that the uniform bonus produced greater deviations than a purely proportional victory bonus, the District Court was clearly correct in concluding that the uniform bonus was impermissible in light of its asserted justification.

5. However, we cannot conclude that the District Court was correct in its initial determination that the victory bonus was a permissible concept. We first note that it is not transparently obvious that the purpose of the victory bonus is to measure Republican strength in terms of state victory. First, there is the fact that 11% of the delegates are apportioned on the basis of a uniform victory bonus, a bonus system which is plainly not necessary to and is perhaps inconsistent with an ascertainment of party strength by state voting power in the electoral college. Second, party strength in the electoral college is measured on the basis of only one election — the most recent presidential election for virtually all of the bonus delegates. When the Republicans have an unusually good or an unusually bad presidential election year, e. g. 1964 and 1972, party strength in the electoral college is not correctly measured by the most recent presidential election. Furthermore, the brief for the Republican National Committee and the National Republican Party foregoes any reliance on a party strength justification such as was conceptualized by the District Court. We could, therefore, simply assume for purposes of decision the validity of a victory bonus geared to ascertainment of party electoral strength and rule only on the justification offered by the defendants for the bonus. We are, however, hesitant to do this since the two justifications — the party strength justification and the justification offered by defendants — are closely related and should be held invalid for similar reasons.

6. The purpose of the victory bonus, both uniform and proportional, as set forth by the defendants is not based on an approximation of party strength but rather on ideological considerations. As defendants state in their brief:

Political treatment of states on a state-to-state basis, without regard to their wealth, size or population (or alleged party strength) is practically an American tradition. Each state, no matter its size, is represented by two United States Senators, each state is represented by its own distinguishable delegation within the Electoral College and each state is granted only one vote in the United States House of Representatives in the event of a failure by the Electoral College to vote a majority for any single presidential candidate. In a word, each state is sovereign, requiring even-handed and equal treatment in the Federal System. Defendants submit that the recognition of these facts, coupled with the element that identical achievements [in a state party receiving enough Republican votes to win the electoral votes of its state] deserve identical rewards, constitutionally justifies the inclusion as a part of a presidential bonus formula [a uniform bonus].

This argument is buttressed by the assertion that such recognition of individual states and their "achievements" is a non-justiciable accommodation of competing ideological interests. Implicitly the Republican Party claims the authority to create for its party an analogue of the Connecticut Compromise whereby smaller Republican states are guaranteed "a say" in Party affairs and large states are given proportionately more of the amount of delegates to be apportioned on the basis of population. A complete evaluation of these contentions requires a consideration of the nature of the one person-one vote principle and its relation to a major political party.

IV. THE INVALIDITY OF TERRITORIAL DISCRIMINATIONS FOR IDEOLOGICAL REASONS

The argument advanced by the defendants is exactly that which was rejected in Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964) and Lucas v. Forty-Fourth General Assembly of Colorado, 377 U.S. 713, 84 S.Ct. 1459, 12 L.Ed.2d 632 (1964). Indeed, defendants' argument virtually tracks the dissents of Justice Harlan in Reynolds and Justice Stewart in Lucas. The essence of this argument is that deviations from the one person-one vote standard are permissible since they represent a political balance of economic, social, historical and ideological factors. Such a balance is obviously not for the Court, as everyone must agree. The process of apportionment so viewed is simply an extension of election victory: since the people elect persons representing definable economic, social, historical and ideological positions, they may similarly elect to give a disproportionate (in terms of a hypothetical one person-one vote standard) weight through an increased number of political representatives to the victorious positions. This view of the defendants' argument is perfectly illustrated by Lucas which involved an attempt to malapportion the Colorado legislature through a popular referendum.

The Court understood this argument when it decided Reynolds and it rejected it. It rejected it because voting rights are "preservative of other basic civil and political rights." and because, in language which is now immortal:

Legislators represent people not trees or acres. Legislators are elected by voters, not farms or cities or economic interests. As long as ours is a representative form of government . . ., the right to elect legislators in a free and unimpaired fashion is a bedrock of our political system. It could hardly be gainsaid that a constitutional claim had been asserted by an allegation that certain otherwise qualified voters had been entirely prohibited from voting . . . . And, if a State should provide that the votes of citizens in one part of the State should be given two times, or five times, or 10 times the weight of votes of citizens in another part of the State, it could hardly be contended that the right to vote of those residing in the disfavored areas had not been effectively diluted. It would appear extraordinary to suggest that a State could be constitutionally permitted to enact a law providing that certain of the State's voters could vote two, five, or 10 times . . . while voters living elsewhere could vote only once.

The center of this reasoning is that the right to an equal vote is the starting place for battle over the proper weight to be given economic, social, historical and ideological interests in the legislative process; the right to vote is not merely another arena in which those various interests may assert their power. The right to an equal vote serves to prevent an entrenchment of any one group of interests to the exclusion of others, even if freely chosen in the most democratic fashion because such an entrenchment in the very process of political choice is contrary to the democratic ideal. In each election, warring interest groups must theoretically re-contest the balance struck at the last election. In this manner, democratic change is permitted if such change is desired. This concept was easily recognized in the first reapportionment cases which presented the most horrendous examples of entrenchment of certain interest groups. As Justice Clark forcefully stated in regard to the Tennessee malapportionment: "[The] legislative policy has riveted the present seats in the Assembly to their respective constituencies, and by the votes of their incumbents a reapportionment of any kind is prevented. The people have been rebuffed at the hands of the Assembly. . . ." The Supreme Court did not think even a democratic choice in the present should be permitted to begin such an entrenchment and so held in Lucas.

But the policy against entrenchment of political interests of whatever shape or form, liberal or conservative, desirable or not from any point of view, does not extend to the results of an election, nor to the type of district in which a representative is to stand, nor, indeed, to the amount of votes needed to carry a particular measure. The policy against entrenchment operates solely to insure "full and effective participation by all citizens"; it does not operate to control what ideological shading or interest group balance that citizen participation produces. The one person-one vote standard is simply the starting line which all interest groups must toe. So viewed, the one person-one vote standard is implicit in the concept of a democracy, as necessary to representative government as the concept that individuals may not be disenfranchised because of their social or economic views and as such a constitutional principle of the highest order.

With this understanding of the one person-one vote standard, the invalidity of a victory bonus scheme either as an approximation of party strength or as an ideological commitment to state sovereignty becomes apparent. In either case, the victory bonus operates to entrench the prevailing powers of the Republican Party through a territorial discrimination that weights the votes of some actual and potential Republicans more than others. As such, the victory bonus is, as its name suggests, directed to the past; it does not provide a starting line for the future. Of course, the Republican Party, equitably apportioned, may choose whatever balance of social, economic, historical or ideological interests it thinks fit to adopt. Nothing in the one person-one vote principle prevents it from doing so. But that choice must be one made by all potential and actual Republicans and not those which are given additional strength solely because of a balance struck in the past.

It remains to be seen whether any reasons exist for exempting a national and major political party from this principle. We turn first to the most insistent argument the defendants present — that this principle infringes upon the ideological choice of a political association in contravention of the First Amendment and in opposition to the intimation by the Supreme Court in O'Brien and the holding of Cousins v. Wigoda. However, we hold above that the apportionment of delegates is not a subject mete for control by major party ideology. For reasons expressed on page ___ of 173 U.S.App.D.C., page 564 of 525 F.2d infra, this holding is consistent with the First Amendment. The ideological coloration of the delegates once equitably apportioned, as determined by the voters or by the party, is a distinct issue. This is clearly recognized in the reapportionment cases and inherent in their principle as that principle has been elucidated above. O'Brien and Cousins dealt with the character of the delegates and thus dealt with "internal party affairs" in a manner clearly not contemplated by our reasoning in this case. Cousins was expressly limited to cases where "the convention itself [was] the proper forum for determining intra-party disputes as to which delegates [should] be seated." We conclude that the District Court correctly distinguished O'Brien.

A second concern is whether our reasoning is consistent with Bode and Georgia to the extent those cases held that a party may attempt to approximate party strength in apportionment of delegates. The argument is that since we approved apportionment schemes based in part on past performance, we implicitly approved recognition of past success in the apportionment of delegates. We disagree. Both Georgia and Bode hold only that the relevant constituency upon which the one person-one vote standard is to be applied at a party convention is not the total population. More important, Bode recognizes that constituency can be measured by counting potential as well as past Democratic voters. The project in determining party strength is thus not based on past "success" in winning votes by a particular political program alone but rather is an attempt to loosely define the probable constituency for the future, an attempt which requires some consideration of past voting patterns. The difference between this sort of attempt and a victory bonus scheme is obvious enough: the victory bonus excludes consideration of that part of the Republican constituency which resides in a state which did not vote Republican in the Electoral College. The only relevant constituency under the victory bonus is not the potential Republican votes that may exist but only the past Republican votes in states which voted Republican in the Electoral College.

A further argument is that the Electoral College system, by recognizing a unit voting scheme, and Article II of the Constitution, by adopting the Connecticut Compromise, explicitly permit a malapportionment based on recognition of state sovereignty and on recognition that only Electoral College votes are the "constituency" of a national political party. We held in Bode and reaffirm here that the Electoral College principle does permit a malapportion of delegates in regard to either total population or party vote. However, it is quite a step beyond this to calculate the Republican constituency on the basis of Electoral College votes. We think in this day and age the Republican constituency consists of Republican voters and not Republican Electors. The modern adoption of popular suffrage requires no less. Calculation of party strength, both actual and potential, thus must be directed toward individual voters.

As to the use of the analogue to the Senate to permit malapportionment of delegates to a political party, we do not think a political party can be assimilated to a bicameral legislature. The party acts to elect a President, not the Senate, and in the actual election under the Electoral College system the only deviation from one person-one vote is the uniform grant of two Electoral votes representing the state's Senators. That marginal increase is permitted by Bode. We do not think any greater deviation can be permitted simply by analogy to the United States Senate. Whether the principle of giving smaller states a sufficient "say" in the convention may permit a relatively minor deviation from the one person-one vote standard, such as was permitted in Mahan v. Howell, 410 U.S. 315, 93 S.Ct. 979, 35 L.Ed.2d 320 (1973), is a question we need not decide on the record in this case.

A final argument for distinguishing a major, national political party for purposes of applying the one person-one vote standard is that a political party is not exclusive. The gist of this argument is that if a citizen is denied equal access to the Republican decision-making process through malapportionment, he or she may simply join another party or vote for another party: that is, as long as there is equal access at the general election, the one person-one vote standard is fulfilled and we need not be concerned with the operations of the party. This argument would have great force if there were a truly free marketplace in political parties at the general election. However, such is not the case. As discussed in regard to "state action", the narrowing process utilized by the major political parties effectively limits the range of choice at and, indeed, the importance of the general election. "The convention [of major political parties] serves the pervasive national interest in the selection of candidates for national office. . . . The paramount necessity for effective performance of the Convention's task is underscored by Mr. Justice Pitney's admonition `that the likelihood of a candidate succeeding in an election without a party nomination is practically negligible. . . . As a practical matter, the ultimate choice of the mass of voters, is predetermined when the nominations have been made.'" For better or worse, the major political parties are an integral part of the election process and if the one person-one vote standard is to apply at the general election, it must also apply to the primary process to be truly effective.

We might conclude our discussion by further reference to the concept of justiciability. We have held above that courts may intervene into the affairs of major, national political parties to insure that delegates to their conventions are apportioned fairly, on the basis of the one person-one vote standard as modified by Bode and Georgia. This limited judicial intervention is no less manageable than judicial intervention into apportionment at the general election. Since the holding is that the principle underlying the one person-one vote standard is not a subject mete for party ideology, there is no judicial involvement in party ideology. This judicial intervention protects only the process of political choice not the results. Whether the process has any effect on the results is an interesting but to us irrelevant question. We would reach a similar result even if the one person-one vote standard changed nothing.

This limited intervention to insure equal participation is consonant with the institutional structure of the First Amendment and, indeed, forwards its underlying purposes. The right to an equal vote, the principle enunciated in the reapportionment decisions and which we apply here, is itself rooted in the First Amendment or, put another way, the First Amendment is designed to make the right to vote effective and to permit its intelligent exercise. It can hardly be deemed a rational development of First Amendment doctrine to dilute the vote in the name of the freedom whose exercise is designed to make the vote more effective, at least when the entity which attempts to dilute the vote is colored with state action. One may conceptualize the matter as a reconciliation of the First Amendment with other sections of the Constitution which establish a democratic form of government, as Professor Meicklejohn appears to do, or determine the principle of the reapportionment decisions to be a "compelling" national interest which overrides the freedom of association of national political parties. In either case, the First Amendment as it has developed in the numerous decisions protecting the rights of political associations does not prevent vindication of the commanding constitutional principle of the reapportionment decisions. We, therefore, have no difficulty concluding that this judicial intervention is fully consistent with concepts of justiciability. Whether a different result would obtain if a different type of intervention were attempted is beyond the scope of this case.

V. REMEDY AND ATTORNEYS FEES

Plaintiffs present us with two alternative plans for the apportionment of the 1976 Republican Convention delegates. They request that this Court either order the defendants to implement one of these plans or order that the defendants take timely action to adopt a revised formula and submit the revised formula to the District Court for approval before effectuation. We first note that "primary jurisdiction" over reapportionment lies with the body being reapportioned. The policies underlying the doctrine of justiciability and state action require this result. Since the defendants proceeded in good faith to obtain a judicial determination of the validity of their current reapportionment plan and have not had an opportunity to comply with an adverse judgment, we must therefore refrain from implementing any plan suggested by plaintiffs until the defendants have had a reasonable opportunity to comply with our mandate. We similarly do not think that implementation of this plan should be conditioned on approval by the District Court. If the defendants fail to comply with the mandate of this Court, they will, of course, be subject to further proceedings on the basis of our judgment. If the defendants implement a new plan not controlled by our judgment here but which plaintiffs consider raises new constitutional issues, the plaintiffs may file an appropriate complaint in the District Court and preliminary relief may be granted if permissible under the applicable precedents. We have not had sufficient experience with the difficult issues in this area of law to order that the defendants comply with the standards approved in Bode and Georgia. We think the defendants should have another opportunity to consider the constitutionality of their apportionment plan. As to the request for timely reconsideration, the defendants have not shown any propensity for delay in compliance with the mandate of the District Court and we would not infer such a course of action. If revision of the formula for the 1976 convention is not forthcoming within a reasonable time after entry of judgment, plaintiffs may move the District Court for appropriate relief.

The District Court declared that the use of uniform victory bonuses was improper and enjoined their use in the apportionment of delegates to the 1976 convention. We affirm that order and further declare that use of proportional victory bonuses is illegal and remand the cause with instructions to enjoin the use of such victory bonuses in the apportionment of delegates to the 1976 Republican Convention.

Plaintiffs moved the District Court for allowance of attorneys fees and certain expenses, a motion which the District Court denied without opinion. Plaintiffs assert three grounds in support of their argument for allowance of fees: (1) that the defendants have acted in bad faith; (2) that the plaintiffs' suit confers a benefit on an ascertainable class and the award of fees will serve to spread the costs of litigation among the beneficiaries; and (3) that the plaintiffs have acted as private attorneys-general and vindicated a policy which Congress considered of the highest priority. We do not think the first two grounds justify an award of fees. The defendants have not acted in bad faith in refusing to accept the District Court decision as final and instead insisting on their right to appellate review. Defendants are not correct in asserting that the plan adopted at the 1972 Republican Convention was a good faith effort to comply with the District Court's first order, but they were under no duty to comply with that order since it was stayed. On proceedings in regard to the second complaint, defendants were entitled to maintain that the District Court was mistaken in its view of the uniform bonus and thereby perfect the issue for appeal. Absent Supreme Court review of this decision, further refusal to comply with the judgment of the District Court or refusal to comply with our judgment here may be viewed in a different light. As to the second asserted ground, it is not applicable where imposition of attorneys fees on the defendants will not spread the costs of litigation proportionately among the beneficiaries.

As to the third asserted ground, we have no doubt that the one person-one vote standard is, through operation of 42 U.S.C. § 1983 (1970), a Congressional policy of the highest priority. There is a good deal of authority for the proposition that this fact alone justifies the award of attorneys fees. However, we agree with the Fourth Circuit that the logical extension of these holdings is that virtually all constitutional claims entitle the proponent to attorneys fees. We are unsure what, if any, limitations should be placed on these holdings. We are satisfied, however, that the beginning point for analysis is our recent decision in Wilderness Society v. Morton, 161 U.S.App.D.C. 446, 495 F.2d 1026, cert. granted sub nom., Alyeska Pipeline Service Company v. Wilderness Society, 419 U.S. 823, 95 S.Ct. 39, 42 L.Ed.2d 47 (1974). Since the District Court did not have the benefit of this opinion when it ruled on plaintiffs' motion and since it must exercise its equitable discretion in deciding upon the amount of fees, if any, to be granted, we think the better course is to remand this issue to the District Court for further consideration.

Affirmed in Part, Reversed in Part And Remanded With Instructions.

Senior Circuit Judge DANAHER dissented and filed an opinion with the Clerk.

BAZELON, Chief Judge:

ORDER

PER CURIAM:

It is ordered by the Court, en banc, sua sponte, that the opinions and judgment filed herein this date are hereby vacated, and it is

Further ordered by the Court, en banc, sua sponte, that these cases shall be reheard by the Court sitting en banc.