This suit against Indiana’s election officials, brought by the Communist Party, U.S.A., some of its candidates, and some people who would like to vote for its candidates, seeks to enjoin as a violation of the First Amendment an Indiana law that requires minor parties wanting to be listed on the ballot for national or statewide office to submit petitions equal to 2 percent of the number of people who voted in the last election for secretary of state. Ind.Code, § 3-1-11-1. The district court dismissed the suit after a hearing and the Communist Party (as we shall for brevity denote all the plaintiffs) has appealed. Although brought before the 1984 election, which is now some months in the past, the suit is not moot, since the Communist Party hopes to field candidates in future presidential elections. See, e.g., Moore v. Ogilvie, 394 U.S. 814, 816, 89 S.Ct. 1493, 1494, 23 L.Ed.2d 1 (1969); Gjertsen v. Board of Election Comm’rs, 751 F.2d 199, 202 (7th Cir.1984).

Before 1980 the required percentage for nomination by the petition route was only .5 percent. On March 3 of that year the legislature increased the percentage to 2 percent, effective January 1, 1983. Eight presidential candidates, including the Communist Party’s, appeared on the ballot in the November 1980 election, for which the old requirement of .5 percent was still in force. That was the most presidential candidates to have appeared on the Indiana ballot since 1932, when there had been six; since then there had never been more than five, until 1980. In 1984, however, with the new law in effect, there were only four presidential candidates on the Indiana ballot; and there might have been only two except for a since-dispelled ambiguity in the law which allowed two small parties that had polled .5 percent of the vote for secretary of state in the 1982 election, but much less than 2 percent, to get on the ballot. The Communist Party could not use this route (since closed) in 1984 because it had not put up a candidate for secretary of state in 1982. The Party submitted some 9,000 verified signatures, which fell far short of the 35,000 required but would just have made it under the old .5 percent rule. Since Indiana does not permit write-in votes, the Communist Party was completely excluded from the ballot. Incidentally, only very small parties follow the petition route to get on the ballot. A party that polls more than 2 percent but fewer than 10 percent of the votes cast for secretary of state is required to hold a convention to nominate its candidates, while a party that polls 10 percent or more must hold a primary election.

Although, with immaterial exceptions such as the Fifteenth Amendment’s prohibition of abridging the right to vote on racial grounds, the federal Constitution does not explicitly create a right to vote, or to stand for office, or to associate in a political party (political parties were not foreseen when the original Constitution,[*1173] and the Bill of Rights, were adopted), the Supreme Court in recent years has found such rights to be implicit in various constitutional provisions. Of particular relevance here, the Court has allowed minor parties to challenge under the equal protection clause of the Fourteenth Amendment, or (as here) under the First Amendment (which has of course been held applicable to state action by virtue of the Fourteenth Amendment), provisions of state law that make it harder to get the candidates of such parties on the ballot. See, e.g., Illinois State Bd. of Elections v. Socialist Workers Party, 440 U.S. 173, 184, 99 S.Ct. 983, 990, 59 L.Ed.2d 230 (1979). But the Court has not settled on the standard to be applied in ruling on such challenges. Some cases suggest a strict standard: the state must use “the least drastic means” of restricting access to the ballot in pursuit of legitimate state interests — must avoid “overbroad restrictions.” Id. at 185-86, 99 S.Ct. at 990-91; see also Williams v. Rhodes, 393 U.S. 23, 31, 89 S.Ct. 5, 10, 21 L.Ed.2d 24 (1968). Other cases suggest a looser standard: “the State’s important regulatory interests are generally sufficient to justify reasonable, nondiscriminatory restrictions.” Anderson v. Celebrezze, 460 U.S. 780, 788, 103 S.Ct. 1564, 1569-70, 75 L.Ed.2d 547 (1983) (footnote omitted); see also id. at 788 n. 9, 103 S.Ct. at 1570 n. 9, and cases cited there. The uncertainty about the standard for judging such restrictions has been noted by the Fifth Circuit in Dart v. Brown, 717 F.2d 1491,1501-02 (5th Cir.1983), and by several commentators. See, e.g., Tribe, American Constitutional Law 783 (1978); Case Comment, 18 Suffolk U.L.Rev. 24, 30 (1984).

If a strict standard were applied to this case, Indiana would be in trouble. While it certainly has an interest in limiting the length of the ballot in order to prevent confusion of voters, it has not shown that the protection of this interest required it to raise the petition requirement from .5 to 2 percent. If the 1984 election can be considered representative in this regard — and only time will tell — the new law will eliminate all minor parties from the presidential ballot. Maybe eight presidential candidates are too many, though Justice Harlan, concurring in Williams v. Rhodes, supra, 393 U.S. at 47, 89 S.Ct. at 19, thought not; and maybe the legislators foresaw when they raised the percentage in March 1980 that there would be eight candidates in the fall election. Still, to go from .5 percent all the way to 2 percent in one jump was abrupt, and seems to have brought the number of candidates well below the level that had seemed tolerable for at least 40 years. Since Indiana does not permit write-ins, the ballot as a method of expressing political views may be completely closed to minor parties in Indiana. Compare Jenness v. Fortson, 403 U.S. 431, 438, 91 S.Ct. 1970, 1974, 29 L.Ed.2d 554 (1971).

The consequences seem difficult to square with a test that requires that the state use the “least drastic means” possible to regulate access to the ballot. But unfortunately for the Communist Party, there is abundant judicial authority, much in the Supreme Court itself and therefore beyond our power to reexamine, for allowing states to set even higher minimum percentages than Indiana has done. See Storer v. Brown, 415 U.S. 724, 738, 746, 94 S.Ct. 1274, 1283, 1286, 39 L.Ed.2d 714 (1974) (5 percent); Jenness v. Fortson, supra (same); Populist Party v. Herschler, 746 F.2d 656, 660 (10th Cir.1984) (per curiam) (same); Libertarian Party v. Florida, 710 F.2d 790 (11th Cir.1983) (3 percent); Arutonoff v. Oklahoma State Election Bd., 687 F.2d 1375 (10th Cir.1982) (5 percent); Beller v. Kirk, 328 F.Supp. 485 (S.D.Fla.1970) (three-judge panel) (per curiam), aff’d without opinion under the name of Beller v. Askew, 403 U.S. 925, 91 S.Ct. 2248, 29 L.Ed.2d 705 (1971) (3 percent); Wood v. Putterman, 316 F.Supp. 646 (D.Md.) (three-judge panel), aff’d without opinion, 400 U.S. 859, 91 S.Ct. 104, 27 L.Ed.2d 99 (1970) (3 percent); Populist Party v. Orr, 595 F.Supp. 760 (S.D.Ind.1984) (upholding Indiana’s 2 percent requirement). In many of these cases, moreover, the base to which the percentage was applied was larger than it is here. In Jenness, Libertarian Party [*1174] v. Florida, and Wood, the base was all registered voters. Since many registered voters do not actually vote, this is a larger base than actual voters in a particular election, as here. In Arutunoff the base was the vote for President or (in off-year elections) for governor; here it is the vote for the secretary of state, a more obscure officer chosen in an off-year election. If we compare the 35,000 signatures required by the Indiana law with the 2.2 million votes cast in Indiana in the 1980 presidential election, see U.S. Dept, of Commerce, Bureau of the Census, Statistical Abstract of the United States 1985, at 240 (tab. 403), we see that the Indiana law required signatures equal to only 1.5 percent of the number of votes for President, which is less than a third of the requirement upheld in Arutunoff.

Granted, numbers aren’t everything. In McLain v. Meier, 637 F.2d 1159, 1163-64 (8th Cir.1980), a requirement of only 15,000 signatures, fewer than in this case though equal to 3.3 percent of the electorate, was struck down — but in conjunction with other restrictions. The regulatory scheme must be considered in its entirety. Thus Jenness, where a 5 percent requirement was upheld, can be distinguished from this case by reference to the existence of a write-in alternative, a factor stressed by the Court, see 403 U.S. at 438, 91 S.Ct. at 1974, and absent here. But in many of the cases we have cited, the difficulty for the minor party of getting on the ballot was greater than in this case, once the practical difficulties of the write-in route are recognized. In Libertarian Party v. Florida, for example, which involved a more populous state than Indiana and a signatures requirement computed in terms of registered voters, the state’s 3 percent requirement meant that a party had to collect more than 140,000 signatures to get on the ballot, which is four times the number required by Indiana. Even the 1 percent requirement upheld in American Party of Texas v. White, 415 U.S. 767, 783, 94 S.Ct. 1296, 1307, 39 L.Ed.2d 744 (1974), entailed two-thirds, rather than one-half, as many signatures as Indiana’s 2 percent requirement (22,000 versus 35,000). Of course in one sense the more populous the state, the easier it is to get signatures— there is a larger pool to fish from. On this theory only the percentage should matter, a proposition for which McLain, which invalidated a requirement of only 15,000 signatures, offers some support. But this overlooks a number of things: it costs money to circulate petitions; the more signatures that are required, the higher the cost is; and minor parties usually are strapped for funds.

The courts may sometimes talk the language of least drastic means but they only strike down ballot-access regulations that are unreasonable, such as Ohio’s 15 percent requirement in Williams v. Rhodes. Of course the existence of a less restrictive alternative may be relevant to an assessment of reasonableness; one way in which a requirement may be unreasonable is that it is unnecessary in light of another requirement that could be imposed instead. In any event, it seems apparent that the courts do not actually subject ballot-access regulations to the same stringent review to which they subject racial discrimination and regulations of the content of political communications, to give just two examples of strict review. And while it might be possible to reconcile the divergent formulations in the Supreme Court’s ballot-access cases on a purely verbal level by drawing an analogy to the semi-strict review that the Court uses in cases involving sexual discrimination alleged to violate the Fourteenth ' Amendment’s equal protection clause, see, e.g., Kirchberg v. Feenstra, 450 U.S. 455, 459, 101 S.Ct. 1195, 1198, 67 L.Ed.2d 428 (1981), that would suggest a stricter standard than the Court actually applies. See Dart v. Brown, supra, 717 F.2d at 1498-1504, an exhaustive examination of the issue that reaches the same conclusion as we.

We must follow what the Supreme Court does, and not just what it says, especially when what it says is not altogether consistent from case to case; and while as an original matter a 2 percent requirement in[*1175] the circumstances of this case might be thought an undue restriction on minor parties’ access to the ballot if the test is “the least drastic means,” the lawfulness of such a restriction follows a fortiori from the decisions upholding higher requirements. In general the danger zone does not even begin until a state goes above 2 percent; and while under the precedents a 2 percent requirement, or in special cases an even smaller requirement, see Libertarian Party v. Beermann, 598 F.Supp. 57, 60 (D.Neb.1984), could be struck down if combined with other restrictions — and some states are ingenious in devising such restrictions, see Blackman, Third Party President? An Analysis of State Election Laws, ch. 4 (1976) — we do not think that barring write-in votes, the only other restriction in Indiana’s law that the Communist Party has cited us to, can put the plaintiff over the hump; as a practical matter it is a trivial restriction. Professor Tribe has suggested that the danger zone does not begin till the requirement goes above 5 percent. Tribe, supra, at 784. The percentage here is much lower, and is, as we have stressed, applied to votes in an off-year election for a relatively minor post, that of the secretary of state, and hence yields only a modest number of required signatures.

The Communist Party points out, however, that the history of Indiana elections provides a kind of laboratory, unavailable in other states, in which to assess the need for the 2 percent minimum. In states that have always had a 2 or a 3 or a 5 percent requirement it would be speculative to determine the effects of reducing it. A party that knows it can’t get 2 percent may not try to get any signatures; hence it will be difficult to say whether, if the 2 percent requirement were relaxed, the number of candidates would increase — and if so to impossible numbers. But we know that in Indiana a much lower requirement, a .5 percent requirement, kept the number of candidates down to eight — indeed, usually to five — and that raising the requirement to 2 percent may have cut it down to two. This is not certain, though. The two small parties that qualified because they got- .5 percent in the secretary of state election in 1982 might have been able to collect enough signatures to meet the 2 percent requirement if that had been necessary. In any event, if Indiana is to be penalized for having formerly had a lower requirement, an arbitrary difference will be created among the states. States that had always had high requirements could retain them; states like Indiana that had experimented with lower ones might, depending on the results of the experiments, be frozen into them. Although Libertarian Party v. Fainter, 741 F.2d 728, 730 (5th Cir.1984) (per curiam), gives a bit of support to this position, Arutunoff v. Oklahoma State Election Bd,., supra, 687 F.2d at 1380, decisively, and we think correctly, rejects it; any other result would create an unacceptable constitutional ratchet.

Weighing the burdens on minor parties and their supporters, actual and potential, along with the benefits of restricting access to the ballot, see Dart v. Brown, supra, 717 F.2d at 1504, we conclude that Indiana’s 2 percent requirement does not violate the Constitution. The 1980 campaign year, which of course began well before March, was marked by great ferment that culminated later in the year in John Anderson’s campaign for President as an independent. The legislators could well have foreseen a crowded ballot, and indeed the November ballot was crowded; there were eight presidential candidates on the ballot — an unusual number. The problem of voter confusion cannot be dismissed as trivial, given the advantage that being listed first on the ballot gives a candidate. See Sangmeister v. Woodard, 565 F.2d 460 (7th Cir.1977); Comment, Protecting the Rationality of Electoral Outcomes: A Challenge to First Amendment Doctrine, 51 U.Chi.L.Rev. 892, 900-02 (1984). No doubt most voters make up their minds whom they are going to vote for before they get into the voting booth, but some of them either do not, or become confused when they get into the booth; the longer the ballot, the greater the confusion is like[*1176] ly to be. And even if we thought the problem a rather minor one, we would have to bow to the contrary view of the Supreme Court. See, e.g., Anderson v. Celebrezze, supra, 460 U.S. at 788 n. 9, 103 S.Ct. at 1570 n. 9; Libertarian Party v. White, supra, 415 U.S. at 782 n. 14, 94 S.Ct. at 1307 n. 14.

The increase in the required percentage left Indiana with a lower requirement than many other states have, especially when regard is had for the base to which the required percentage is applied, as well as the percentage itself. While no parties met the requirement in 1984, the two small parties that got on the ballot by virtue of their 1982 vote might have been able to get the required number of signatures if necessary. They, and the Communist Party if it tries harder, may well be able to qualify for the 1988 ballot. Thirty-five thousand signatures in a state of millions of registered voters are not a lot to get; and a party that cannot get that many signatures is not likely to make much of an impression on the electorate even if it gets on the ballot. This is especially so since Indiana does not impose such onerous requirements as that the petitions be distributed geographically in a specified way rather than being concentrated in the areas where the party has its strongest support (such a .“distribution” requirement was struck down in the Beermann case, supra), or that the petitions contain detailed information about the signers, or that all the signatures must be collected long before the election (as in McLain) or within a brief period of time, or the other burdening devices discussed in the Blackman monograph; certainly the Communist Party has pointed to none. Apart from requiring that signers be registered voters and that their signatures be attested — which seem perfectly appropriate requirements — the state imposes only the 2 percent restriction; and it is a lesser restriction than many other states impose, and less than the percentages that the courts regularly uphold.

It is quite possible that the Indiana law ratifies rather than restricts the amount of political competition. It may keep off the ballot just those parties so lacking in electoral appeal that they contribute little to the political diversity of the state or nation. Cf. Comment, Legal Obstacles to Minority Party Success, 57 Yale L.J. 1276, 1286 (1948). John Anderson, leading a brand-new political movement, was able to get on every state’s ballot in 1980, though it has been argued that the effort to do so debilitated his campaign. See Frampton, Challenging Restrictive Ballot Access Laws on Behalf of the Independent Candidate, 10 Rev.Law & Social Change 131 (1980). The Communist Party is an established party, but one with a proven lack of political appeal in the United States, especially in conservative states such as Indiana. It appeared on only 22 states’ ballots (and that of the District of Columbia) in 1984.

Finally, there is no evidence that the Indiana legislature acted with an invidious intent in raising the minimum requirement from .5 to 2 percent — that it was hostile to minor parties in general or to the Communist Party or any other party in particular. The district court held an evidentiary hearing at which the Communist Party could have presented evidence of such an intent; it presented none. There is no published legislative history of Indiana laws but the Communist Party could have sought evidence of invidious intent elsewhere if it had thought that a promising route, cf. Wallace v. Jaffree, — U.S. -,-- -, 105 S.Ct. 2479, 2489-93, 86 L.Ed.2d 29 (1985), which evidently it did not. Granted, limitations on ballot access that are as stringent as found in many of the cases may be motivated less by concern with voter confusion than by the traditional American antipathy to the coalition governments that minor parties precipitate in systems of proportional representation. Although the Supreme Court has allowed states to base restrictions on access to the ballot for state elections on a fear of “unrestrained factionalism,” Storer v. Brown, supra, 415 U.S. at 736, 94 S.Ct. at 1282, it has questioned this rationale for state restrictions on the ballot for federal elections. See Anderson v. Celebrezze, supra, 460[*1177] U.S. at 804, 103 S.Ct. at 1577. No state’s election laws will measurably affect the danger of destabilizing the national political process through “unrestrained factionalism,” and presumably therefore the motivations for such laws lie elsewhere (though one might argue that if the effects are good, motivation should not matter). Moreover, it can be argued that, based on the historical record, minor parties pose little threat to the stability of our political system, and indeed seek access to the ballot not to win elections but to have a more prominent platform from which to preach ideas that may in time enter the political mainstream on the back of one of the major parties. But the overriding purpose of the ballot is electoral, and the states can limit access to the ballot in order to ensure that its electoral purpose is fulfilled. Georges v. Carney, 691 F.2d 297, 300 (7th Cir.1982).

Affirmed.