Opinion by Judge Norris; Dissent by Judge Wiggins.

This appeal requires us to decide the constitutionality of certain provisions of Proposition 73, a campaign financing reform measure for elections to state and local offices that was approved by California voters in 1988.1 Appellants are the California Fair Political Practices Commission (“FPPC”), which was named as a defendant in the action below, and the authors of Proposition 73, Assemblyman Ross Johnson and Senator Quentin Kopp (“Authors”), who intervened as parties defendant. Appellees are the Service Employees International Union, other labor organizations, elected officials, and individual campaign contributors, the plaintiffs that brought the action, and the California Democratic Party, which intervened as a party plaintiff.

Proposition 73 limits the amount individuals and groups may contribute to candidates for state and local office each fiscal year.2 Appellees do not, however, focus their constitutional attack on Proposition 73’s contribution limits per se. Instead they focus on the fact that Proposition 73 limits the amount a contributor may give during each fiscal year rather than follow*1315ing the federal model3 of limiting the amount a contributor may give during each election cycle. Appellees argue that in tying the contribution limits to fiscal years, Proposition 73 discriminates in favor of incumbents and their supporters and against challengers and their supporters for a very practical reason: Challengers do not typically decide to run for office years in advance of the election. As a result, they are unable to engage in fundraising during each fiscal year between elections as incumbents commonly do. Thus, for example, an incumbent state legislator may tap an individual contributor for $1,000 in each fiscal year of her four-year term, while a potential opponent cannot realistically do the same until the fiscal years in which the primary and general elections occur. As a result, appellees conclude, Proposition 73 effectively limits the contributions of an individual who chooses to support a non-incumbent to $1,000 in each of two fiscal years, while an individual who chooses to support an incumbent may contribute $1,000 in each of four fiscal years.

After a six-day bench trial, the district court agreed with appellees that limiting contributions on a fiscal year basis unconstitutionally discriminated against challengers. The district court found that incumbents “raise substantial amounts of money each of the years of incumbency, while as a general matter challengers cannot, and generally do not, do so.” Service Employees International Union v. Fair Political Practices Commission, 747 F.Supp. 580, 588 (E.D.Cal. 1990). Because the district court held that the fiscal year provisions of Proposition 73 were not severable from the contribution limitations themselves, it struck down the contribution limitations under the First and Fourteenth Amendments and permanently enjoined their enforcement. Id. at 593-94.

Appellees also challenged three provisions of Proposition 73 in addition to the fiscal year contribution limitations. First, appellees challenged Proposition 73’s carryover provision, which prohibits the expenditure of campaign funds raised prior to January 1989. On a partial motion for summary judgment, the district court held that this provision constituted an unconstitutional expenditure limitation and permanently enjoined its enforcement. Service Employees International Union v. Fair Political Practices Commission, 721 F.Supp. 1172 (E.D.Cal. 1989).

Appellees also challenged Proposition 73’s ban on intra-candidate transfers: transfers of funds between controlled committees of a single candidate. Cal.Gov’t Code § 85304. After trial, the court held that this provision was also an unconstitutional expenditure limitation. Service Employees International Union, 747 F.Supp. at 591. Finally, appellees challenged Proposition 73’s ban on inter-candidate transfers: transfers of funds between candidates. Cal.Gov’t Code § 85304. Analyzing this provision as a contribution limitation as opposed to an expenditure limitation, the district court ruled that the ban was unconstitutional because it served no purpose once Proposition 73’s contribution limitations had been declared unconstitutional. 747 F.Supp. at 591-93. The district court permanently enjoined enforcement of both the intra-candidate and the inter-candidate transfer bans. Id. at 593-94.4

*1316Appellants defend the constitutionality of every provision of Proposition 73, except that appellant FPPC does not appeal the district court’s invalidation of the carryover ban.

I

Before we consider the question whether limiting campaign contributions on a fiscal year basis unconstitutionally discriminates in favor of incumbents and against challengers, we must first address two threshold questions raised by appellants: (1) that appellees lack standing to raise the question of discrimination against challengers because none of them is a challenger,5 and (2) that the district court’s findings of fact on the issue of discrimination are clearly erroneous. See Fed.R.Civ.P. 52(a).

A

Appellants argue that appellees lack standing to question the constitutionality of a law that discriminates against challengers because no appellee is a challenger. However, we reject this argument because appellees have standing to assert their own rights as contributors. As the district court pointed out, 747 F.Supp. at 588, the Supreme Court held in Buckley v. Valeo that contributing money is an act of political association that is protected by the First Amendment because the act of contributing serves to associate the contributor with a candidate as well as with like-minded contributors. 424 U.S. 1, 22, 96 S.Ct. 612, 636, 46 L.Ed.2d 659 (1976) (per curiam). If Proposition 73 discriminates against challengers by limiting their opportunities to accept contributions, then it necessarily discriminates against contributors who wish to associate themselves with challengers. Appellees therefore have standing to assert their own associational rights and to challenge the fiscal year contribution limits as discriminatory. See Renne v. Geary, — U.S. —, 111 S.Ct. 2331, 2338, 115 L.Ed.2d 288 (1991) (“Respondents of course have standing to claim that [a law] has been applied in an unconstitutional manner to bar their own speech.”).

B

We now turn to appellants’ argument that the district court’s findings of fact on the issue of discrimination are clearly erroneous.6 The district court found that Proposition 73’s fiscal year contribution limits tend to discriminate against challengers as a class. Finding 155. While “[ijncumbents raise a significant amount of their campaign funds early in an election cycle,” challengers “raise very little money early in the election cycle. Most challengers do not decide to run until relatively late in the cycle because the prospects for success, which depend on national or state trends and information about the incumbent, cannot be assessed years in advance of the election.” Findings 128, 129. The court went on to find that this difference between challengers and incumbents leads to great disparities in fundraising during non-election years:

In state races in the off-years 1983, 1985, and 1987, all incumbents, but very few challengers, engaged in fundrais-ing_ In statewide constitutional office races in the off-years 1983 and 1985, incumbents outraised challengers by an average of almost 9 to 1. In state Senate races in the off-years 1983, 1985 and *13171987, incumbents outraised challengers by an average of more than 40 to 1. In State Assembly races in the off-years 1983, 1985 and 1987, incumbents out-raised challengers by an average of more than 70 to 1.

Finding 133.

We reject appellants’ contention that these findings are clearly erroneous.7 The district court’s findings are derived from the testimony of two expert witnesses, Professors Gary Jacobson and Bruce Cain. Jacobson’s testimony was based on his study of campaign finance and the fundraising patterns of incumbents and challengers in elections for congressional offices. See 2 Reporter’s Transcript (“RT”) at 271, 274. Cain’s testimony was based on his study of campaign finance in general, as well as Plaintiffs’ Trial Exhibit 6, an examination of contribution patterns in several California legislative and statewide elections. See 3 RT at 407, 412-13; Appellees’ Supplemental Joint Excerpts of Record (“ASJER”), Tab B.

Professor Jacobson testified that challengers “tend to raise their money later in the campaign,” 2 RT at 299-300, because “challengers often don’t decide until relatively late whether or not to make a race.” Id. at 301.8 Challengers wait, Jacobson testified, because the prospects for success cannot meaningfully be assessed two or three years before an election. Id.

Professor Cain prepared a report of fundraising in California based on FPPC annual reports that documented vast dis*1318parities between incumbents and challengers during the off years. ASJER, Tab B, at 6. This report supports Finding 133 as well as more general findings about fund-raising patterns. Like Jacobson, Cain testified that “fund raising in the off years is primarily an incumbent activity.” 3 RT at 413. Specifically, Cain noted that while few challengers “have even begun to fund raise in the off year,” incumbents had “raise[d] significant fractions of their money in the off year, and indeed raised about half of it in the first six months of the election cycle.” Id. He testified that few challengers even begin to think about running before the end of the off-year period, “let alone make the final decision and raise money.” Id. at 421-22. Thus, the testimony of Professors Jacobson and Cain provides ample evidentiary support for the district court’s crucial finding that Proposition 73’s fiscal year contribution limits discriminate against challengers as a class.

Appellants characterize the expert testimony as speculative because it was not based on elections conducted with Proposition 73’s contribution limitations in effect.9 Appellants’ argument is unpersuasive. The district court was not required to discount the testimony of Professors Cain and Jacobson merely because their data was not collected after Proposition 73 went into effect. Experts may make reasonable projections of future harm based on reliable data collected under similar circumstances and they need not wait to measure the actual discriminatory effects of a law for their testimony to have probative value. Indeed, in Buckley, the Court relied on data taken prior to the enactment of the federal contribution and expenditure limitations. See, e.g., Buckley, 424 U.S. at 21 n. 23, 96 S.Ct. at 636 n. 23.

C

Having determined that appellees have standing to challenge the fiscal year contribution limits and that the district court’s findings of fact on the issue of discrimination are not clearly erroneous, we now turn to the question whether viewpoint and content neutral contribution limits that discriminate against challengers and their supporters offend the Constitution. In Buckley, the Court upheld the constitutionality of federal contribution limits based on an election cycle. See supra note 3. The Court, however, refrained from deciding the constitutionality of contribution limits that discriminate against challengers because the record contained no evidence of “invidious discrimination against challengers as a class.” 424 U.S. at 31, 96 S.Ct. at 641.

However, the Court in Buckley did address the issue of discrimination in the context of expenditure limitations. The government asserted an interest in leveling the political playing field by limiting expenditures, which in practice had the effect of burdening expressive activities of the wealthy to a greater degree than those of the poor. In rejecting this argument, the Court said: “the concept that government may restrict the speech of some elements of our society in order to enhance the relative voice of others is wholly foreign to the First Amendment.” Id. at 48-49, 96 S.Ct. at 648-49 (1976).10 Thus, Buckley teaches *1319that government must remain scrupulously neutral when regulating expressive activities protected by the First Amendment. Moreover, a line of other Supreme Court cases from Police Dept. of Chicago v. Mosley, 408 U.S. 92, 92 S.Ct. 2286, 33 L.Ed.2d 212 (1972), to Austin v. Michigan Chamber of Commerce, 494 U.S. 652, 110 S.Ct. 1391, 108 L.Ed.2d 652 (1990), firmly establishes the principle that discrimination is permissible in the First Amendment context only when the discrimination is itself necessary to achieve a substantial governmental interest.11 Our court has reaffirmed this principle only recently. Harwin v. Goleta Water District, 953 F.2d 488 (9th Cir. 1991).

In Mosley, the Court struck down a Chicago ordinance that prohibited picketing on school grounds except for labor picketing. The Mosley Court recognized that Chicago could legitimately “prohibit[] some picketing to protect public order,” 408 U.S. at 98, 92 S.Ct. at 2291, and that “[cjities certainly have a substantial interest in stopping picketing which disrupts a school.” Id. at 99, 92 S.Ct. at 2292. Thus, the question in Mosley was not whether Chicago could burden expressive activity, but whether it could do so in a discriminatory manner. The answer, the Court held, depended on whether the discrimination itself was necessary to achieve the city’s purposes. “[T]he crucial question is whether there is an appropriate governmental interest suitably furthered by the differential treatment.” Id. at 95, 92 S.Ct. at 2290. The answer in Mosley was no. Chicago had not shown that non-labor picketing was “clearly more disruptive” than labor picketing. Id. at 100, 92 S.Ct. at 2292. The legitimate governmental interest that could support a general burden on expressive activity did not justify a selective burden on such activity.

In Carey v. Brown, 447 U.S. 455, 100 S.Ct. 2286, 65 L.Ed.2d 263 (1980), the Court relied upon Mosley in striking down an Illinois statute that prohibited the picketing of residences but exempted “the peaceful picketing of a place of employment involved in a labor dispute.” Id. at 457, 100 S.Ct. at 2288. While recognizing that “[t]he State’s interest in protecting the well-being, tranquility, and privacy of the home is certainly of the highest order in a free and civilized society,” id. at 471, 100 S.Ct. at 2296, the Court held that this interest would not support an exclusion for labor picketing. “[NJothing in the content-based labor-nonlabor distinction has any bearing whatsoever on privacy.” Id. at 465, 100 S.Ct. at 2293. Because the discrimination served no substantial govern*1320mental interest, the statute was unconstitutional.

While both Mosley and Carey involved content based discrimination, other cases establish that even viewpoint and content neutral statutes that discriminate violate the Constitution unless the discrimination is itself necessary to serve a legitimate government interest. In Minneapolis Star v. Minnesota Commissioner of Revenue, 460 U.S. 575, 103 S.Ct. 1365, 75 L.Ed.2d 295 (1983), the Court struck down a state law that taxed purchases of paper and ink in excess of $100,000 per year. Id. at 578 n. 2, 103 S.Ct. at 1368 n. 2. While the tax was facially neutral, in practice it fell on only a small number of newspapers. Id. at 578-79, 103 S.Ct. at 1368. This, the Court held, was impermissible. Id. at 591-92, 103 S.Ct. at 1375. Although the Court recognized that Minnesota’s interest in raising revenue was “critical,” id. at 586, 103 S.Ct. at 1372, that interest did not justify levying a tax on some newspapers but not others.12

While each of these cases speaks to the issue of discrimination in the First Amendment context, the case that is most directly on point is Austin v. Michigan Chamber of Commerce, 494 U.S. 652, 110 S.Ct. 1391, 108 L.Ed.2d 652 (1990). Austin requires that when a statute regulating political campaigns discriminates against a class of participants in the political process, the dis-' crimination must be independently justified, even where the statute is viewpoint and content neutral. The statute at issue in Austin prohibited corporate expenditures in elections for state office except expenditures made from segregated funds. Id. 110 S.Ct. at 1395. Thus, it burdened the First Amendment rights of corporations more heavily than the rights of others but was viewpoint and content neutral. The Court upheld the Michigan law, but only because the discriminatory treatment of corporations was itself necessary to serve “the compelling state interest of eliminating from the political process the corrosive effect of political ‘war chests’ amassed with the aid of the legal advantages given to corporations.” Id. 110 S.Ct. at 1401. Thus, Austin confirms that discrimination is permissible in the context of the First Amendment only if the discrimination itself is necessary to serve a substantial governmental interest.

In sum, Buckley, Mosley, Austin, and the other cases we have discussed establish that government must remain scrupulously neutral when it regulates activity protected by the First Amendment. The Court has not hesitated to strike down laws that are facially neutral but have a discriminatory impact on First Amendment rights. See Minneapolis Star, 460 U.S. at 591-92, 103 S.Ct. at 1375. The government may deviate from its neutral role only when it can satisfy a heightened standard of judicial scrutiny.13

*1321Appellants argue that Proposition 73 s contribution limits serve the compelling governmental interest in preventing corruption and the appearance of corruption. Although this interest will support some limits on political contributions, Buckley, 424 U.S. at 26-28, 96 S.Ct. at 638-39, appellants have made no showing that limiting contributions on a fiscal year basis advances this interest in any way. Thus, this case is much like Mosley. In Mosley, the Court recognized that Chicago had a legitimate interest in limiting picketing to prevent the disruption of schools but held that this governmental interest would not support a discriminatory ban on picketing. Here, we recognize that the state has a legitimate interest in preventing corruption and the appearance of corruption, but hold that this interest will not support a discriminatory formula for limiting contributions. In this case, as in Mosley, “the crucial question is whether there is an appropriate governmental interest suitably furthered by the differential treatment.” Mosley, 408 U.S. at 95, 92 S.Ct. at 2290. In both cases, the answer is no.

D

The district court held that the fiscal year feature of Proposition .73 is not sever-able from the contribution limits themselves. We agree. Appellants have given us no reason to believe that “the legislation would have been enacted if it had not in-eluded the unconstitutional provision[ ].” National Advertising Co. v. Town of Babylon, 900 F.2d 551, 557 (2d Cir.) (citing United States v. Jackson, 390 U.S. 570, 585 n. 27, 88 S.Ct. 1209, 1218 n. 27, 20 L.Ed.2d 138 (1968)), cert. denied, — U.S. -, 111 S.Ct. 146, 112 L.Ed.2d 112 (1990). As the district court pointed out, were we simply to strike the word “fiscal” from the statute, contribution limits would still be measured on an annual basis, raising the same problems of discrimination. 747 F.Supp. at 590. Were we to rewrite the statute to limit contributions on an election cycle basis, we would be at a loss to know what the dollar amounts of the limitations should be. In short, to save the statute, we would have to rewrite it substantially, “a practice that is decidedly disfavored.” Id. (citing Thornburgh v. American College of Obstetricians & Gynecologists, 476 U.S. 747, 764-65, 106 S.Ct. 2169, 2180-81, 90 L.Ed.2d 779 (1986)). Accordingly, we affirm the district court's decision that all of Proposition 73’s contribution limits that are measured on a fiscal year basis are constitutionally infirm.14

II

Having concluded that Proposition 73’s contribution limits based on a fiscal year are unconstitutional, we consider whether its ban on inter- and intra-candidate transfers likewise violate the First and Fourteenth Amendments.15 The district court’s *1322decision invalidating these provisions was not grounded on their discriminatory effect, and appellees have not argued on appeal that the transfer bans are discriminatory.16 We therefore shift from the discrimination analysis used in part I to a straightforward application of the tests for contribution and expenditure limitations enunciated in Buckley and its progeny.

A

We agree with the district court that the ban on intra-candidate transfers operates as an expenditure limitation because it limits the purposes for which money raised by a candidate may be spent.17 Expenditure limitations are subject to strict scrutiny and will be upheld only if they are “narrowly tailored to serve a compelling state interest.” Austin, 110 S.Ct. at 1396.

Appellant FPPC asserts that the ban is justified by the government’s interest in preventing funds from being raised for one office and spent for another. Even if we were to recognize this to be a compelling state interest, we would invalidate the ban as violative of the First Amendment because it is not narrowly-tailored. We agree with the district court that this interest in ensuring that contributors are not misled could be served simply by requiring candidates to inform contributors that their contributions might be spent on other races. 747 F.Supp. at 591. Concerns about the unintended use of contributors’ money can be met “by means far more narrowly tailored and less burdensome than [a] restriction on direct expenditures: simply requiring that contributors be informed that their money may be used for such a purpose.” FEC v. Massachusetts Citizens for Life, 479 U.S. 238, 261, 107 S.Ct. 616, 629-30, 93 L.Ed.2d 539 (1986). We hold, therefore, that the intra-candidate transfer ban fails the narrowly-tailored prong of the strict scrutiny test.18

B

We turn now to the inter-candidate transfer ban, which, in contrast to the intra-candidate ban, operates as a contribution limitation because it limits the amount one candidate may contribute to another. The Supreme Court has applied a somewhat less stringent test than strict scrutiny to decide the constitutionality of contribution limitations. Buckley, 424 U.S. at 20-21, 23, 96 S.Ct. at 635, 636; see also Massachusetts Citizens for Life, 479 U.S. at 259-60, 107 S.Ct. at 628-29 (“We have consistently held that restrictions on contributions require less compelling justification than restrictions on independent spending.”). However, the test is still a “rigorous” one. Buckley, 424 U.S. at 29, 96 S.Ct. at 640. Proposition 73’s ban on inter-candidate transfers may be sustained, therefore, only if the state “demonstrates a sufficiently important interest and employs means closely drawn to avoid unnecessary abridgment of associational freedoms.” Id. at 25, 96 S.Ct. at 638.

The FPPC asserts that the inter-candidate transfer ban is necessary to prevent contributors from circumventing the contribution limits by funneling contributions through one candidate to another. As the district court pointed out, 747 F.Supp. at 593, the inter-candidate transfer ban cannot serve this purpose in the absence of valid contribution limits.

*1323The Authors argue that the inter-candidate transfer ban may be justified by the state’s interest in preventing corruption or the appearance of corruption by “political power brokers.” Brief of Defendants in Intervention-Appellants [Amended] at 49. Even if we assume this to be an important state interest, the ban is not “closely drawn to avoid unnecessary abridgment of associational freedoms.” Buckley, 424 U.S. at 25, 96 S.Ct. at 638. The potential for corruption stems not from campaign contributions per se but from large campaign contributions. Id. at 28, 96 S.Ct. at 639. The inter-candidate transfer ban prohibits small contributions from one candidate to another as well as large contributions. We hold, therefore, that the inter-candidate transfer ban is unconstitutional because it fails the “rigorous” test used in Buckley, 424 U.S. at 29, 96 S.Ct. at 640.

Ill

Finally, we consider Proposition 73’s prohibition on the expenditure of funds raised prior to January 1989. The Authors, but not the FPPC, appeal the district court’s invalidation of this provision.19

The Authors concede that the ban on the use of pre-1989 funds operates as a restriction on expenditures. As an expenditure limitation, the ban on pre-1989 funds must be “narrowly tailored to serve a compelling state interest.” Austin, 110 S.Ct. at 1396. The only purpose of the ban advanced by the Authors is to purge previously unregulated contributions from a regulated system under Proposition 73. Given our decision that Proposition 73’s contribution limits are invalid, this purpose is no longer served by the prohibition on the expenditure of pre-1989 funds.

The judgment of the district court is AFFIRMED.

dissenting:

The majority’s remarkable opinion is best understood by examining the interests of the parties in this litigation. The plaintiff/ appellee in this case is the Service Employees International Union, AFL-CIO, a special interest group classified as a person under Proposition 73 and subject to the $1,000 contribution limitation. See Cal. Gov’t Code §§ 85102(b), 85301(a). There is no doubt that Proposition 73 will severely limit the Service Employees International Union’s campaign contributions. The Union is represented by Rencho, Johansen & Purcell, a law firm that has a long relationship with the California Democratic Party and has served as counsel to the California Democrats. Indeed, the California Democratic Party itself has joined this suit as a plaintiff in intervention, undemocratically placing itself in direct opposition to a majority of the voters in California. When this suit was filed on March 24, 1989, 24 members (60%) of the California Senate were Democrats, while 15 were Republicans and 1 was independent, and 47 members (59%) of the California State Assembly were Democrats, while 32 were Republicans. Thus, the California Democratic Party had a strong interest in keeping its incumbents in office and retaining control of the State Assembly and Senate.

The Union, the Democratic Party, Democratic incumbents, and other plaintiffs with closely related interests brought this challenge to Proposition 73 before Judge Karlton. 747 F.Supp. at 581. Incredible as it may seem, the plaintiffs, who would be disadvantaged by Proposition 73 because it harms incumbents and special interests, argued that Proposition 73 was unconstitutional because it discriminates against challengers.1 Judge Karlton accepted this *1324unbelievable argument and struck down Proposition 73 as unconstitutional. Now Judge Poole and Judge Norris sanction this argument on appeal.

The majority determines that Proposition 73 is nothing more than invidious discrimination against challengers, while ignoring the district court’s conflicting findings showing that Proposition 73 usually works against incumbents and therefore benefits challengers. The majority then cloaks its decision as a finding of fact that this court should be hesitant to disturb. This treatment of the district court’s decision as a sacrosanct finding of fact is instructive. Because the majority decision cannot stand on its own merit, the decision is justified by suggesting that we must follow the district court’s conclusion unless that conclusion was clearly erroneous.

To the contrary, I maintain that whether Proposition 73 works an invidious and unconstitutional discrimination on challengers and their supporters is a conclusion of law. At a minimum, this issue is a mixed question of law and fact. Thus, the district court’s conclusion is not entitled to any special deference. See United States v. Spillone, 879 F.2d 514, 520 (9th Cir. 1989) (mixed questions of' law and fact are reviewed de novo).

The district court’s conclusion that Proposition 73 discriminates against challengers is not a finding of fact but an ultimate conclusion “to be distinguished from the findings of primary, evidentiary or circumstantial facts.” Helvering v. Tex-Penn Oil Co., 300 U.S. 481, 491, 57 S.Ct. 569, 574, 81 L.Ed. 755 (1937), cited with approval in United States v. McConney, 728 F.2d 1195, 1202 (9th Cir.) (en banc), cert. denied, 469 U.S. 824, 105 S.Ct. 101, 83 L.Ed.2d 46 (1984). The relevant factual findings are undisputed, and the discrimination conclusion turns on the legal interpretation of the undisputed facts.2 This court is in just as good a position as the district court to review the facts and determine whether Proposition 73 works an unconstitutional discrimination on challengers.

Even if we assume, however, that the supposed discrimination is a finding of fact, it would still be clearly erroneous. The record is completely barren of any evidence of invidious discrimination. Proposition 73 limits campaign contributions to reduce both corruption and the appearance of corruption in the California political system. Findings of Fact 46 & 47. The contribution limitations are based on a July 1 through June 30 fiscal year to correspond to California’s June primaries. Thus, a contributor who gave the maximum allowable contribution to a candidate in the primary may make another contribution for the general election as soon as the primary is over. See 747 F.Supp. at 589. At trial, the plaintiffs failed to produce any evidence that this system was a result of invidious or purposeful discrimination. At most, the plaintiffs show that the fiscal year provision may have a disparate effect on some challengers. Standing alone, however, a disparate impact does not establish invidious discrimination. See Washington v. Davis, 426 U.S. 229, 242, 96 S.Ct. 2040, 2049, 48 L.Ed.2d 597 (1979).

Moreover, the record in this case demonstrates that Proposition 73 helps rather than harms challengers. There is ample evidence suggesting that on the whole the contribution limits in this case favor challengers by disproportionately burdening incumbents. When the district court’s findings are reviewed together, it is clear that any negative effects a fiscal year basis may have on challengers are overcome by the benefits challengers receive from contribution limits.

In its findings of fact, the district court made findings demonstrating that incumbents are more likely than challengers to receive contributions in excess of Proposition 73 limits: “In the three general elec*1325tion years immediately preceding the effective date of Proposition 73, incumbents in contested California legislative races received a greater percentage of contributions in excess of the limits established by Proposition 73 than did their challengers.” Finding of Fact 87. Thus, the overall effect of Proposition 73 is to narrow the gap between incumbent and challenger fund-raising.

The district court found that incumbents are likely to raise more money than challengers in non-election years, in part because challengers begin their campaigns later. Thus, the district court and majority theorize that Proposition 73 disadvantages .challengers by preventing them from overcoming their late fundraising start and narrowing the contribution gap in the election year. This theory is absurd given that on average incumbents consistently receive more contributions than challengers in election years. Because incumbents are more likely to receive contributions in excess of the Proposition 73 limits, even in election years, the net effect of Proposition 73 is to narrow rather than exacerbate the fund-raising gap between incumbents and challengers.

Moreover, in those campaigns where incumbents are raising funds in early non-election years but challengers are not, Proposition 73 helps challengers by, reducing the amount incumbents would otherwise raise in such circumstances. Given that this fundraising gap already exists and will continue to exist with or without Proposition 73, challengers are much better off with contribution limits to reduce the amounts incumbents raise before challengers begin their campaigns.

By striking down Proposition 73 because it supposedly disadvantages challengers, the majority eliminates contribution limits and places challengers in an even more disadvantageous position. Indeed, the majority disingenuously bolsters its decision by comparing Proposition 73 with an alternative that does not exist.

There are three basic alternatives for regulating campaign contributions presented in the context of this case. The first alternative — no limits on contributions — is the most advantageous to incumbents and harmful to challengers because on average incumbents receive more large contributions. The second alternative — Proposition 73 — benefits challengers because it places limits on contributions and narrows the fundraising gap between challengers and incumbents. The third alternative — contribution limits based on election cycles (the federal model)3 — may be even more advantageous to challengers. It would still help to reduce the advantage- incumbents obtain by receiving more large campaign contributions, while also allowing challengers to receive more money from contributors in election years to the extent those contributors did not contribute earlier in the campaign (usually because the challengers did not start-their campaigns earlier).

The majority’s decision rests primarily on the conclusion that the third alternative is more advantageous to challengers than Proposition 73. At bottom, the majority’s conclusion is that Proposition 73 unconstitutionally discriminates against challengers because it is not as advantageous to challengers as contribution limits on an election cycle basis (the federal model) would be.

The problem with this reasoning is that the third alternative is a strawman alternative. The voters of California adopted Proposition 73, not the federal model. Although the federal model may be even more advantageous to challengers, Proposition 73 is still better for challengers than the first alternative of no contribution limits at all. Nevertheless, because the majority focuses on the federal model, they reach the incredible conclusion that a law which benefits challengers by helping to close the fundraising gap between incumbents and challengers is unconstitutional because it discriminates against challengers. The very groups that would be harmed most by contribution limits — special interests and *1326incumbents — are able to avoid Proposition 73 through the disingenuous argument that it discriminates against challengers.

Even more troubling than the majority’s reasoning, however, is the majority’s disregard for Supreme Court precedent. The Supreme Court has made it clear that “a court should generally be hesitant to invalidate [campaign contribution] legislation which on its face imposes evenhanded restrictions.” Buckley v. Valeo, 424 U.S. 1, 31, 96 S.Ct. 612, 641, 46 L.Ed.2d 659 (1976). A record showing invidious discrimination against challengers will rebut the general presumption of validity, but the record in this case simply does not show that Proposition 73 discriminates against challengers. As discussed above, and identical to the reasoning in Buckley, “to the extent that incumbents generally are more likely than challengers to attract very large contributions, [Proposition 73’s] ceiling has the practical effect of benefiting challengers as a class.” 424 U.S. at 32, 96 S.Ct. at 641 (emphasis added).. Moreover, even if Proposition 73 could be viewed as discriminatory, there is simply no showing of any invidious discrimination.

At best, the plaintiff/appellees in this case show that some challengers may be disadvantaged in some elections. However, the evidence shows that in most elections Proposition 73 actually disadvantages incumbents and benefits challengers. Again, similar to Buckley, “the record provides no basis for predicting that such adventitious factors will invariably and invidiously benefit incumbents as a class.” Id. at 33, 96 S.Ct. at 641.

Like the federal model at issue in Buckley, Proposition 73’s “primary purpose — to limit the actuality and appearance of corruption resulting from large individual campaign contributions” — is a sufficiently compelling interest to justify interference with protected rights. 424 U.S. at 26, 96 S.Ct. at 638. To overcome the evils associated with large campaign contributions,4 it is necessary that the rights of incumbents, special interests, contributors, and even challengers will be burdened to some extent. Under Buckley's First Amendment discrimination analysis, these burdens are constitutional unless the contribution limitation “invariably and invidiously” discriminates against one class for the benefit of another. Id. at 33, 96 S.Ct. at 641. Thus, to survive the constitutional challenge in this case, it is enough that Proposition 73 (1) serves its corruption-fighting purpose and (2) does not place challengers in a worse position.

Apparently, the majority believes that campaign contribution limitations are unconstitutional unless every part of the limitation is carefully tailored to benefit challengers. Although Proposition 73 actually benefits challengers and requires incumbents to bear a disproportionate share of the burden of eliminating actual and apparent corruption, the majority strikes down Proposition 73 because the fiscal year provision does not skew the balance even farther in favor of challengers. This reasoning misses the point that benefitting or harming incumbents or challengers is not a legitimate purpose for regulating campaign contributions.

Simply put, the majority’s reasoning is not a faithful or even a reasonable application of Buckley. Indeed, to justify this same line of reasoning, the district court suggested that Buckley has been overruled by Citizens Against Rent Control v. Berkeley, 454 U.S. 290, 102 S.Ct. 434, 70 L.Ed.2d 492 (1981). See 747 F.Supp. at 588, 589 n. 20. Although the majority is not as direct as the district court, the ma*1327jority shows the same disregard for the Buckley mandate.

Even if the majority were not overlooking Supreme Court direction, the majority’s discrimination analysis would still be in error because the decision sanctions such a transparent attempt by incumbents to overcome a law that inadvertently benefits challengers. However, given the clear mandate of Buckley, the majority decision is especially egregious. With this decision, I believe the majority disregards the judgments of the Supreme Court and the people of California as well as common sense.

In Summary, Proposition 73 represents a balanced effort to eliminate a significant political problem — the actuality and appearance of corruption resulting from large campaign contributions. Under Buckley’s discrimination analysis, this purpose may be served absent a showing that Proposition 73 invidiously discriminates against challengers as a class. Because (1) the record shows no invidious discrimination and (2) Proposition 73 actually benefits challengers as a class, Proposition 73 does not violate the Constitution. Incumbents and special interests should not be able to avoid this law by arguing that the law does not benefit challengers enough.5

I began this dissent by observing that this litigation can best be understood by examining the interest of the parties who brought it. Without question, it is an effort by the Democratic Party and its major financial supporters to strike down a measure adopted by the people of California to curtail perceived abuses in campaign financing. The majority opinion affirms the district court in finding the popularly adopted measure unconstitutional. Thus, the majority returns the law in California to where it was before Proposition 73 was enacted — no limitations on campaign contributions. In a most bizarre twist of reasoning, the majority permits the law to return to a condition where the maximum benefit is afforded the incumbents, by accepting their argument that the challenged law provides insufficient advantages to challengers.

It requires no special insight to conclude that this case turned out exactly as the Democratic plaintiffs intended: to preserve the seats of incumbent Democratic members of the state legislature and to insulate them from successful challenge.