McComish v. Bennett

611 F.3d 510 | 9th Cir. | 2010

Lead Opinion

ORDER

The opinion filed May 21, 2010, slip op. 7319, is amended by deleting footnote 1, slip op. at 7326, and substituting the following footnote in its place:

1 Governor Symington was sentenced to a 30-month term of imprisonment, but his conviction was reversed and the case remanded for a new trial. See United States v. Symington, 195 F.3d 1080 (9th Cir.1999). He was then granted a presidential pardon just as his retrial was set to commence.

OPINION

TASHIMA, Circuit Judge:

This is a challenge to the constitutionality of the “matching funds” provision of Arizona’s Citizens Clean Elections Act, Ariz.Rev.Stat. § 16-952. The Act establishes a legal framework within which the State of Arizona may provide public financing to candidates for state political offices. A candidate who chooses to participate in the Act’s voluntary public financing scheme relinquishes her or his right to raise private campaign contributions. Instead, she or he receives an initial grant of funds from the state to spend on her or his campaign. The challenged provision ensures that if the participating candidate has an opponent who is not participating in the public financing system and whose campaign expenditures or contributions exceed a threshold set by the Act, she or he receives additional matching funds from the State.

Six past and future candidates for Arizona political office who have, or plan to, run privately-financed campaigns, as well as two political action committees who fund such candidates, brought suit to enjoin the Act’s matching funds provision, alleging that it violates their rights under the First Amendment and the Equal Protection Clause of the Fourteenth Amendment of the U.S. Constitution. These Plaintiffs claim that the matching funds provision severely burdens them exercise of protected political speech by punishing them for making, receiving, or spending campaign contributions. As nonparticipating candidates, if they exceed the Act’s matching funds threshold, they will trigger the disbursement of matching funds to their opponents. They allege that their fear of triggering matching funds to their opponent causes them to curb their campaign fundraising or spending, thereby chilling their speech. They also claim that because the Act treats candidates differently based on whether or not they participate in the public financing scheme, it denies them the equal protection of the law.

The district court held that the matching funds provision of the Act violated the First Amendment. It did not reach Plaintiffs’ equal protection claim. After determining that the matching funds provision of the Act could not be severed from the Act as a whole, the district court granted Plaintiffs’ motion for summary judgment, issued a declaratory judgment that the Act violates the First Amendment, and enjoined its enforcement.

Applying Supreme Court precedent analyzing campaign finance laws under the First Amendment, see Citizens United v. FEC, - U.S. -, 130 S.Ct. 876, - L.Ed.2d - (2010); Davis v. FEC, - U.S. -, 128 S.Ct. 2759, 171 L.Ed.2d 737 (2008); Buckley v. Valeo, 424 U.S. 1, 96 S.Ct. 612, 46 L.Ed.2d 659 (1976) (per curiam), we conclude that the matching funds provision of the Act imposes only a minimal burden on First Amendment rights. It survives intermediate scrutiny because it bears a substantial relation to the State’s important interest in reducing quid pro quo political corruption. Because the Act conforms to the requirements of the First Amendment and must be upheld, we reverse. We decline to address Plaintiffs’ equal protection claim in the first instance; instead, we remand to the district court so that it may consider the issue.

I. Statutory Background

Prior to the passage of the Citizens Clean Elections Act (the “Act”), Arizona had already adopted campaign contribution limits. Ariz.Rev.Stat. § 16-905 (historical and statutory note). In 1986, the State’s voters passed an initiative measure establishing individual contribution limits of $200 for legislative candidates and $500 for statewide candidates, per election. See id. Even with these campaign contribution limits in place, Arizona experienced a series of massive political corruption scandals.

In 1988, Governor Evan Mecham was indicted on multiple criminal charges, including perjury and fraud for allegedly hiding a campaign loan. He was later impeached on charges of misuse of public funds and obstruction of justice, and ousted from office. Next, the “Savings and Loan Scandal” led to a United States Senate Ethics Committee investigation of certain activities of both U.S. Senators from Arizona, who had received contributions and favors from Arizona savings and loan tycoon Charles Keating.

Then, in 1991, AzScam erupted. A sting operation caught state legislators on videotape accepting campaign contributions and bribes in exchange for agreeing to support gambling legislation. The video footage was generally seen as outrageous. For example, Representative Don Kenny was seen stuffing a $55,000 cash bribe into a gym bag after joking, “are you sure there are no hidden cameras up there?” AzScam resulted in the indictment of twenty-one individuals, including lobbyists, political activists, and seven state legislators. Two additional state legislators were named in a civil racketeering suit. In total, nearly ten percent of the Arizona Legislature at the time faced civil or criminal charges related to AzScam. Former Arizona Governor J. Fife Symington testified that the scandal was highly publicized around the state.

In the late nineties, Governor Symington himself was embroiled in scandal and indicted on twenty-three counts, including using the power of his office to extort concessions from a pension fund to which he owed $10 million. He became Arizona’s second governor in a row to leave his office in disgrace when he resigned in 1997, after being convicted on seven counts of filing false financial statements.¹

In the wake of these scandals, the voters of Arizona passed the Citizens Clean Elections Act, Ariz.Rev.Stat. §§ 16-940-16-961, an initiative measure, in the 1998 statewide election. Ariz.Rev.Stat. § 16-940 (historical and statutory note). There is no evidence that the Act was intended solely to remedy Arizona’s apparent susceptibility to political corruption. Plaintiffs, however, introduced ample evidence indicating that when the Act was adopted, voters were aware of, and concerned about, continuing and repeated political corruption in Arizona.

The voter information pamphlet for the 1998 election argued that the Act should be approved because it would free politicians “to represent the best interests of all the citizens, not just the large financial contributors who can trade their cash for political support.” Ballot Propositions Publicity Pamphlet for the 1998 Arizona General Election, at *87, available at http://www.azsos.gov/election/1998/info/ pubpamphlet/ prop200.pdf. One justification for the Act was that it would change Arizona’s “reputation [as] a state rife with corruption and the abuse of money in politics .... [and] restore confidence in our political system.” Id. at *88. The pamphlet further noted that under the existing election law regime, “[o]ur elected officials are going to jail and this cycle of abuse seems endless.”² Id.

The Act’s findings state that the State’s previous election financing system, “[u]ndermine[d] public confidence in the integrity of public officials,” because it “[a]l-low[ed] Arizona elected officials to accept large campaign contributions from private interests over which they have governmental jurisdiction.” Ariz.Rev.Stat. § 16-940. The Act’s stated purpose is to “create a clean elections system that will improve the integrity of Arizona state government by diminishing the influence of special-interest money, will encourage citizen participation in the political process, and will promote freedom of speech.” Id.

Plaintiffs argue that the Act was not aimed at reducing corruption, and that legislative reforms designed to prevent another AzScam had already been adopted at the time the Act was passed. Instead, Plaintiffs contend that the Act’s true purposes were to level the political playing field and reduce campaign spending.³ Voters are motivated by varied and conflicting motivations. Generally, a diverse electorate cannot be said to share one true intent in adopting an initiative measure. More specifically, the extent to which Arizona’s various corruption scandals led to the passage of the Act cannot be precisely determined. Based on the record before us, however, we conclude that one of the principal purposes of the Act was to reduce quid pro quo corruption.

The Act created a system of public financing for political campaigns. A candidate who chooses not to participate in this system may raise unlimited funds from private donations, subject to contribution limits and disclosure requirements, which existed before the Act. If a candidate opts to participate in the public financing system, she or he agrees to forfeit her or his right to fund her or his campaign with private contributions. Instead, she or he must collect a required number of five-dollar “qualifying contributions” during a specified time period to demonstrate that she or he has the voter support to be a viable candidate. Ariz.Rev.Stat. § 16-946. The number of required qualifying contributions varies between two hundred and four thousand, depending on the office that the candidate is seeking. Ariz.Rev.Stat. § 16-950(D).

If the candidate qualifies for public financing, she or he will then receive a lump-sum grant for her or his primary campaign, which varies depending on whether she or he is running opposed or unopposed in a party primary, or whether she or he is an independent candidate. Ariz.Rev.Stat. § 16-951. If the participating candidate has a nonparticipating opponent in the primary who spends more than her or his initial grant, or whose expenditures, combined with the value of independent expenditures in opposition to her or his candidacy or in support of her or his nonparticipating opponent,⁴ exceed the amount of her or his initial grant, the participating candidate will receive “matching funds” in the amount of the combined spending of her or his nonparticipating opponent, plus the value of independent expenditures against her or him or in support of her or his nonparticipating opponent, reduced by six percent and reduced by the amount of “early contributions” raised by the nonparticipating opponent during the preprimary fundraising period.⁵ Ariz.Rev.Stat. §§ 16-952, 16-945.

If the participating candidate wins her or his primary and continues on to the general election, she or he receives a second initial lump-sum grant. Ariz.Rev. Stat. § 16-952. If she or he has a nonparticipating opponent in the primary whose contributions received, combined with the value of independent expenditures in supporting him or her and opposing the participating candidate, less his or her expenditures during the primary campaign, exceed the amount of the participating candidate’s second initial grant, the participating candidate will receive matching funds, calculated in the same manner as in the primary campaign. Ariz.Rev. Stat. § 16-952(B). During both elections, matching funds, combined with the initial grant, may not exceed three times the amount of the initial grant. Ariz.Rev.Stat. § 16-952(E). This means that a nonparticipating candidate who is able to raise funds in excess of three times the amount of his or her participating candidate’s initial grant gains a potentially unlimited financial advantage in the campaign.

II. Factual Background

Plaintiffs are John McComish and Nancy McLain, current members of the Arizona House of Representatives who are seeking reelection in 2010, Tony Bouie, a non-incumbent candidate for the Arizona House of Representatives, and Robert Burns, an Arizona State Senator seeking reelection in 2010. Joining them in this challenge to the Act are plaintiffs-intervenors, Dean Martin, a former Arizona State Senator, current State Treasurer, and candidate for Governor in 2010, Rick Murphy a member of the Arizona House of Representatives seeking reelection in 2010, the Arizona Free Enterprise Club’s Freedom Club PAC (“Freedom Club PAC”), a political action committee (“PAC”) which makes contributions to Arizonans for a Sound Economy, an independent expenditure committee that funds independent expenditures in races subject to the Act, and the Arizona Taxpayers Action Committee (“Arizona Taxpayers”), an independent expenditure committee that makes independent expenditures in races subject to the Act. All candidate-plaintiffs have declined to participate in the Act’s public funding program and will be running in the 2010 election as privately-financed candidates.

Plaintiffs sued Jan Brewer, in her official capacity as Arizona Secretary of State,⁶ and Gary Scaramazzo, Royann Parker, Jeffrey Fairman, Donald Lindholm, and Lori Daniels, in their official capacities as members of the Arizona Citizens Clean Elections Commission (collectively, the “State”). The Clean Elections Institute, Inc., a nonprofit organization formed to continue the work of Arizonans for Clean Elections, the campaign committee that sponsored the initiative that created the Act, intervened as a defendant.

Since 2002, between fifty-two percent and sixty-seven percent of candidates for state office during general elections have chosen to participate in the Act’s public funding program. Plaintiffs contend that matching funds have dampened overall campaign spending in Arizona. It is undisputed, however, that overall campaign spending in Arizona has increased since the Act’s passage.

Plaintiffs also allege that matching funds have specifically deterred them from engaging in political expression through monetary expenditures. They argue that the fear of triggering matching funds causes privately-funded candidates to refrain from or delay raising and spending money in support of their campaigns. They contend that matching funds burden the speech of independent expenditure committees by causing them to delay making independent expenditures in order to avoid triggering matching funds until later in the campaign, when the publicly-financed candidate will have less time to use them, in an attempt to minimize the competitive benefit of the matching funds to the publicly-financed candidate.

We agree with the district court’s observation that “Plaintiffs’ testimony is somewhat scattered and shows only a vague interpretation of the burden of the Act.” John MeComish testified that in his 2008 campaign for the House of Representatives, he triggered matching funds to his opponents. In order to avoid triggering additional matching funds to his opponents until later in the campaign, he “deci[ded] to wait and not make any substantial expenditures until September 2, 2008.” MeComish describes this as an act of “self-censorship” in order to avoid the consequence of “disseminating] viewpoints hostile to my campaign.”

Nancy McLain testified that during her 2008 campaign for House of Representatives she “made a conscious decision to not raise or spend a significant amount because it could not be spent without resulting in ‘equalizing funds.’” Although she avoided triggering matching funds in the general election, she triggered matching funds in the primary. She claims that her “reluctance to fundraise and spend money during the primary election meant that [she] was not able to broadcast [her] message either on air or in newspapers as often as[she] thought necessary to win the election,” although she was victorious. She describes matching funds as imposing a “competitive disadvantage” on her campaign, which her fundraising and spending choices “seek to minimize.”

Tony Bouie also triggered matching funds during his 2008 campaign for House of Representatives. He testified that because he had “no control over independent expenditures that the Commission deems [to] benefit him” he “fe[lt] compelled to conserve money for damage-control in anticipation of poorly-conceived independent expenditures backfiring after they trigger matching funds” to his opponents. As a result he “change[d his] campaign strategy” by “holding [his] campaign speech until [his] expenditures could be timed to minimize the impact of the matching funds.” He believes that matching funds thereby “placed [him] at a continuous tactical disadvantage.”

Robert Burns “believefs] that the existence of matching funds coerces traditionally funded candidates into changing their message and the timing of getting out their message, even if ultimately the amount of messaging a traditional candidate chooses to engage in does not change.” His deposition testimony indicated that matching funds had not actually chilled his speech during his most recent 2008 election for State Senate. He admitted that he was willing to trigger matching funds. Further, he also indicated that despite the matching funds provision, he simply communicated his message to the extent he felt was necessary to win, stating “[i]f I had to spend X number of dollars to get out a — a mailer, and I had that amount of money, I would go ahead and do the mailer.”

Dean Martin stated that “the matching funds provision of the Act forced [him] to self-censor” in his most recent race for State Treasurer in 2006. He reasoned that the “only way [he] could avoid being massively outspent by [his] opponent was to prevent the triggering of additional matching funds to her campaign” and “was thus coerced into accepting the same limits as [his] government-funded opponent ... [without] receiving] any of the benefits of being a government-funded candidate, such as the receipt of matching funds for independent expenditures that opposed [his] candidacy and the label ‘participating’ or ‘clean’ candidate.” Despite Martin’s testimony that he felt matching funds severely burdened his speech, he curiously could not even recall whether he had ever triggered matching funds to an opponent. The record indicates that Martin is fundamentally opposed to the policy of publicly financing political campaigns, regardless of whether a given public financing scheme includes a provision for matching funds.

Rick Murphy ran successfully for House of Representatives as a participating candidate in 2004. He later ran for reelection in 2006 and 2008 as a nonparticipating candidate. He testified that he attempted to not trigger matching funds during his 2006 campaign because, since his opponent would receive additional matching funds once he surpassed the threshold amount, “[i]t didn’t seem like it made a lot of sense for me to do that, so I curtailed my speech and curtailed my fundraising in order to prevent it.” Still, Murphy testified that matching funds have never prevented him from accepting a political contribution. Murphy’s testimony that he curtailed his fundraising for fear of triggering matching funds was directly contradicted by his own campaign consultant, who testified that he advised Murphy to “raise as much money” as he could from “everywhere,” that he never advised Murphy to stop raising money, and that Murphy never curtailed or stopped his fundraising efforts.

The treasurer of the Freedom Club PAC testified that the PAC has never been prevented from making an independent expenditure for fear of triggering matching funds. He further stated that he takes into account various campaign finance “reporting requirements, notifications and filings” upon which matching funds disbursement determinations by the Citizens Clean Elections Commission are based, in determining when to spend money.

Arizona Taxpayers stated that it “decided not to speak in opposition to [a] participating candidate ... in the 2006 primary election because such speech would have triggered matching funds.” However, the groups’s founder admitted that it had never decided against making an expenditure because it would trigger matching funds. Further, the evidence demonstrates that Arizona Taxpayers could not afford to make expenditures during the 2006 primary election, because it had only $52.72 in cash on hand.

III. Standard and Scope of Review

We review the district court’s grant of summary judgment to Plaintiffs and denial of summary judgment to the State de novo.⁷ See Block v. City of L.A., 253 F.3d 410, 416 (9th Cir.2001). This means that we will look at the case “anew, the same as if it had not been heard before, and as if no decision previously had been rendered,” and “giving no deference to the district judge’s determinations.” Freeman v. DirecTV, Inc., 457 F.3d 1001, 1004 (9th Cir.2006); Ditto v. McCurdy, 510 F.3d 1070, 1075 (9th Cir.2007).⁸ We apply the same legal standard that the district court did, “[viewing the evidence in the light most favorable to the nonmoving party,” and granting summary judgment “only if no genuine issues of material fact remain for trial and the moving party is entitled to judgment as a matter of law.” Block, 253 F.3d at 416. We “may affirm on any ground supported by the record.” N. Mariana Islands v. United States, 399 F.3d 1057, 1060 (9th Cir.2005). Because the district court did not rule on the parties’ evidentiary objections and did not strike any evidence from the record, all of the evidence adduced below remains a part of the record before us on appeal. See Vinson v. Thomas, 288 F.3d 1145, 1152 & n. 8 (9th Cir.2002).

IV. Analysis

First, we must determine what level of constitutional scrutiny applies to the Act’s matching funds provision. The level of scrutiny that applies to a law which implicates First Amendment concerns is “dictated by both the intrinsic strength of, and the magnitude of the burden placed on, the speech and associational freedoms at issue.” Lincoln Club v. City of Irvine, 292 F.3d 934, 938 (9th Cir.2002). We must, therefore, engage in a two-step inquiry. First, we determine whether the type of speech the law affects is fully protected by the First Amendment. Next, we determine the type of burden, if any, that the law places on that speech. Laws that place severe burdens on fully protected speech are subject to strict scrutiny. Laws that “placet ] only a minimal burden on fully protected ... freedoms” or that apply to “speech and associational freedoms [that] are not fully protected by the First Amendment” receive intermediate scrutiny. Id.

Buckley, the Supreme Court’s landmark case on the constitutionality of campaign finance laws under the First Amendment, provides guidance in determining whether the Act affects fully protected speech. Buckley held that campaign contributions are not fully protected political speech. 424 U.S. at 21-22, 96 S.Ct. 612; see also Lincoln Club, 292 F.3d at 938-39. The Supreme Court reasoned that a “contribution serves as a general expression of support for the candidate and his views, but does not communicate the underlying basis for the support. ... [T]he transformation of contributions into political debate involves speech by someone other than the contributor.” Buckley, 424 U.S. at 21, 96 S.Ct. 612. By contrast, campaign expenditures are fully protected speech because a “restriction on the amount of money a person can spend on political communication during a campaign necessarily reduces the quantity of expression by restricting the number of issues discussed, the depth of their exploration, and the size of the audience reached.” Id. at 19, 96 S.Ct. 612.

Following Buckley, a law that has a hybrid effect on both contributions and expenditures is interpreted as though it affects fully protected speech. See Lincoln Club, 292 F.3d at 939. The matching funds provision of the Act affects both contributions and expenditures. During primary elections, the Act explicitly ties participating candidate’s matching fund disbursements to nonparticipating candidates’ expenditures. Ariz.Rev.Stat. § lb-952. During a general election, the Act bases matching fund calculations on contributions received by nonparticipating candidates. Id. This includes contributions that come from a nonparticipating candidate’s personal funds, which, under Buckley’s framework, are considered expenditures, and fully protected speech. See 424 U.S. at 51-55, 96 S.Ct. 612. Because the Act affects both contributions and expenditures, we analyze it as though it affects fully protected speech.

Next, we determine what level of burden, if any, the Act imposes on that speech. Plaintiffs argue that the Supreme Court’s recent decision in Davis, 128 S.Ct. 2759, controls this case, and compels the conclusion that the Act places a severe burden on their speech triggering the application of strict scrutiny. The State and DefendanU-Intervenor counter that the Act places only a minimal or indirect burden on Plaintiffs’ speech and that the Supreme Court’s latest campaign finance decision, Citizens United, 130 S.Ct. 876, reaffirmed that intermediate scrutiny applies to such laws.

In Buckley, the Supreme Court evaluated the constitutionality of provisions of the Federal Election Campaign Act of 1971, and related provisions of the Internal Revenue Code of 1954, as amended, which allowed for the public financing of presidential elections. See 424 U.S. at 6, 85-109, 96 S.Ct. 612. The Court found that the public financing scheme in that case was “a congressional effort, not to abridge, restrict, or censor speech, but rather to use public money to facilitate and enlarge public discussion and participation in the electoral process, goals vital to a self-governing people. Thus [it] furthers, not abridges, pertinent First Amendment values.” Id. at 92-93, 96 S.Ct. 612 (footnote omitted). Therefore, the public financing of elections itself does not create any burden on speech.

In Davis, the Court struck down the Millionaire’s Amendment to the Bipartisan Campaign Reform Act of 2002 (“BCRA”), 2 U.S.C. § 441a-l(a). The Millionaire’s Amendment applied to campaigns for the United States House of Representatives, in which all candidates are privately financed and there is no public funding. But if one candidate financed his campaign with personal funds in excess of a threshold amount, “a new, asymmetrical regulatory scheme e[ame] into play.” 128 S.Ct. at 2766. While the self-financing candidate continued to be subject to the normal contribution limits, his opponent, the “non-self-financing” candidate became eligible to “receive individual contributions at treble the normal limit ... even from individuals who may have reached the normal aggregate contributions cap, and [could] accept coordinated party expenditures without limit.” Id. The Davis Court held that this “scheme of discriminatory contribution limits,” “imposes a substantial burden on the exercise of the First Amendment right to use personal funds for campaign speech.” Id. at 2772. Having found the Millionaire’s Amendment to be a substantial burden on fully protected speech, the Court applied strict scrutiny. Id.

Plaintiffs urge us to adopt the rationale of the district court and conclude that, under the logic of Davis, they have “established a cognizable burden” under the First Amendment. The district court reasoned that “[i]f the mere potential for your opponent to raise additional funds is a substantial burden, the granting of additional funds to your opponent must also be a burden.” We disagree. As discussed below, we conclude that Davis is easily and properly distinguished from the case at bench.

The regulatory framework the Supreme Court examined in Davis is different from the one we confront under the Act. Davis says nothing about public “funding schemes and therefore says nothing about their constitutionality.” Comment, 122 Harv. L.Rev. 375, 383 (Nov.2008). All of the candidates in Davis were subject to the same scheme regulating privately financed candidates. “Under the usual circumstances, the same restrictions apply to all the competitors for a seat and their authorized committees.” Davis, 128 S.Ct. at 2765. But if one candidate heavily self-funded his campaign, it triggered an entirely new “asymmetrical regulatory scheme” that required him to file three additional types of campaign finance disclosures not required of his opponents. Id. at 2766-67. “Failure to comply with the reporting requirements may [have] resulted] in civil and criminal penalties.” Id. at 2767. It also triggered new, more generous campaign contribution limits, but only for the self-financing candidate’s opponents. Id. at 2771.

In striking down the Millionaire’s Amendment, the Supreme Court noted that had the law “simply raised the contribution limits for all candidates, Davis’ argument would plainly fail.” Id. at 2770; see also id. at 2771 (“Consequently, if § 319(a)’s elevated contribution limits applied across the board, Davis would not have any basis for challenging those limits.”). Instead, the law constituted a burden on Davis’ speech only because it treated candidates running against each other under the same regulatory framework differently based on a candidate’s decision to self-finance his or her campaign, which is a manner of engaging in fully protected speech. Id. at 2771.

In contrast, the Supreme Court has held that it is constitutional to subject candidates running against each other for the same office to entirely different regulatory schemes when some candidates voluntarily choose to participate in a public financing system. Buckley, 424 U.S. at 97, 96 S.Ct. 612. “[T]he Constitution does not require Congress to treat all declared candidates the same for public financing purposes.” Id. “Sometimes the grossest discrimination can lie in treating things that are different as though they were exactly alike.” Id. at 97-98, 96 S.Ct. 612.

In Davis, the Millionaire’s Amendment was an attempt to “level electoral opportunities for candidates of different personal wealth.” 128 S.Ct. at 2773. This was problematic because Buckley had held that “the First Amendment simply cannot tolerate [a] restriction upon the freedom of a candidate to speak without legislative limit on behalf of his own candidacy.” 424 U.S. at 54, 96 S.Ct. 612. In Citizens United, the Supreme Court elaborated on its decision in Davis, indicating that the Millionaire’s Amendment was unconstitutional because it specifically sought to disadvantage the rich. “The rule that political speech cannot be limited based on a speaker’s wealth is a necessary consequence of the premise that the First Amendment generally prohibits the suppression of political speech based on the speaker’s identity.” Citizens United, 130 S.Ct. at 905. Under the Act, while matching funds are calculated based on the total contributions received and expenditures made by a nonparticipating opponent, they are not distributed specifically to the opponents of wealthy candidates. Matching funds do not distinguish between different sources of nonparticipating candidates’ financing at all. The law in Davis was problematic because it singled out the speakers to whom it applied based on their identity. The Act’s matching funds provision makes no such identity-based distinctions.

Plaintiffs, perhaps recognizing that they have not demonstrated any actual chilling of their speech by the Act, argue that under Davis, we could strike down the matching funds provision without any proof that their speech has been deterred or punished. But Davis does not require this Court to recognize mere metaphysical threats to political speech as severe burdens. We will only conclude that the Act burdens speech to the extent that Plaintiffs have proven that the specter of matching funds has actually chilled or deterred them from accepting campaign contributions or making expenditures.

Based on the record before us, we conclude that any burden the Act imposes on Plaintiffs’ speech is indirect or minimal.⁹ Since the Act’s adoption, campaign spending in Arizona has increased. Several Plaintiffs testified that they would have made increased expenditures or undertaken increased fundraising but for the matching funds provision. No Plaintiff, however, has pointed to any specific instance in which she or he has declined a contribution or failed to make an expenditure for fear of triggering matching funds. The record as a whole contradicts many of Plaintiffs’ unsupported assertions that their speech has been chilled. Robert Burns admitted he was willing to trigger matching funds and spent as much money as he needed to in order to communicate his message. Dean Martin claimed his speech was chilled by matching funds, but could not even recall whether he had triggered them in the past. Rick Murphy admitted he had never turned a contribution away, and his political consultant testified that he never slowed his fundraising. The Freedom Club PAC’s treasurer testified that he has never been dissuaded from making an independent expenditure by the possibility of triggering matching funds. Arizona Taxpayers’s claim that it declined to speak in the 2006 primary for fear of triggering matching funds seems disingenuous in light of the fact that it only had $52.72 cash on hand. Plaintiffs have not demonstrated that any chilling effect exists. Their own experiences campaigning under the Act highlight that it “in no way limits the quantity of speech one can engage in or the amount of money one can spend engaging in political speech, nor does it threaten censure or penalty for such expenditures.” Daggett, 205 F.3d at 464.

Plaintiffs bemoan that matching funds deny them a competitive advantage in elections. The essence of this claim is not that they have been silenced, but that the speech of their opponents has been enabled. We agree with the First Circuit that the First Amendment includes “no right to speak free from response — the purpose of the First Amendment is to secure the widest possible dissemination of information from diverse and antagonistic sources.” Id. (internal quotations omitted).

Plaintiffs’ assertions that they have delayed making certain expenditures in order to avoid triggering matching funds to their opponents until later in the campaign cycle is evidence that they continue to seek strategic advantages under the Act, but does not support the argument that their speech has been chilled. Many campaign finance regulations, particularly disclosure requirements, lead candidates to engage in such strategic behavior, but this does not make them unconstitutional. As the Court has observed, “the public begins to concentrate on elections only in the weeks immediately before they are held.” Citizens United, 130 S.Ct. at 895. Consequently, waiting until that time to make campaign expenditures would not necessarily be evidence of coerced behavior. Rather, such decisions likely reflect a rational strategy of maximizing the impact of one’s campaign funds.

Although Plaintiffs cannot point to any specific instances in which their speech has been chilled because of the Act, we recognize that under the Supreme Court’s jurisprudence, even laws that create only potential chilling effects impose some First Amendment burden. For example, the Supreme Court has held that campaign contribution disclosure requirements and campaign advertisement funding disclosure and disclaimer requirements impose a minimal burden on speech and are therefore subject to intermediate scrutiny. See Buckley, 424 U.S. at 64, 96 S.Ct. 612 (disclosure requirements); Citizens United, 130 S.Ct. at 914 (disclosure and disclaimer requirements). In Buckley, the law at issue required candidates and PACs, as well as some other individuals or groups making political contributions or expenditures, to file disclosure reports with the FEC. 424 U.S. at 63, 96 S.Ct. 612. The Court recognized that these reporting requirements “can seriously infringe on pi'ivacy of association and belief guaranteed by the First Amendment.” Id. at 64, 96 S.Ct. 612. The Court also realized that the pub-lie disclosure of political contributions “will deter some individuals who otherwise might contribute. In some instances, disclosure may even expose contributors to harassment or retaliation.” Id. at 68, 96 S.Ct. 612. Even though these were “not insignificant burdens on individual rights,” the Court analyzed them under intermediate scrutiny rather than strict scrutiny because the disclosure laws “impose no ceiling on campaign-related activities.” Id. at 64, 68, 96 S.Ct. 612.

Recently, in Citizens United, the Court applied the same type of analysis to the BCRA’s requirement that political advertisements must state who “is responsible for the content of [certain] advertising.” 130 S.Ct. at 913-14. The Court reaffirmed Buckley’s reasoning that such requirements may impose a cognizable burden under the First Amendment even though they “do not prevent anyone from speaking.” Id. at 914.

The Court rejected the plaintiffs’ arguments that the requirement “decreases both the quantity and effectiveness of the group’s speech by forcing it to devote four seconds of each advertisement to the spoken disclaimer.” Id. at 915. Notably, in the absence of any evidence “of harassment or retaliation,” the Court also dismissed the plaintiffs’ arguments that the disclosure requirements would “chill donations to an organization.” Id. at 916.

In this case, as in Buckley and Citizens United, the burden that Plaintiffs allege is merely a theoretical chilling effect on donors who might dislike the statutory result of making a contribution or candidates who may seek a tactical advantage related to the release or timing of matching funds. The matching funds provision does not actually prevent anyone from speaking in the first place or cap campaign expenditures. Also, as in Buckley and Citizens United, there is no evidence that any Plaintiff has actually suffered the consequence they allege the Act imposes. We conclude that the burden created by the Act is most analogous to the burden of disclosure and disclaimer requirements in Buckley and Citizens United. Following the Supreme Court’s precedents in those cases, because the Act imposes only a minimal burden on fully protected speech, intermediate scrutiny applies.

Examining the matching funds provision under intermediate scrutiny, because there is a “ ‘substantial relation’ ” between the Act’s matching funds provision and a “ ‘sufficiently important’ governmental interest,” we conclude that it does not violate the First Amendment. See Citizens United, 130 S.Ct. at 914 (quoting Buckley). The State has a sufficiently important interest in preventing corruption and the appearance of corruption. See Citizens United, 130 S.Ct. at 909. The record demonstrates that Arizona has a long history of quid pro quo corruption. AzScam, in which legislators literally sold their votes for cash bribes, was just one of many substantial, wide-spread, and highly-publicized political scandals that Arizona experienced in the late 1980s and 1990s. These incidents occurred despite the contributions limits in place prior to the Act. Regardless of whether quid pro quo corruption continued to be a problem at the time of the Act’s passage, the appearance of quid pro quo corruption to the electorate was undeniable. Arizona voters were justified in concluding that contribution limits alone were not sufficient to combat corruption and its appearance. As the Supreme Court has recognized, the State’s interest in eradicating the appearance of quid pro quo corruption to restore the electorate’s confidence in its system of government is not “illusory,” it is substantial and compelling. See Buckley, 424 U.S. at 26-27, 96 S.Ct. 612.

Furthermore, the State has an interest in providing matching funds to encourage participation in its public funding scheme. Because Buckley held that public financing of elections furthers First Amendment values, federal courts have found that states may structure them in a manner which will encourage candidate participation in them. See, e.g., Rosenstiel, 101 F.3d at 1553 (“the State has a compelling interest in stimulating candidate participation in its public financing scheme”). The Eighth and First Circuits have found this interest to be so compelling as to withstand even the strictest scrutiny.¹⁰ See id.; Vote Choice, Inc. v. DiStefano, 4 F.3d 26, 39-40 (1st Cir.1993) (finding that the state has a “compelling” interest in “having candidates accept public financing”).

The district court misapprehended how the Act functions to reduce corruption. It assumed that the Act works by reducing nonparticipating candidates’ incentive to fundraise private contributions, thereby reducing the appearance of corruption among nonparticipating candidates. Thus, it concluded that the Act did not further an anticorruption interest by providing matching funds to participating candidates triggered by non-participating candidates making contributions to their own campaigns from their own private funds. In doing so, it relied on the Court’s holding that “discouraging use of personal funds[ ] disserves the anticorruption interest.” Davis, 128 S.Ct. at 2773.

The fact is, however, that the Act is aimed at reducing corruption among participating candidates. The relevant inquiry thus is whether matching funds bear a substantial relation to reducing corruption among participating candidates. In exchange for public funding, participating candidates relinquish their right to raise campaign contributions from private donors. They therefore have both reduced opportunities and reduced incentives to trade legislative favors for financial favors. The Supreme Court has held “[i]t cannot be gainsaid that public financing as a means of eliminating the improper influence of large private contributions furthers a significant governmental interest.” Buckley, 424 U.S. at 96, 96 S.Ct. 612.

Viewing the Act from this perspective, it is clear that the Act’s anticorruption interest is further promoted by high participation in the program. The more candidates that run with public funding, the smaller the appearance among Arizona elected officials of being susceptible to quid pro quo corruption, because fewer of those elected officials will have accepted a private campaign contribution and thus be viewed as beholden to their campaign contributors or as susceptible to such influence.

It is not relevant under this analysis what the source of a nonparticipating candidate’s campaign contributions is when he or she triggers matching funds. In order to promote participation in the program, and reduce the appearance of quid pro quo corruption, the State must be able to ensure that participating candidates will be able to mount competitive campaigns, no matter what the source of their opponent’s funding. If matching funds were not triggered by independent expenditures or expenditures from a nonparticipating candidate’s own funds, the Act’s public funding plan would not attract participants. It would be an enormous political risk to participate in such a system.

In this way, matching funds bear a substantial relation to the State’s anticorruption interest. A public financing system with no participants does nothing to reduce the existence or appearance of quid pro quo corruption. If participants were not given matching funds, they would not join the program because they would not be viable candidates in their elections. If the State were to adopt the district court’s recommendation of structuring matching funds by tying them “solely to contributions made by third parties to a candidate,” it would have the effect of deterring the participation of any candidate who feared she or he might have an opponent who was able to self-finance or would be the beneficiary of large independent expenditures. This would substantially diminish the Act’s ability to attract participants, thereby undermining its ability to prevent corruption.

In contrast, if the Act were to raise the amount of its lump-sum grants and do away with matching funds altogether, it would make the Act prohibitively expensive and spell its doom. By linking the amount of public funding in individual races to the amount of money being spent in these races, the State is able to allocate its funding among races of varying levels of competitiveness without having to make qualitative evaluations of which candidates are more “deserving” of funding beyond the base amounts provided to all publicly-funded candidates. The State must walk a fine line between providing too much and too little funding to participating candidates, and we cannot conclude that the Act’s matching funds provision has failed in this effort.

V. Conclusion

For the reasons set forth above, we conclude that the matching funds provision of the Act does not violate the First Amendment. We decline to reach the equal protection claim, and reverse and remand to the district court for further proceedings consistent with this opinion.

REVERSED and REMANDED.

1 . Governor Symington was sentenced to a 30-month term of imprisonment, but his conviction was reversed and the case remanded for a new trial. See United States v. Symington, 195 F.3d 1080 (9th Cir.1999). He was then granted a presidential pardon just as his retrial was set to commence.

2 . One of the Act’s original supporters testified that another one of its goals was to "promote freedom of speech because ... [under the Act,] more candidates would have more opportunity to speak."

3 . We recently recognized in Long Beach Area Chamber of Commerce v. City of Long Beach, that in light of Supreme Court precedent, " 'preventing corruption or the appearance of corruption are the only legitimate and compelling government interests thus far identified for restricting campaign finances.' ” Long Beach, 603 F.3d 684, 693-94 (9th Cir.2010) (quoting FEC v. Nat’l Conservative Political Action Comm., 470 U.S. 480, 496-97, 105 S.Ct. 1459, 84 L.Ed.2d 455 (1985)). In particular, we observed that the so-called "anti-distortion rationale," under which the government claims an "interest in combating 'the corrosive and distorting effects of immense aggregations of wealth,’ ” is not a legitimate basis for campaign finance restrictions after the Supreme Court partially overruled Austin v. Mich. Chamber of Commerce, 494 U.S. 652, 110 S.Ct. 1391, 108 L.Ed.2d 652 (1990), in its Citizens United decision. Id. at 693 (quoting Austin, 494 U.S. at 660, 110 S.Ct. 1391 and discussing Citizens United, 130 S.Ct. at 904, 912-13). Likewise, the "time protection rationale," under which the government claims an interest in " ‘protecting] candidates from spending too much time raising money rather than devoting that time to campaigning among ordinary voters' ” may not serve as the basis for restricting campaign finance activity. Id. at 693-94 (quoting Randall v. Sorrell, 548 U.S. 230, 243-45, 126 S.Ct. 2479, 165 L.Ed.2d 482 (2006)).

We need not, however, address the potential legitimacy of arguments that the State has not raised before us in defense of the Act. Further, even if the State had raised these arguments, "[i]t is unnecessary to look beyond the Act's primary purpose to limit the actuality and appearance of corruption resulting from large individual financial contributions in order to find a constitutionally sufficient justification” for the matching funds provision. Buckley, 424 U.S. at 26, 96 S.Ct. 612. Because the Act is justified by the State’s legitimate and non-illusory interest in reducing quid pro quo corruption, we need not consider the constitutional legitimacy of other potential rationales that might support it. See id. at 24-27, 96 S.Ct. 612.

4 . An independent expenditure is a campaign expenditure made by a third party that "expressly advocates the election or defeat of a clearly identified candidate, that is made without cooperation or consultation with any candidate or committee or agent of the candidate and that is not made in concert with or at the request or suggestion of a candidate, or any committee or agent of the candidate." Ariz.Rev.Stat. § 16-901(14).

5 . The matching funds that the participating candidate receives are reduced by six percent to take into account the fact that nonparticipaling candidates incur expenses in order to raise money. The assumption underlying this provision is that a nonparticipating candidate must spend six dollars (on postage, event costs, etc.) in order to raise one hundred dollars, while a participating candidate has no fundraising costs. Therefore, in order for both candidates to have the same net amount available to spend on non-fundraising campaign expenses, the matching funds that the participating candidate receives are reduced by six percent.

6 . Ken Bennett has succeeded Jan Brewer as Secretary of State since the commencement of this action.

7 . "Ordinarily a denial of a motion for summary judgment is not a final order and thus not appealable. 28 U.S.C. § 1291. However, the district court's grant of summary judgment was a final decision giving us jurisdiction to review its denial of [the State’s] motion for summary judgment.” Abend v. MCA, Inc., 863 F.2d 1465, 1482 n. 20 (9th Cir. 1988).

8 . We reject Plaintiffs' argument that Brown v. Cal. Dep’t of Transp., 321 F.3d 1217, 1221 (9th Cir.2003), requires us to review the district court's findings of fact for clear error because its opinion was one which struck down a restriction on speech. Brown, and the line of cases that it follows, apply to this Court’s review of cases in which the district court made findings of fact in the process of issuing a preliminary injunction or holding a bench trial, or when questions of fact were submitted to a jury. See Planned Parenthood v. Am. Coalition of Life Activists, 290 F.3d 1058, 1066-1070 (9th Cir.2002) (en banc). This case, arising from a grant of summary judgment, is in an entirely different procedural posture to which these cases do not apply.

9 . Prior to, and without the benefit of, the Supreme Court's recent decisions in Citizens United and Davis, three of our sister circuits considered the constitutionality of state public financing schemes with matching funds provisions similar to the one we now confront. See N.C. Right to Life Comm. Fund for Indep. Political Expenditures v. Leake, 524 F.3d 427 (4th Cir.2008); Daggett v. Comm’n on Gov'tal Ethics and Election Practices, 205 F.3d 445 (1st Cir.2000); Day v. Holahan, 34 F.3d 1356 (8th Cir.1994). The Fourth and First Circuits concluded that the respective statutes imposed no First Amendment burden at all. See N.C. Right to Life, 524 F.3d at 437-39; Daggett, 205 F.3d at 464-65. The Eighth Circuit, however, concluded that the Minnesota law at issue imposed a substantial burden on fully protected political speech and, because it could not survive strict scrutiny, struck it down. See Day, 34 F.3d at 1360, 1363.

Plaintiffs argue that because the Supreme Court’s decision in Davis cites Day, Davis somehow transformed the decision of one of our sister circuits into controlling precedent. Davis cites Day only once for a single, limited proposition. See Davis, 128 S.Ct. at 2772 ("Many candidates who can afford to make large personal expenditures to support their campaigns may choose to do so despite [the Millionaire’s Amendment], but they must shoulder a special and potentially significant burden if they make that choice. See Day v. Holahan, 34 F.3d 1356, 1359-60 (C.A.8 1994) (concluding that a Minnesota law that increased a candidate’s expenditure limits and eligibility for public funds based on independent expenditures against her candidacy burdened the speech of those making the independent expenditures)’’). In so citing Day, Davis did not affirm or adopt the Eighth Circuit’s approach, nor did it overturn sub silento the decisions of the First and Fourth Circuits. It merely cited Day for what it is — a decision in which our sister circuit found that a law chilled protected speech and held that such a burden could not survive strict scrutiny. Our decision is not controlled by Day, nor are we persuaded by it.

In Day, plaintiffs demonstrated that the Minnesota statute they were challenging had a "chilling effect” on political speech and that its “mere enactment ... ha[d] prevented many if not most potential political expenditures from ever being made.” 34 F.3d at 1360. As discussed infra, Plaintiffs have not established the existence of such a chilling effect in the instant case. Having found that the Minnesota law imposed a substantial burden on fully protected political speech, the Eighth Circuit proceeded to apply strict scrutiny. Id. at 1361. The court concluded that the law could not survive that test because it found that the state’s professed interest supporting the law, that of encouraging participation in the public financing program, was “not legitimate,” “no matter how compelling in the abstract,” because it was not based in fact, but was rather “contrived for purposes of [] litigation.” Id. The record there demonstrated that the challenged provision was “not necessary to encourage candidates' involvement in public campaign financing, as participation was approaching 100% before the new campaign finance laws were passed.” Id.

Two years after it decided Day, the Eighth Circuit, recognizing the unusual factual background of that case, declined to extend its reasoning and upheld other provisions of Minnesota's campaign finance law under strict scrutiny. See Rosenstiel v. Rodriguez, 101 F.3d 1544, 1555 (8th Cir.1996). In the case at bench, because the interest Arizona asserts in support of the Act is not "contrived for purposes of this litigation,” Day, 34 F.3d at 1361, even under current Eighth Circuit law "the circumstances surrounding the enactment of the [Act] make Day inapposite.” Rosenstiel, 101 F.3d at 1555. We decline to follow the Eighth Circuit down a road that even it refused to follow.

10 . Because we hold that the Act is subject only to intermediate scrutiny, we need not decide whether the Act would survive strict scrutiny, as the First and Fourth Circuits have held with respect to public financing schemes similar to the Act.

---

Concurrence Opinion

KLEINFELD, Circuit Judge,

concurring:

I concur. My reasoning is slightly different. In my view, the historical and policy considerations do not matter to the outcome. Although the complaint asserts both a facial and an as-applied challenge, the district court injunction and the arguments amount to a facial rejection of the statute.

The fact that matters is that the Arizona public financing scheme imposes no limitations whatsoever on a candidate’s speech. Davis v. Federal Election Commission¹ is easily distinguished because there the scheme did indeed impose a limit on the candidate’s speech, at least indirectly. If the candidate spent too much of his own money promoting his political position, then his opponent would benefit from a tripled contribution limit per donor.² The Arizona scheme does not manipulate the limits on private donors’ contributions according to whether a competing candidate is participating in the government funding scheme. Had it done so, Davis would apply by analogy.

Other circuits have divided on whether schemes like Arizona’s violate the First Amendment.³ The Supreme Court cited with apparent approval the Eighth Circuit decision,⁴ which may be contrary to the view we take today. But as the majority correctly says, it is not clear that the Supreme Court meant to adopt the Eighth Circuit view, and Davis did not require the Court to address a scheme like the one addressed by the Eighth Circuit.

I doubt that the level of scrutiny or risk of corruption matter in this case. The Arizona public financing scheme’s matching funds provision imposes no limit at all on contributions or expenditures for one’s own campaign. The limits are separate from the public financing scheme and not challenged. The plaintiffs do not challenge the contribution limits.⁵ Contribution limits are what invoke corruption and level of scrutiny considerations.

Plaintiffs challenge not contribution limits, but the benefits that the public financing scheme gives to participating candidates running against privately financed candidates and the strategic concerns the scheme creates for privately financed candidates. Because the challenged scheme imposes no contribution or spending limits, it does not restrict speech at all, so I cannot see why heightened scrutiny would apply.⁶ All that the evidence shows is that candidates who forego public funding make strategic decisions in response to the public funding scheme.⁷ That is not a restriction on speech. Intelligent, ambitious people seeking political office, or any other goal, are likely to study the rules and develop strategies taking maximum advantage of the rules. The kinds of strategic choices generated by the Arizona rules do not differ in kind from the choices presented to candidates by other election laws. For example, candidates will run their campaigns differently according to whether there is a fixed election day or an extended period for mail-in ballots, at large elections of multiple candidates in one district or single winner elections in multiple smaller districts, or partisan or nonpartisan elections. That different laws generate different strategies does not make them restrictions on speech.

The only speech-related concern I can see to the Arizona scheme is that a privately funded candidate has to raise a lot more money to swamp a publicly funded candidate. Someone not serious about politics, but serious about traveling, eating meals, and seeing his picture on billboards and in the newspapers for months, all at taxpayer expense, might pose a genuine challenge to a privately funded candidate, unless that candidate could raise funds well beyond the ceiling in the public funding scheme. If the notion is that campaign contributions amount to disguised bribery, I suppose this might mean the Arizona scheme would promote rather than inhibit corruption. But it is hard to see the materiality of that because contribution limits have not been challenged, and all the corruption theory has been used to justify in First Amendment law is contribution limits. As for the privately funded candidate, his or her speech is not limited by this increased burden of fundraising. Rather, his chances of winning are inhibited if he makes an incorrect judgment about whether he is going to be able to raise enough money to swamp the publicly funded candidate. But the First Amendment does not protect the candidate’s interest in winning, just his interest in being heard. There is no First Amendment right to make one’s opponent speak less,⁸ nor is there a First Amendment right to prohibit the government from subsidizing one’s opponent,⁹ especially when the same subsidy is available to the challenger if the challenger accepts the same terms as his opponent.¹⁰

As for the practical effects of the Arizona scheme, I have no idea whether they are good or bad, and nothing in the record throws any light on whether the good outweighs the bad or vice versa. Intent and motivation do not show effect, because the unintended effects of reform frequently outweigh the intended effects. Nor does it matter whether the scheme is a good or bad one, for First Amendment purposes. Arizona is entitled to make good or bad laws so long as they do not violate the Constitution. Since this law does not limit speech, it does not violate the First Amendment.

Perhaps public funding is good because it helps candidates run for office without relying on contributions from people who will want something in return. Perhaps it is bad because in a state with a large electorate and mass media, large contributors are the only people involved who know a candidate, and their money gives the public the benefit of what amounts to quality screening by people who actually know the candidate. As for the notion that contributions amount to bribery, at least when they are large, that may be so, or it may be that large contributions amount to protection money that those especially vulnerable to government power pay to those they fear will wield it. Perhaps both, or neither, is true. Doubtless many contributions result from contributors’ beliefs about who will provide good government, and gratitude to those willing to accept elected office and competent to discharge it well. None of this matters to this case. The only significance of the corruption concern to First Amendment law is that in Buckley v. Valeo, the concern was used to justify a limit on contributions to campaigns.¹¹ Since the case before us does not involve a contribution limit or any other burden on speech, the concern is immaterial.

In sum, the Arizona public financing scheme does not limit speech directly or indirectly. The only Supreme Court case touching at all closely upon the issues is Davis, and Davis has to be distinguished because the scheme in that case affected contribution limits and this scheme does not. No Ninth Circuit case speaks to the constitutionality of the Arizona scheme or anything like it. Our sister circuits are divided. Accordingly, I conclude that Arizona is not constitutionally barred from using the scheme before us.

1 . - U.S. -, 128 S.Ct. 2759, 171 L.Ed.2d 737 (2008).

2 . Id. at 2766.

3 . Compare N.C. Right to Life Comm. Fund for Indep. Political Expenditures v. Leake, 524 F.3d 427, 437 (4th Cir.2008) (holding that matching funds to publicly funded candidates that trigger based on contributions and ex-pendilures for privately funded candidates do not impose any burden on First Amendment rights) and Daggett v. Comm’n on Gov’tal Ethics and Election Practices, 205 F.3d 445, 464 (1st Cir.2000) (same) with Day v. Holahan, 34 F.3d 1356, 1360-62 (8th Cir.1994) (holding that providing additional public funds in response to independent expenditures advocating for a privately financed candidate burdens the speech of the independent advocacy groups, subjecting the law to strict scrutiny, and striking it down).

4 . Davis, 128 S.Ct. at 2772.

5 . Contribution limits are imposed by Arizona Revised Statutes § 16-905. Part of the referendum that created Arizona’s public funding scheme alters the contribution limits by providing new rules for participating candidates, § 16-941(A), and uniformly decreasing the contribution limits for nonparticipating candidates, § 16-941(B). The complaint does not allege that either § 16-905 or § 16-941 is unconstitutional. It alleges that § 16-952, the matching funds provision of the referendum, is unconstitutional. Section 16-952 does not impose contribution limits.

6 . Accord N.C. Right to Life, 524 F.3d at 437; Daggett, 205 F.3d at 464.

7 . N.C. Right to Life, 524 F.3d at 438; Daggett, 205 F.3d at 464.

8 . See Daggett, 205 F.3d at 464.

9 . Cf. Nat’l Endowment for the Arts v. Finley, 524 U.S. 569, 587-88, 118 S.Ct. 2168, 141 L.Ed.2d 500 (1998) (holding that the National Endowment for the Arts may decide who to subsidize based on aesthetic value of artworks because a subsidy is not a regulation on who can speak).

10 . Cf. Rosenberger v. Rector and Visitors of the Univ. of Va., 515 U.S. 819, 829-30, 115 S.Ct. 2510, 132 L.Ed.2d 700 (1995) (holding that a public university must make its printing facilities available to all qualifying student publications regardless of viewpoint).

11 . 424 U.S. 1, 25-26, 96 S.Ct. 612, 46 L.Ed.2d 659 (1976) (per curiam); see also Citizens United v. Fed. Election Comm'n, - U.S. -, 130 S.Ct. 876, 901, - L.Ed.2d -(2010).