Full opinion text
JOSÉ A. CABRANES, Circuit Judge: This is the first of two opinions in which we consider a constitutional challenge to certain provisions of Connecticut’s Campaign Finance Reform Act (CFRA). The CFRA, enacted in 2005, represents a comprehensive effort by the Connecticut General Assembly to change the way that campaigns for state office in Connecticut are financed. We consider here a challenge to the Citizens Election Program (CEP), a part of the CFRA that provides public money to candidates running for state office. In our second opinion, which we file separately, we consider a constitutional challenge to restrictions imposed by the CFRA on campaign contributions (and the solicitation of campaign contributions) by state contractors, lobbyists, and their families. See Green Party of Conn. v. Garfield, 616 F.3d 189 (2d Cir.2010). After a bench trial, the United States District Court for the District of Connecticut (Stefan R. Underhill, Judge) ruled, in part, that the CEP violated the First Amendment and the Equal Protection Clause of the Fourteenth Amendment by invidiously discriminating against so-called minor political parties and their candidates. See Green Party of Conn. v. Garfield, 648 F.Supp.2d 298 (D.Conn.2009) (“Green Party II ”). We reverse that part of the District Court’s judgment and hold that the CEP does not, on this record, invidiously discriminate against minor parties and their candidates. The District Court also ruled that certain discrete components of the CEP — its so-called “trigger provisions,” which include the CEP’s “excess expenditure provision” and “independent expenditure provision” — violate the First Amendment by impermissibly restricting the right of candidates and other individuals and organizations to spend their own funds on campaign speech. We affirm that part of the District Court’s judgment because we agree that the CEP’s trigger provisions violate the First Amendment. BACKGROUND We first describe the history of the CEP. We then outline its provisions and briefly recount the procedural history of this action. I. The History of the CEP The CFRA — which includes the CEP— was passed in response to several corruption scandals in Connecticut. Id. at 306— 07. The most widely publicized of the scandals involved Connecticut’s former governor, John Rowland. In 2004, Rowland was accused of accepting over $100,000 worth of gifts and services from state contractors, including vacations, flights on a private jet, and renovations to his lake cottage. Rowland accepted the gifts, it was alleged, in exchange for assisting the contractors in securing lucrative state contracts. Rowland resigned amidst the allegations, and in 2005 pleaded guilty — along with two aides and several contractors-to federal charges in connection with the scandal. Rowland was fined and sentenced to a year and a day in federal prison. See id. at 307. Sadly, the ignominy of public corruption was not limited to Rowland. As the District Court discussed in detail, the “Rowland scandal was but one of the many corruption scandals involving elected officials in state and local government that helped earn the state the nickname ‘Corrupticut.’ ” See id. at 307-08 (cataloging the scandals); see also id. at 307 n. 9 (discussing the decline of the reputation of Connecticut’s state government). It was in the wake of those scandals that Connecticut lawmakers resolved to enact “expansive campaign finance reforms.” Id. at 309. In the summer of 2005, Governor M. Jodi Rell established the Campaign Finance Reform Working Group (the “Working Group”), a collection of six state representatives and six state senators who were charged with drafting a new campaign finance reform law. After holding televised hearings for three months, the Working Group proposed an expansive bill, much of which would be incorporated into the final version of the CFRA. See id. at 309-10. In the fall of 2005, Governor Rell called a special session of the General Assembly for the sole purpose of considering the Working Group’s proposed bill. After a month of debate, the General Assembly passed the CFRA, and Governor Rell signed it into law. See id. at 300-11. As the District Court set forth in detail, several contemporaneous statements from General Assembly members, as well as Governor Rell, explain that the CFRA was passed “to combat actual and perceived corruption in state government.” Id. at 311. Much of the CFRA went into effect on January 1, 2006, but “2008 marked the first election cycle with candidates participating in the CEP public financing scheme.” Id. at 330; see also Conn. Gen. Stat. § 9-702(a) (providing that the CEP becomes effective for the legislative elections in 2008 and for the statewide elections in 2010). Before it went into effect, the CEP was twice amended. See Green Party II, 648 F.Supp.2d at 311, 319-20. II. The Provisions of the CEP The CEP is a complicated statutory scheme, see Conn. Gen.Stat. § 9-702 et seq., and the District Court took great care in explaining each of its provisions. See Green Party II, 648 F.Supp.2d. at 311-20. We describe only those provisions of the CEP that are relevant to our decision here. A. Qualification Criteria Candidates qualify for CEP funding by satisfying one of two types of qualifying criteria — one type for “major party” candidates and one type (with two subtypes) for “minor party” candidates. Under what we will refer to as the CEP’s “statewide qualifying criteria,” candidates qualify for CEP funding if they are running on the ticket of a major party. See Conn. Gen.Stat. § 9-702(a). A “major party” is defined by the CEP as a party that either (a) had a candidate for governor in the last election who received at least 20% of the vote, or (b) has as members at least 20% of the registered voters in the state. See id. § 9-372(5). There are, and have been for some time, only two parties that have achieved “major party” status in Connecticut: the Republican Party and the Democratic Party. Go-een Party II, 648 F.Supp.2d at 311. For candidates who are not running on the ticket of a major party- — -that is, for candidates who are running on the ticket of a minor party or who have no party affiliation — there are alternative ways of qualifying for CEP funding. Under what we will refer to as the CEP’s “single-election qualifying criteria,” a minor-party candidate can qualify for funding in a specific race if a member of his or her party achieved a certain threshold percentage of the vote in the same race in the last election. See Conn. Gen.Stat. § 9 — 705(c)(1), (g)(1). A minor-party candidate can qualify for a full grant of CEP funding if a member of his or her party received 20% of the vote in the same race in the last election; a candidate can qualify for two-thirds of the full amount if a member of his or her party received 15% of the vote in the same race in the last election; and a candidate can qualify for one-third of the full amount if a member of his or her party received 10% of the vote in the same race in the last election. See id. Under what we will refer to as the “petitioning criteria,” minor-party candidates can also qualify for CEP funding by collecting a certain number of signatures of those eligible to vote in the race in which they are running. A minor-party candidate can receive a full CEP grant if he or she collects a number of eligible signatures equal to 20% of the votes cast in the same race in the last election; the candidate can receive two-thirds of the full amount if he or she collects a number of eligible signatures equal to 15% of the votes cast in the same race in the last election; and the candidate can receive one-third of the full amount if he or she collects a number of eligible signatures equal to 10% of the votes cast in the same race in the last election. See id. § 9 — 705(c)(2), (g)(2). Finally, all candidates — whether they qualify under the statewide criteria, the single-election criteria, or the petitioning criteria — must raise a specified amount of money through small “qualifying contributions” of $100 or less. See id. § 9-704. The required amount that candidates must raise varies depending on the office sought: candidates for governor, for instance, must raise $250,000 in qualifying contributions, whereas candidates for state representative must raise $5,000 in qualifying contributions. Id. § 9-704(a)(l), (4). Otherwise-qualified candidates do not receive CEP funding until they have raised the required qualifying contributions. B. Distribution Formulae Once a candidate qualifies for public funds under the CEP, the amount of public money that he or she receives is determined by the CEP’s “distribution formulae.” 1. Primary Election Grants Candidates seeking the endorsement of a major party must run in primary elections that are governed by state law. Those candidates receive CEP funding for the primary election in the following amounts: candidates for governor receive $1.25 million; candidates for other statewide offices receive $375,000; candidates for the state senate receive $35,000; and candidates for the state house of representatives receive $10,000. Id. § 9-705(a)(l), (b)(1), (e)(1), (f)(1). Like all CEP grants, those amounts will, in the future, be adjusted for inflation. Id. § 9 — 705(d), (h). A candidate running for the General Assembly receives more money for the primary election if the election takes place in a district that is considered “one-party dominant” and the candidate is a member of the “dominant” party. (As discussed in greater detail below, we will also refer to “one-party dominant” districts as “safe” districts.) A “one-party dominant” district is defined as a district in which there is a difference of twenty percentage points or more between the number of registered voters for the two major parties. For example, if 55% of the voters in a district were registered Democrats and 35% of the voters were registered Republicans (with 10% unaffiliated or registered with a minor party), there would be a twenty-percentage-point difference in the number of Democratic and Republican voters, and the candidates running in the Democratic primary would receive extra money: the grant for the Democratic candidate for the state senate would increase to $75,000, and the grant for the Democratic candidate for the state house of representatives would increase to $25,000. See id. § 9-705(e)(1)(A), (f)(1)(A). Currently, no minor party in Connecticut selects its candidates by means of primary elections, but defendants contend that, if a minor party were to hold primary elections, that party’s candidates would be eligible for CEP funding. See Green Party II, 648 F.Supp.2d at 312 n. 16. 2. General Election Grants For the general election, the CEP provides the following “full” grants: candidates for governor receive $3 million; candidates for other statewide offices receive $750,000; candidates for the Connecticut Senate receive $85,000; and candidates for the Connecticut House of Representatives receive $25,000. See Conn. GemStat. § 9-705(a)(2), (b)(2), (e)(2), (f)(2). Those full grants may be reduced in certain circumstances. For instance, if a major-party candidate is running unopposed, the CEP grant is reduced to 30% of the full amount. See id. § 9 — 705(j)(3). If a major-party candidate has no major-party competitor but is running against a minor-party candidate who has not qualified for (or accepted) CEP funding, the major-party candidate receives 60% of the full amount. See id. § 9-705(j)(4). If a major-party candidate is running against a minor-party candidate who has, in fact, qualified for CEP funding (or if the minor-party candidate has raised or spent nonpublic funds equal to the amount of funding the candidate would have received under the CEP), the major-party candidate receives the full grant. See id. C. Expenditure Limits By participating in the CEP and accepting public funds, candidates agree to accept certain limits on the total amount of money they may spend on their campaigns. In essence, candidates that participate in the CEP may spend only the amount they receive in public funds, plus the amount they raise through the required “qualifying contributions.” See Conn. GemStat. § 9-702(c). Participating candidates are also permitted to spend a small amount of their own personal funds in certain circumstances. See id. §§ 9-702(c), 9-710(c). D. Trigger Provisions Finally, under the CEP’s so-called “trigger provisions,” candidates receive additional funding when certain conditions are triggered. There are two trigger provisions: the “excess expenditure” provision and the “independent expenditure” provision. The District Court concisely explained the excess expenditure provision: The CEP provides matching funds for participating candidates who are outspent by a non-participating opponent' — ■ who is not bound by any expenditure limit — in the primary or the general election (“excess expenditure trigger”). Conn. Gen.Stat. § 9-713. If a non-participating candidate receives contributions or spends more than an amount equal to the participating candidate’s expenditure limit, then the participating candidate is eligible to receive up to four additional grants, each worth 25% of the full grant. Id. The excess expenditure grants are distributed whenever the non-participating candidate receives contributions or makes expenditures exceeding 100%, 125%, 150%, and 175% of the expenditure limit for that particular office. Green Party II, 648 F.Supp.2d at 315-16. The independent expenditure provision is similar to the excess expenditure provision, but it applies when non-candidate individuals and organizations make independent expenditures advocating against the election of a candidate. Again, the District Court concisely explained this provision: The CEP also contains a trigger provision tied to independent expenditures made by non-candidate individuals and political advocacy groups.... Conn. Gen.Stat. § 9-714. A qualifying independent expenditure is “an expenditure that is made without the consent, knowing participation, or consultation of, a candidate or agent of the candidate committee and is not a coordinated expenditure,” id. § 9-601(18), and that is made “with the intent to promote the defeat of a participating candidate.” Id. § 9-714(a). Matching funds under this provision are triggered when non-candidate individuals or groups make independent expenditures advocating the defeat of a participating candidate, that in the aggregate, and when combined with the spending of the opposing non-participating candidates in that race, exceed the CEP grant amount. Id. § 9-714(c)(2). Funds are distributed to the participating candidate on a dollar-per-dollar basis to match the amount of the independent expenditure(s) in excess of the full grant amount. Id. § 9-714(a). Notably, independent expenditures made in support of a candidate (without expressly advocating the defeat of an opponent) do not count towards the independent expenditure trigger, meaning individuals and groups are entitled to make unlimited independent expenditures in support of a candidate without triggering CEP matching funds for that candidate’s opponents. See generally id. § 9 — 714[.] Id. at 316. III. This Action Plaintiffs-appellees (“plaintiffs”) brought this action in 2006 claiming that certain provisions of the CFRA (including the CEP) violated the First and Fourteenth Amendments to the United States Constitution. A. The Parties Plaintiffs include two minor parties operating in Connecticut: the Green Party of Connecticut and the Libertarian Party of Connecticut. Plaintiffs also include several Connecticut-based lobbyists and state contractors, as well as Michael DeRosa, a member of the Green Party who has run, in the past, for the state senate and for Secretary of the State on the Green Party ticket. See Green Party II, 648 F.Supp.2d at 302-06; J.A. 49-52 (Compl.n 10-17). Defendants-appellants (“defendants”) include Jeffrey Garfield, who is named in his official capacity as the Executive Director and General Counsel of the State Elections Enforcement Commission, and Richard Blumenthal, who is named in his official capacity as the Attorney General of the State of Connecticut. See Green Party II, 648 F.Supp.2d at 306; J.A. 52 (Compl.M 18-19). The parties in this action also include several individuals and entities who successfully moved to intervene as defendants. The intervenor-defendants-appellants include three former major-party candidates for state office and two advocacy groups: Connecticut Common Cause and Connecticut Citizens Action Group. See Green Party II, 648 F.Supp.2d at 306. The intervenor-defendants defend the constitutionality of the CEP. B. The Claims Plaintiffs have organized their claims into five counts. In Count One, plaintiffs claim that the CEP’s qualification criteria and distribution formulae, Conn. GemStat. §§ 9-702(b), 704-05, violate the First Amendment and the Equal Protection Clause of the Fourteenth Amendment by invidiously “discriminating]” against minor parties and their candidates. See J.A. 66 (Comply 53). In Counts Two and Three, plaintiffs assert First Amendment challenges to the CEP’s excess expenditure provision, Conn. Gen.Stat. § 9-713 (Count Two), and the CEP’s independent expenditure provision, id. § 9-714 (Count Three). See J.A. 66-67 (Compilé 54-55). In Counts Four and Five, plaintiffs assert First Amendment challenges to aspects of the CFRA that do not involve the CEP. In Count Four, plaintiffs challenge the CFRA’s bans on contributions (and the solicitation of contributions) by state contractors, lobbyists, and their families. Conn. Gen.Stat. §§ 9-610(g)-(h), 9-612(g). In Count Five, plaintiffs challenge disclosure requirements imposed by the CFRA on state contractors. Id. § 9 — 612(h)(2); see J.A. 67 (Compile 56-57). This opinion addresses Counts One, Two, and Three. Our second, separately filed opinion addresses Count Four. Plaintiffs have not pursued Count Five in these appeals; thus we do not address it. C. Proceedings in the District Court The District Court disposed of plaintiffs’ claims by means of two separate judgments. The District Court first granted summary judgment for defendants on Count Four, holding that the CFRA’s contribution and solicitation bans did not violate the First Amendment. See Green Party of Conn. v. Garfield, 590 F.Supp.2d 288 (D.Conn.2008) (“Green Party I”). On February 11, 2009, the District Court entered a partial final judgment for defendants with respect to Count Four. See Fed.R.Civ.P. 54(b). Plaintiffs filed a timely appeal of that partial final judgment (2d Cir. Docket No. 09-0599-cv(L)), which we address in our separately filed opinion. The District Court then held a bench trial and, at the end of the trial, granted judgment to plaintiffs on the remaining counts — Counts One, Two, and Three. See Green Party II, 648 F.Supp.2d 298. With respect to Count One, the District Court determined that “the CEP impose[d] an unconstitutional, discriminatory burden on minor party candidates’ First Amendment-protected right to political opportunity.” Id. at 300. With respect to Counts Two and Three, the District Court “conclude[d] that the CEP’s excess expenditure and independent expenditure provisions ... unconstitutionally burden[ed] the plaintiffs’ exercise of their First Amendment rights.” Id. at 302. Accordingly, in a September 9, 2009 final judgment, the District Court declared the CEP unconstitutional and entered a permanent injunction prohibiting defendants from enforcing each of the CEP’s provisions. See id. at 374. The District Court then stayed the injunction pending this appeal. See Green Party of Conn. v. Garfield, No. 3:06-cv-01030, Docket Entry No. 399 (D.Conn. Sept. 29, 2009). Defendants filed a timely appeal of the District Court’s September 9, 2009 judgment on Counts One, Two, and Three, and we address that appeal in this opinion. DISCUSSION “We review the district court’s findings of fact after a bench trial for clear error and its conclusions of law de novo.” Arch Ins. Co. v. Precision Stone, Inc., 584 F.3d 33, 38-39 (2d Cir.2009) (quotation marks omitted). There were, in this case, very few factual disputes for the District Court to resolve at trial. Instead, much of the record in this case consisted of undisputed facts, and in any event, nearly all of the District Court’s assessment of plaintiffs’ claims involved either pure issues of law or the “application of ... facts to draw conclusions of law.” Scribner v. Summers, 84 F.3d 554, 557 (2d Cir.1996). We therefore review much of the District Court’s analysis de novo. See id. (“The district court’s application of ... facts to draw conclusions of law ... is subject to de novo appellate review.” (citing Travellers Int'l AG. v. Trans World Airlines, Inc., 41 F.3d 1570, 1575 (2d Cir.1994))); see also Bose Corp. v. Consumers Union, 466 U.S. 485, 501, 104 S.Ct. 1949, 80 L.Ed.2d 502 (1984); In re Complaint of Messina, 574 F.3d 119, 128 (2d Cir.2009); Davis v. N.Y. C. Hous. Auth., 278 F.3d 64, 79 (2d Cir.2002). COUNT ONE: Whether the CEP Unconstitutionally Discriminates Against Minor-Party Candidates In Count One, plaintiffs claim that the CEP violates the First Amendment and the Equal Protection Clause of the Fourteenth Amendment by invidiously “discriminat[ing]” against minor-party candidates. See J.A. 66 (ComplJ 53). Plaintiffs’ challenge is focused on the CEP’s “qualification criteria,” which are the criteria by which candidates qualify to receive CEP funding, as well as the CEP’s “distribution formulae,” which are the formulae that establish the amount of money that the CEP provides to participating candidates. See id. According to plaintiffs, the CEP’s qualifying criteria and distribution formulae violate the Constitution because they impermissibly burden the “political opportunity” of minor-party candidates. See Buckley v. Valeo, 424 U.S. 1, 95-96, 96 S.Ct. 612, 46 L.Ed.2d 659 (1976) (holding that a public financing system may violate equal protection if it “unfairly or unnecessarily burden[s] the political opportunity of any party or candidate”). The District Court granted judgment for plaintiffs on Count One. The Court determined that the CEP imposed “a severe, discriminatory burden on the political opportunity of minor party candidates,” and it held that “despite presenting compelling government interests, the state ha[d] failed to demonstrate how the CEP [was] narrowly tailored to advance those government interests.” Green Party II, 648 F.Supp.2d at 361-62. In our view, the District Court erred in its judgment for plaintiffs on Count One. We conclude that the Connecticut General Assembly enacted the CEP “in furtherance of sufficiently important governmental interests,” and we hold that the CEP’s qualification criteria and distribution formulae do not, on this record, “unfairly or unnecessarily burden[ ] the political opportunity of any party or candidate.” Buckley, 424 U.S. at 95-96, 96 S.Ct. 612. We therefore reverse the District Court on Count One and grant judgment for defendants. I. The Legal Standard for Plaintiffs’ Claim of Unconstitutional Discrimination In determining the legal standard to apply to Count One, we hew to the Supreme Court’s analysis in Buckley v. Valeo, which is the principal binding precedent addressing whether a system of public financing for elections unconstitutionally discriminates against minor-party candidates. Buckley considered, in part, a 1970 federal statute that created a system of public financing for presidential election campaigns. See Buckley, 424 U.S. at 85, 96 S.Ct. 612. Several individuals and entities, including minor parties and prospective candidates, see id. at 7-8, 96 S.Ct. 612, challenged the law. They claimed, among other things, that it violated the First Amendment and the equal protection component of the Due Process Clause of the Fifth Amendment by “discriminating” against minor-party candidates. See id. at 93, 96 S.Ct. 612. As we set forth in greater detail below, the CEP differs in some ways from the presidential-candidate financing system at issue in Buckley, and our analysis must account for those differences. We are, nonetheless, compelled to apply the legal standard articulated in Buckley, as that case addressed exactly the type of claim raised in Count One: a challenge to a public financing system on the ground that it unconstitutionally “discriminates” against minor-party candidates. We acknowledge that another Supreme Court decision, issued after Buckley, ruled on a similar challenge to a public financing system. See Bang v. Chase, 442 F.Supp. 758 (D.Minn.1977), summarily aff'd sub nom., 436 U.S. 941, 98 S.Ct. 2840, 56 L.Ed.2d 782 (1978). That ruling, however, was a summary affirmance of a district court judgment and therefore provides little guidance. As the Supreme Court clarified a year before issuing its summary affirmance in Bang, an “unexplicated summary affirmance settles the issues for the parties, and is not to be read as a renunciation by th[e] Court of doctrines previously announced in [its] opinions after full argument.” Mandel v. Bradley, 432 U.S. 173, 176, 97 S.Ct. 2238, 53 L.Ed.2d 199 (1977) (quotation marks omitted). We therefore heed the Court’s warning that “ ‘[ascertaining the reach and content of summary actions may itself present issues of real substance,’ ” id. (quoting Hicks v. Miranda, 422 U.S. 332, 345 n. 14, 95 S.Ct. 2281, 45 L.Ed.2d 223 (1975)), and we do not attempt to divine whether the Supreme Court adopted the district court’s reasoning in Bang or whether the Court affirmed on an entirely different rationale. See, e.g., Bush v. Vera, 517 U.S. 952, 996, 116 S.Ct. 1941, 135 L.Ed.2d 248 (1996) (Kennedy, J., concurring) (“We do not endorse the reasoning of the district court when we order summary affirmance of the judgment.” (emphasis added)); Mandel, 432 U.S. at 176, 97 S.Ct. 2238 (“Because a summary affirmance is an affirmance of the judgment only, the rationale of the affirmance may not be gleaned solely from the opinion below.” (emphasis added)); see also Morse v. Republican Party, 517 U.S. 186, n. 21, 116 S.Ct. 1186, 134 L.Ed.2d 347 (1996) (“We ... note that a summary affirmance by this Court is a ‘rather slender reed’ on which to rest future decisions.” (quoting Anderson v. Celebrezze, 460 U.S. 780, 784-85 n. 5, 103 S.Ct. 1564, 75 L.Ed.2d 547 (1983))). We also decline plaintiffs’ invitation to look for guidance from Davis v. Federal Election Commission, — U.S. —, 128 S.Ct. 2759, 171 L.Ed.2d 737 (2008), in deciding plaintiffs’ discrimination claim in Count One. Davis involved an entirely different claim: the Davis plaintiffs challenged the so-called “Millionaire’s Amendment,” which imposed a “penalty”- — -in the form of a disadvantageous “asymmetrical regulatory scheme” — on candidates for Congress who spent large amounts of their own money on their campaigns. Id. at 2766, 2771. Davis accordingly addressed a law that burdened the “fundamental” First Amendment right to spend one’s own money on one’s own campaign. See id. at 2771 (“[W]e agree with Davis that this scheme impermissibly burdens his First Amendment right to spend his own money for campaign speech.”); see also id. (recognizing “the fundamental nature of the right to spend personal funds for campaign speech”). Putting aside the CEP’s trigger provisions, which we address below in connection with Counts Two and Three, the CEP does not impose a penalty on a candidate who spends his or her own money on a campaign, for in every race candidates can decline to participate in the CEP. See id. at 2772 (“[T]he choice involved in Buckley was quite different from the choice imposed by [the Millionaire’s Amendment]. In Buckley, a candidate, by forgoing public financing, could retain the unfettered right to make unlimited personal expenditures.... The choice imposed by [the Millionaire’s Amendment] is not remotely parallel to that in Buckley.”). Davis, therefore, is inapposite. In any event, Davis in no way suggested that it was overruling Buckley. Yet if Davis’s analysis were applied here, it could not be reconciled with Buckley. As we discuss in greater detail below, Buckley placed the burden on the plaintiffs to “show[ ] that the election funding plan disadvantage[d] nonmajor parties by operating to reduce their strength below that attained without any public financing.” 424 U.S. at 98-99, 96 S.Ct. 612 (emphasis added). Davis, on the other hand, put the burden on the government to defend the statute in question. See 128 S.Ct. at 2772-74. Buckley, moreover, required that the presidential-campaign financing system be justified by a “sufficiently important” state interest, 424 U.S. at 95-96, 96 S.Ct. 612; see note 7, post; whereas Davis applied a more searching standard and required that the Millionaire’s Amendment be justified by a “compelling state interest,” 128 S.Ct. at 2772 (quotation marks omitted). Because Buckley, not Davis, addressed the same type of claim as the one raised here, and because there is no indication that Davis was meant to overrule Buckley’s analysis of the presidential-campaign financing system (even sub silentio), we look to Buckley for the legal standard by which to assess plaintiffs’ claim of unconstitutional discrimination in Count One. We therefore closely examine the legal standards applied in Buckley, and we describe how the District Court did — and did not — apply those standards correctly. A. The Standard Set Forth in Buckley v. Valeo Buckley first determined that the public financing system 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.” 424 U.S. at 92-93, 96 S.Ct. 612. Accordingly, Buckley rejected the plaintiffs’ First Amendment challenge out of hand, holding that the presidential-candidate financing system only “further[ed],” and did “not abridge[],” the “pertinent First Amendment values.” Id. Turning to the discrimination claim— that is, the claim that the presidential-candidate financing system violated the requirement of equal protection of the laws in its differential treatment of minor-party candidates and major-party candidates— Buckley initially questioned whether the “exacting scrutiny” standard should apply to the system. Id. at 93-94, 96 S.Ct. 612. Buckley cited several precedents and observed that, at the time, a “principle ha[d] been developed that restrictions on access to the electoral process must survive exacting scrutiny.” Id. Yet Buckley distinguished those precedents, finding them inapplicable because they “dealt primarily with state laws requiring a candidate to satisfy certain requirements in order to have his name appear on the ballot.” Id. at 94, 96 S.Ct. 612. Such laws, Buckley reasoned, were “direct burdens not only on the candidate’s ability to run for office but also on the voter’s ability to voice preferences regarding representative government and contemporary issues.” Id. A public financing system, in contrast, was “not restrictive of voters’ rights and less restrictive of candidates’ [rights],” because the system did “not prevent any candidate from getting on the ballot or any voter from casting a vote for the candidate of his choice.” Id. As a result, Buckley determined that “public financing is generally less restrictive of access to the electoral process than the ballot-access regulations dealt with in prior cases.” Id. at 95, 96 S.Ct. 612 (emphasis added). Buckley did not, however, complete that line of reasoning and establish a less searching standard for equal protection challenges to public financing systems. Instead, Buckley determined that the presidential-candidate financing system could be upheld even assuming, for the sake of analysis, that the correct standard was “exacting scrutiny.” After distinguishing the precedents that had applied the “exacting scrutiny” standard, Buckley held: “In any event, Congress enacted [the presidential-candidate financing system] in furtherance of sufficiently important governmental interests and has not unfairly or unnecessarily burdened the political opportunity of any party or candidate.” Id. at 95-96, 96 S.Ct. 612 (emphasis added). Following Buckley, therefore, the starting point for a court in determining whether a public financing system unconstitutionally discriminates against minor parties is to assume, for the sake of analysis, that the correct standard is the version of “exacting scrutiny” articulated in Buckley. Under that standard, a court must first examine whether the system was “enacted ... in furtherance of sufficiently important governmental interests.” Id. at 95, 96 S.Ct. 612. The court must then determine whether the system “burden[s] the political opportunity of any party or candidate” in a way that is “unfairf]” or “unneeessar[y].” Id. at 96, 96 S.Ct. 612. If the public financing system fares favorably under that two pronged test, the inquiry is over-the system does not violate the Constitution. If, however, the public financing system fails under Buckley’s version of the “exacting scrutiny” standard' — that is, if the system furthers insufficiently important governmental interests, or if the system does, in fact, burden the political opportunity of a party or candidate in a way that is unnecessary or unfair — then the court must proceed to a second step of the inquiry: the court must finish the line of reasoning that Buckley left unresolved and determine whether a less searching standard applies. Here, in evaluating plaintiffs’ claim in Count One, we are not required to perform that second step of the inquiry because, as we set forth in greater detail below, we, like the Supreme Court in Buckley, reject plaintiffs’ claim of unconstitutional discrimination even applying Buckley’s version of “exacting scrutiny.” Nonetheless, we conclude that if, in another case, a court determines that a public financing system cannot withstand Buckley’s version of “exacting scrutiny,” the court must proceed to the second step of the inquiry, finish the line of reasoning that Buckley left unresolved, and determine whether a less searching standard applies. In sum, when a plaintiff claims that a public financing system violates the First Amendment and the Equal Protection Clause in its differential treatment of minor-party candidates and major-party candidates, a court should employ the following analysis: The court should first assume that Buckley’s version of “exacting scrutiny” applies and determine (a) whether the system was enacted in furtherance of a sufficiently important governmental interest and (b) whether the system burdens the political opportunity of a party or candidate in a way that is unfair or unnecessary. If the system fails under Buckley’s version of the “exacting scrutiny” standard, the court should then complete Buckley’s unresolved line of reasoning and determine whether a less searching standard applies. If the court determines that a less searching standard applies, the court should then evaluate the public financing system under that less searching standard. B. The District Court’s Erroneous Application of Strict Scrutiny Before proceeding to the merits of plaintiffs’ discrimination claim, we must clarify that the District Court erred in applying strict scrutiny to evaluate plaintiffs’ claim. The District Court began its analysis by applying the correct legal standard, as it first examined, at length, whether the CEP “ ‘unfairly or unnecessarily bur-dented] the political opportunity of any party or candidate’” — that, of course, is one part of Buckley’s version of the “exacting scrutiny” standard. Green Party II, 648 F.Supp.2d at 333-34 (quoting Buckley, 424 U.S. at 96, 96 S.Ct. 612). Ultimately, the District Court concluded that the CEP did, in fact, impermissibly burden the political opportunity of minor-party candidates. That is a legal conclusion that we reverse, as set forth below. Nonetheless, assuming, for the sake of analysis, that the District Court was correct to hold that the CEP impermissibly burdened the political opportunity of minor-party candidates, the Court was, at that point, required to proceed to a second step of the inquiry — to determine whether a less searching standard applied in evaluating plaintiffs’ discrimination claim. Yet the District Court did exactly the opposite: it held that “strict scrutiny” — a more searching standard — applied in evaluating plaintiffs’ discrimination claim. In applying strict scrutiny, the District Court relied on two cases from our sister Circuits, Daggett v. Commission on Governmental Ethics & Election Practices, 205 F.3d 445, 466 (1st Cir.2000), and Rosenstiel v. Rodriguez, 101 F.3d 1544, 1553 (8th Cir.1996). In those cases, candidates claimed that a state public financing system violated the First Amendment because it was overly “coercive,” effectively requiring that every candidate accept public money. The courts applied strict scrutiny because they concluded that the right to decline public funds — and to raise and spend one’s own money in an election campaign — was a “fundamental” right protected by the First Amendment. We have no occasion to address whether strict scrutiny was the correct standard to evaluate the claims raised in Daggett and Rosenstiel. We note only that the claims raised in those cases were far different from the claim raised by plaintiffs in Count One: the plaintiffs in Daggett and Rosenstiel claimed that a public financing system was overly “coercive” and thereby violated the First Amendment, whereas plaintiffs here claim that a public financing system unconstitutionally discriminates in its differential treatment of minor-party candidates and major-party candidates. The District Court’s reliance on Daggett and Rosenstiel was, therefore, misplaced. It is, instead, Buckley that provides the best guidance in this context, as Buckley addressed the same type of claim that plaintiffs raise in Count One. Again, as we have explained, in no event does Buckley suggest that “strict scrutiny” — a standard that is more demanding than “exacting scrutiny” — applies to the type of claim raised in Count One. In sum, the District Court erred in applying strict scrutiny. II. The Merits of Plaintiffs’ Claim of Unconstitutional Discrimination Having clarified the legal standard with which to evaluate plaintiffs’ claim of unconstitutional discrimination in Count One, we now turn to the merits of that claim. As explained above, we will follow Buckley’s example and assume for the sake of analysis that Buckley’s version of “exacting scrutiny” applies. Thus we ask (a) whether the CEP was enacted in furtherance of a sufficiently important governmental interest and (b) whether the CEP burdens the political opportunity of a party or candidate in a way that is unfair or unnecessary. See Buckley, 424 U.S. at 95-96, 96 S.Ct. 612. The answer to the first question— whether the CEP furthers a sufficiently important governmental interest — is straightforward. As Buckley held, “public financing as a means of eliminating improper influence of large private contributions furthers a significant governmental interest.” Id. at 96, 96 S.Ct. 612. The District Court found that the CEP was enacted in furtherance of several goals, including to eliminate improper influence on elected officials. See Green Party II, 648 F.Supp.2d at 309 (explaining that the CEP was “[sjpurred in large part by the fall-out from the corruption scandals that culminated in the resignation of Governor Rowland and his subsequent indictment and conviction”). Accordingly, the District Court held that the CEP was enacted to further a sufficiently important governmental interest. See id. at 351. We agree with that holding. The answer to the second question— whether the system burdens the political opportunity of a candidate in a way that is unfair or unnecessary — is more complicated. Plaintiffs claim, primarily, that three aspects of the CEP impermissibly burden their political opportunity: (1) the CEP’s single-election qualification criteria, (2) the CEP’s statewide qualification criteria, and (3) the CEP’s distribution formulae. We address each aspect of the CEP in turn. A. The Single-Election Qualification Criteria The District Court determined that the CEP’s single-election qualification criteria, see generally Conn. Gen.Stat. § 9-705, impermissibly burdened the political opportunity of minor-party candidates because the criteria “ma[de] it extremely difficult for minor party candidates to become eligible for even partial public funding,” Green Party II, 648 F.Supp.2d at 344. We cannot agree with that application of law to fact. 1. The CEP May Condition Public Funds on a Showing of Popular Support in the Previous Election As an initial matter, Buckley held that a public financing system may condition a grant of public money on a showing that the candidate already enjoys a certain threshold level of popular support. The reason is twofold: First, the government has an “interest in not funding hopeless candidacies with large sums of public money,” and that interest “necessarily justifies the withholding of public assistance from candidates without significant public support.” Buckley, 424 U.S. at 96, 96 S.Ct. 612 (citation omitted). Thus the “Constitution does not require the Government to finance the efforts of every nascent political group,” for “[sjometimes 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 (quotation marks omitted); see also id. at 97, 96 S.Ct. 612 (“[T]he Constitution does not require Congress to treat all declared candidates the same for public financing purposes ... [as] there are obvious differences in kind between the needs and potentials of a political party with historically established broad support, on the one hand, and a new or small political organization on the other.” (citation and quotation marks omitted)). In other words, Buckley recognized that if the Constitution were to require the presidential-candidate financing system to fund every minor-party candidate, the Constitution would provide the means for fly-by-night candidates to “raid the United States Treasury.” Id. at 98, 96 S.Ct. 612 (quotation marks omitted). The second reason that a public financing system may condition public money on a showing of popular support is that limiting an election to a small number of strong candidates “serves the important public interest against providing artificial incentives to splintered parties and unrestrained factionalism.” Id. at 96, 96 S.Ct. 612 (quotation marks omitted). That is, to fund every minor-party candidate would risk a fractured and chaotic election, “artificially fosterling] the proliferation of splinter parties.” Id. at 98, 96 S.Ct. 612 (quotation marks omitted). Accordingly, the CEP may, consistent with the First Amendment and the Equal Protection Clause, distinguish between candidates who can, and who cannot, make a preliminary showing of public support, providing funds to those who can and withholding funds from those who cannot. In addition, “popular vote totals in the last election are a proper measure of public support.” Id. at 99, 96 S.Ct. 612 (citing Jenness v. Fortson, 403 U.S. 431, 439-40, 91 S.Ct. 1970, 29 L.Ed.2d 554 (1971)). The CEP’s use of vote totals from the previous election, therefore, is a permissible way to distinguish between candidates who do and do not enjoy the required threshold level of popularity. 2. We Draw from Buckley Four Principles to Evaluate Whether the CEP’s Qualification Criteria Impose a Burden on Political Opportunity That Is “Unfair” or “Unnecessary” Having established that it is permissible for the CEP to condition public funding on a preliminary showing of public support— and that the CEP may use prior vote totals to measure that support — we now ask whether the CEP’s single election qualification criteria of 10%, 15%, and 20% of the vote in the past election are set so high as to burden the political opportunity of a party or candidate in a way that is unfair or unnecessary. Although Buckley did not expressly define “unfair” or “unnecessary,” we draw from Buckley’s analysis four principles that illuminate what Buckley meant by those terms. (i) A public financing system may condition public funds on a threshold level of public support that is relatively high, as Buckley held that a public financing system may “require ‘some preliminary showing of a significant modicum of support’ as an eligibility requirement for public funds.” Buckley, 424 U.S. at 96, 96 S.Ct. 612 (quoting Jenness, 403 U.S. at 442, 91 S.Ct. 1970) (emphasis added, citation omitted); see also id. at 96, 96 S.Ct. 612 (concluding that the government’s “interest in not funding hopeless candidacies ... necessarily justifies the withholding of public assistance from candidates without significant public support” (emphasis added)); cf. id. (noting that the government has been “held to have important interests in limiting places on the ballot to those candidates who demonstrate substantial popular support” (emphasis added)). In other words, a public financing system need not provide funding to every candidate who can demonstrate some public support; the system may, instead, condition public money on a preliminary showing of “significant” public support. That is not to say, of course, that any threshold would pass constitutional muster. Buckley established an important role for courts in evaluating whether public financing systems unconstitutionally discriminate against a candidate or party, instructing courts to determine whether a public financing system is appropriately tailored to avoid a burden on political opportunity that is unfair or unnecessary. (ii) Yet Buckley also cautioned that there was, “[wjithout any doubt[,] a range of formulations” of public financing systems that “would sufficiently protect the public fisc and not foster factionalism, and would also recognize the public interest in the fluidity of our political affairs.” Id. at 103-04, 96 S.Ct. 612. Buckley made clear, moreover, that in establishing the required threshold level of public support, “the choice of the percentage requirement that best accommodate[d] the competing interests involved was for Congress to make.” Id. at 103, 96 S.Ct. 612. Accordingly, although courts play an important role in assessing whether a public financing system is properly tailored, Buckley warned that a court’s constitutional review should be circumscribed by deference to the legislative branch in its choice from among the “permissible range” of qualification criteria. See id. at 103-04, 96 S.Ct. 612. (iii) We also note with care that Buckley placed the evidentiary burden of demonstrating unconstitutional discrimination squarely on the plaintiffs. Buckley’s approval of the presidential-candidate financing system rested largely on the fact that the plaintiffs had “made no shoiuing that the election funding plan disadvantage[d] nonmajor parties by operating to reduce their strength below that attained without any public financing.” Id. at 98-99, 96 S.Ct. 612 (emphasis added). In other words, in this context, the evidentiary burden is not on the government to show that a public financing system comports with the Constitution; it is on the plaintiffs to show that the system does not. To determine whether the plaintiffs have succeeded, moreover,' the central question a court must ask is whether the plaintiffs have shown that the system has “operated] to reduce their strength below that attained without any public financing.” Id. (iv)Finally, in upholding the presidential-candidate financing system, Buckley instructed that courts should avoid reasoning based on speculation and should, instead, require tangible evidence of the “practical effects” of the public financing system. See id. at 101, 96 S.Ct. 612 (upholding the system in part because “[a]ny risk of harm to minority interests is speculative due to [a general] lack of knowledge of the practical effects of public financing”); see also id. at 97 n. 131, 96 S.Ct. 612 (declining to “rule out the possibility of concluding in some future case, upon an appropriate factual demonstration, that the public financing system invidiously discriminates against nonmajor parties” (emphasis added)). Thus, when a court evaluates a claim like the one presented here by asking whether the public financing system has “operat[ed] to reduce the[ ] strength” of minor parties “below that attained without any public financing,” id. at 98-99, 96 S.Ct. 612, the court should avoid speculative reasoning and focus instead on the evidence, if any, of the system’s practical effects. In sum, although Buckley did not expressly define “unfair” or “unnecessary,” we draw from Buckley four principles that clarify the meaning of those terms: (i) A public financing system may establish qualification criteria that condition public funds on a showing of “significant” public support. See Buckley, 424 U.S. at 96, 96 S.Ct. 612. (ii) There is a range of permissible qualification criteria, and although a public financing system must be tailored to avoid an unfair or unnecessary burden on the political opportunity of a party or candidate, a court must defer to a legislature’s choice of criteria so long as those criteria are drawn from the permissible range. See id. at 103-04, 96 S.Ct. 612. (iii) In assessing whether a burden is unfair or unnecessary, the central question is whether the plaintiffs have shown that the system has reduced the “strength” of minor parties below that attained before the system was put in place. Id. at 98-99, 96 S.Ct. 612. (iv) To determine whether the “strength” of minor parties has been reduced, a court should avoid speculative reasoning and instead focus on the evidence, if any, of the system’s “practical effects.” Id. at 101, 96 S.Ct. 612. We bear those principles in mind as we assess the CEP’s single-election qualification criteria. 3. Under Buckley’s Four Principles, the CEP’s Single-Election Qualification Criteria Are, on This Record, Constitutional Acknowledging that the CEP may condition public funds on a “significant” showing of public support in the previous election, Buckley, 424 U.S. at 96, 96 S.Ct. 612, our intuition suggests that the CEP’s single-election qualification criteria — 20% of the vote for full funding, 15% for two-thirds funding, and 10% for one-third funding — come close to the outer edge of the constitutionally permissible range. A public financing system must account for the “potential fluidity of American political life,” id. at 97, 96 S.Ct. 612 (quotation marks omitted), including the fact that minor-party candidates do, occasionally, defeat major-party opponents. Conditioning public funds on too high of a showing in the previous election risks entrenching the major parties and shutting out the rare minor-party candidate who is able to garner enough public support to win an election. Nevertheless, following Buckley’s example, we must look beyond our intuition to the concrete evidence of the CEP’s “practical effects.” In so doing, we find that data from the 2008 election contradict our intuition and show that a substantial number of minor-party candidates will be eligible for public funding in 2010 under the single-election qualification criteria. Indeed, over one third of the minor-party candidates (fifteen out of forty) who ran in the 2008 General Assembly elections received at least 10% of the vote, thereby qualifying themselves (or another member of their party) to receive partial funding in the same race in the 2010 election. Green Party II, 648 F.Supp.2d at 324. Five of those fifteen candidates — representing fully one eighth of all minor-party candidates — received over 20% of the vote and qualified for full funding in 2010. Id. Those record facts show that, although the CEP’s qualification criteria are high, they are not, as our intuition suggested, set so high as to shut-out minor-party candidates who enjoy public support. Furthermore, even if the CEP’s single-election qualification criteria impose some burden on the political opportunity of minor-party candidates, to evaluate whether the burden is unfair or unnecessary we must examine principally whether plaintiffs have shown that the CEP has “operat[ed] to reduce their strength below that attained without any public financing.” Buckley, 424 U.S. at 98-99, 96 S.Ct. 612. Searching the record, we find insufficient evidence in support of that claim. To the contrary, uncontroverted facts in the record show that minor-party candidates as a whole are arguably stronger — and certainly not weaker — under the CEP. In 2006, the election immediately before the CEP went into effect, zero minor-party candidates received between 15% and 19% of the vote and one minor-party candidate received more than 20% of the vote in legislative races. Green Party II, 648 F.Supp.2d at 322-23. Yet in 2008, after the CEP went into effect for legislative elections, minor-party candidates achieved more success at the polls: four minor-party candidates received between 15% and 19% of the vote and five minor-party candidates received more than 20% of the vote in legislative races. Id. at 324. This shows that, insofar as particular minor-party candidates are failing to qualify for public financing because of the CEP’s high qualification criteria, minor-party candidates as a whole are nonetheless just as strong — if not stronger- — than they were before the CEP went into effect. Their political opportunity, therefore, does not appear to have been burdened in a way that is unfair or unnecessary. We recognize that in reaching this conclusion, we have relied on data from only one election. Once the CEP has been in place for additional election cycles, there may develop a more complete picture of its effect on minor-party candidates. Following Buckley, therefore, “we of course do not rule out the possibility of concluding in some future case, upon an appropriate factual demonstration,” that the CEP’s single-election qualification criteria have, in fact, “operated] to reduce the[ ] strength” of minor parties “below that attained without any public financing.” 424 U.S. at 97 n. 131, 98-99, 96 S.Ct. 612. At present, however, and on the record before us, there is insufficient evidence to conclude that the CEP has burdened the political opportunity of minor-party candidates in a way that is unfair or unnecessary. In sum, although our intuition might suggest, as an abstract principle, that the CEP’s single-election qualification criteria — 20% of the vote for full funding, 15% for two-thirds funding, and 10% for one-third funding — come close to the outer edge of the constitutionally permissible range, the facts of record (most importantly, the 2008 election data) show that the qualification criteria are not so onerous as to deny funding to a sizeable number of minor-party candidates who enjoy substantial public support. Moreover, insofar as the CEP’s single-election qualification criteria may impose some burden on the political opportunity of minor-party candidates, the 2008 election shows that minor-party candidates as a whole are arguably stronger — and certainly not weaker — under the CEP. There is, therefore, little reason to think that the CEP has burdened the political opportunity of minor-party candidates in a way that is unfair or unnecessary. Giving proper deference to the Connecticut General Assembly to choose qualification criteria that “best accommodate!] the competing interests involved,” we “cannot say that” the General Assembly’s “choice falls without the permissible range.” Id. at 103-04, 96 S.Ct. 612. Accordingly, we hold that plaintiffs have presented insufficient evidence on this record to establish that the CEP’s single-election qualification criteria violate the First Amendment or the Equal Protection Clause. We therefore hold that the District Court erred in concluding that the single-election qualification criteria unconstitutionally discriminate against minor parties and their candidates. B. The Statewide Qualification Criteria Plaintiffs also claim that the CEP’s statewide qualification criteria, see Conn. Gen.Stat. § 9-702, impose an unfair or unnecessary burden on the political opportunity of minor-party candidates. The District Court agreed with plaintiffs, determining the statewide qualification criteria “substantially enhance[d] the relative strength of major party candidates compared to minor party candidates [by] ... encouraging] major parties to field candidates for historically uncompetitive seats, without regard to their likelihood of success.” Green Party II, 648 F.Supp.2d at 844. The District Court’s analysis focused on the effect of the statewide criteria in so-called safe or uncompetitive legislative districts, which are districts in which the candidate of one of the major parties is essentially assured of winning. In the state senate district encompassing New Haven, for example, less than 5% of all registered voters are Republican. See id. at 326 n. 33. Thus the New Haven state senate district is considered a safe district for the Democratic Party, since the Democratic candidate is almost certain to win the race for that seat. The District Court observed that in safe districts, one of the major-party candidates (e.g., the Republican candidate in New Haven) often fails to achieve 20% of the vote in an election and therefore would not qualify for CEP funding under the single-election qualification criteria. Under the statewide qualification criteria, however, that major-party candidate would, nevertheless, qualify for CEP funding, as the candidate would be on a ticket of a “major” party whose gubernatorial candidate achieved at least 20% of the vote in the last election. See Conn. Gen.Stat. §§ 9-372(5), 9-702. The District Court concluded, as a result, that the statewide qualification criteria “unfairly favor[ed] competition between major party candidates over competition from minor party candidates and thereby burden[ed] the political opportunity of minor party candidates.” Green Party II, 648 F.Supp.2d at 344. Once again, we cannot agree with the District Court’s application of law to fact. 1. There Is Insufficient Evidence to Conclude That the Statewide Qualification Criteria Have Imposed an Unfair or Unnecessary Burden on Minor-Party Candidates in Safe Districts We acknowledge that Buckley did not address the unique circumstances created by safe districts, as safe districts are not a feature of nationwide presidential elections. We do not, however, think that this is reason to abandon Buckley’s basic standard or analytical framework in assessing a burden on political opportunity. Examining, then, the record evidence of how the CEP has affected minor-party candidates, we find insufficient evidence in the record to conclude that the CEP’s statewide eligibility criteria has “reduce[d]” the “strength” of minor-party candidates in safe districts “below that attained without any public financing.” Buckley, 424 U.S. at 99, 96 S.Ct. 612. To the contrary, as we set forth above, the record here reveals that minor-party candidates as a whole, many of them running in safe districts, appear to have done better in 2008, and certainly no worse. See Green Party II, 648 F.Supp.2d at 323-24. Thus we cannot conclude, on this record, that the statewide eligibility criteria impose an unfair or unnecessary burden on minor-party candidates in safe districts. 2. Even if We Were to Speculate About the Effect of the Statewide Qualification Criteria, Our Specu