Landell v. Sorrell

382 F.3d 91

• Court: Court of Appeals for the Second Circuit
• Decided: August 7, 2002
• Docket: Docket Nos. 00-9159(L), 00-9180(CON), 00-9231(XAP), 00-9139(XAP), 00-9240(XAP)
• Status: Published

Opinions

STRAUB, Circuit Judge.

During his 1997 inaugural address, Vermont’s Governor offered the Vermont General Assembly a moment of telling candor: “As I’ve said before, money does buy access and we’re kidding ourselves and Vermonters if we deny it. Let us do away with the current system.” The General Assembly responded by promulgating Act 64, a comprehensive campaign finance reform package. The testimony and statements made during the General Assembly’s debate demonstrated that Vermont lawmakers were concerned with more than just the quid pro quo corruption that preoccupies much of campaign finance reform. Typically, this fear of corruption has involved the danger that politicians will sell their votes for campaign funds. The Vermont debate highlighted something else that public officials can, and apparently do, offer in exchange for funds: time and access. The General Assembly, together with the State’s chief executive, concluded that Vermont needed limitations governing its campaigns for state office with respect to both expenditures and contributions.

This appeal arises from a consolidated suit which brings a First Amendment challenge to key sections of Act 64. The plaintiffs have argued that Vermont’s reform violates the First Amendment guarantee of free speech and association in the political realm. At the conclusion of a bench trial, the District Court enjoined the enforcement of Act 64’s limitations on expenditures, gifts by non-resident contributors, and contributions by political parties to candidates. The District Court upheld all of Act 64’s other contribution limitations, including limits of between $200 and $400 on contributions to candidates by individuals and political action committees, limits of $2000 on contributions to political parties and political action committees, and regulations treating coordinated expenditures by third parties as contributions to a candidate.

All parties have appealed that decision. We are therefore asked to determine whether the First Amendment rights of free speech and political association forbid each of the challenged provisions, including (1) Vermont’s campaign expenditure limitations; (2) the contribution limits applied to candidates; (3) the contribution limits applied to political parties and political associations; (4) the limit on contributions by non-residents; and (5) the regulation of coordinated expenditures by political parties.

After issuance of the original opinion in this case, see Landell v. Sorrell, Nos. 00-9159(L), 00-9180(CON), 00-9231(XAP), 00-9139(XAP), and 00-9240(XAP) (2d Cir. Aug. 7, 2002) (slip op.), in which we upheld in large part both Act 64’s contribution limits and its expenditure limits, plaintiffs filed a petition for rehearing in banc. We withdrew our original opinion on October 3, 2002, pending further proceedings. Landell v. Sorrell, Nos. 00-9159(L), 00-9180(CON), 00-923(XAP), 00-9139(XAP), and 00-9240(XAP), 2002 WL 31268493 (2d Cir. Oct. 3, 2002). Having reconsidered our holding and taking serious note of the views presented during the rehearing process, we now issue this amended opinion, modifying our holding only with regard to Act 64’s expenditure limits. In both instances, our colleague, Judge Winter, has dissented.

As we did in our original opinion, we hold today that the Supreme Court, in Buckley v. Valeo, 424 U.S. 1, 96 S.Ct. 612, 46 L.Ed.2d 659 (1976) (per curiam), did not rule campaign expenditure limits to be per se unconstitutional, but left the door ajar for narrowly tailored spending limits that secure clearly identified and appropriately documented compelling governmental interests.¹ In applying the narrow tailoring test, we hold'that the State has established that the challenged expenditure limits are supported by its compelling interests in safeguarding Vermont’s democratic process from (1) the corruptive influence of excessive and unbridled fundraising² and (2) the effect that perpetual fundraising has on the time of candidates and elected officials. The evidence considered by the District Court and the Vermont legislature demonstrates that, absent expenditure limitations, the fundraising practices in Vermont will continue to impair the accessibility to elected officials which is essential to any democratic political system. The race for campaign funds has compelled public officials to give preferred access to contributors, essentially requiring candidates to sell their time in order to raise campaign funds. In addition, we affirm the District Court’s finding that effective campaigns can be run under Act 64’s limits.

Nevertheless, although we reaffirm these aspects of our original holding, we now conclude that a remand is necessary for further fact-finding on an aspect of the narrow tailoring inquiry that was not fully considered by the District Court: the crucial question of whether Act 64’s expenditure limits provision was the “least restrictive means” of furthering the State’s compelling anti-corruption and time-protection interests — or whether there are other less restrictive mechanisms available that might be as effective in satisfying the compelling interests established by Vermont. On- remand, the District Court should also consider another question that it did not reach in its original examination of this case — -whether treating related expenditures as candidate expenditures is constitutional. We therefore leave in place the District Court’s injunction, while remanding for further proceedings.

As for the remaining issues regarding Act 64’s contribution limitations, our decision remains the same in all material respects. We hold that all of Vermont’s provisions limiting the size of contributions survive scrutiny, including the treatment of a third party’s related expenditures as contributions and the application of contribution limitations to political party donations to candidates. We thus affirm the District Court’s rulings on contribution limits in part, but vacate and remand for further proceedings insofar as the District Court’s injunction prohibits enforcement of the political party limit. We also vacate the judgment and remand for further proceedings on (1) whether the provisions of Act 64 regulate wholly independent expenditures by political action committees (“PACs”) and, if so, whether those provisions are constitutional; and (2) the constitutionality of the law’s regulation of funds transferred from national political parties to state and local party entities.

Finally, we affirm the District Court’s holding that the First Amendment forbids Vermont’s attempt to limit campaign contributions by non-residents to no more than 25 percent of the total contributions received. Vermont has asserted no governmental interest sufficient to justify such a rule.

Due to the number of issues involved in this case, we set out the following table of contents:

CONTENTS

BACKGROUND.99

A. Act 64 .99

B. Procedural History.102

C. The District Court’s Decision.103

DISCUSSION. . 105

I. Act 64’s Expenditure Limitations. t — I

A. The Rule of Buckley (O rH

B. The Requisite Level of Scrutiny. O T — t

C. Compelling Interests . xf 1 I

1. Anti-Corruption. lo r-l

2. Time Protection. 05 T“t

3. Conclusion: Two Compelling Interests . ^ T — 1

D. Narrow Tailoring. lO t-H

1. Are Vermont’s time-protection and anti-corruption interests advanced by campaign spending caps?. to 05

2. Do spending limits at these levels allow for “effective advocacy”? to 00

3. Are mandatory expenditure limits the least restrictive means of advancing the State’s interests?. CO 1 — ¹

a. Type of Regulation . CO tO

b. Basis for Spending Cap Limits. CO CO

E. Conclusion: Remand for Further Findings . CO C7l

II. Act 64’s Contribution Limitations. CO

A. Limitations on Contributions by Individuals to Candidates. CO -3

B. Limitations on Contributions to and by PACs and Political Parties .... CO to

C. The Related Expenditure Provision is Constitutional as to Contributions.

D. The 25 Percent Limit on Out-of-State Donations is Unconstitutional to

CONCLUSION..148

BACKGROUND

A. Act 64

In 1997, Vermont passed a comprehensive campaign reform act known as Act 64. 1997 Vermont Campaign Finance Reform Act, codified at Vt. Stat. Ann. tit. 17, §§ 2801-2883 (“Act 64” or “the Act”). As enacted, Act 64 is a comprehensive campaign finance reform package, regulating contributions, expenditures, and disclosures related to candidates for state office in Vermont and political organizations that participate in Vermont elections. Section 2805a limits the expenditures that a candidate for office may make during a two-year election cycle. Candidates for statewide office are restricted to varying amounts depending on the position sought, with a candidate for governor limited to $300,000, for lieutenant governor to $100,000, and other statewide offices to $45,000. See id. § 2805a(a)(l)-(3). Candidates for governor and lieutenant governor also have the option of receiving public financing for their campaigns, provided they receive a certain number and amount of “qualifying contributions.” See id. §§ 2851-2856. Candidates for state senator and county office are limited to $4000 in expenditures, with state senators permitted an additional $2500 per seat in mul-ti-seat districts. See id. § 2805a(a)(4). Candidates for state representative in single-member districts can spend'no more than $2000, and those in two-member districts no more than $3000. See id. § 2805a(a)(5). Incumbent candidates may spend only 85 percent of the permitted amounts, except for incumbents of the General Assembly who may spend 90 percent. See id. § 2805a(e).

The Act also limits the size of contributions which candidates, political committees, -and political parties may receive from a single source during-a two-year election cycle. Candidates for state representative or local office may accept no more than $200 from a single source, political party, or political action committee. See id. § 2805(a). Slightly higher limits apply to candidates for state senate or county office ($300) and to candidates for statewide office ($400). See id. Political action committees and political parties may accept no contribution greater than $2000.' See id. For the purpose of all of these contribution limits, a political party’s state, county, and local branches (and national and regional affiliates of the party) count as a single unit. See id. § 2801(5).

The Act further imposes limits on the source of such contributions. Although candidates, political parties, and political action committees may accept contributions from out-of-state residents and political organizations, the sum of such amounts may not exceed 25 percent of the total contributions received. See id. § 2805(c).

Finally, the Act treats coordinated expenditures by third parties as both contributions to a candidate (subject to the applicable contribution limits) and expenditures by the candidate (counted against the candidate’s permissible budget). See id. §§ 2809(a)-(b). The Act creates a re-buttable presumption that expenditures made by political parties or political action committees that recruit or endorse candidates are related expenditures if they primarily benefit six or fewer candidates. See id. § 2809(d).

The Vermont General Assembly promulgated Act 64 after extensive legislative consideration. Numerous committees considered the Act, holding over 65 hearings with more than 145 witnesses testifying. Moreover, Act 64 was the latest installment of Vermont’s century-long effort to safeguard the accessibility and accountability of its elected officials.³

The General Assembly closely investigated the history of campaign financing for state races by examining campaign finance summaries for various Senate, House, and statewide races during the period 1978-1996, and reports of spending and contribution patterns in Vermont races. Members of the General Assembly analyzed the current status of Vermont’s campaign finance law, including the disintegration of Vermont’s voluntary expenditure limits. They also spoke with a range of experienced candidates and experts who provided testimony and data regarding the cost of campaigning, including the cost of travel, staff, materials, mailings, phone calls, and television and radio advertisements. Some of these witnesses described the widespread use of manipulative contribution devices, such as “bundling,” which enable special interests to direct large quantities of money by way of individual contributions to particular candidates. Polls demonstrated that citizens held deep reservations and suspicions about the influence of money on the political system, particularly the influence of large contributions. Some witnesses provided testimony detailing the role that big donors have played in advocating or blocking particular pieces of legislation in Vermont.

The record considered by the General Assembly demonstrated how the Vermont system of unbridled expenditures has created a situation where public officials are functionally compelled to sell privileged access through the fundraising system. The Vermont legislature explained that this results in a number of related phenomena, including (1) candidates being forced to spend too much time fundraising; (2) fund-raising requiring candidates to give preferred access to contributors over non-contributors; and (3) the system of increasing expenditures hindering the robust debate of issues, candidate interaction with the electorate, and public involvement and confidence in the electoral process.

The evidence adduced in those hearings also demonstrated broad and powerful support among the Vermont electorate for fundamental reform to the State’s campaign financing scheme. These legislative hearings culminated in passage of the Act by an overwhelming majority and with strong bipartisan support.

Based on these hearings, reports and data, the General Assembly set forth specific findings which, in its view, indicated the need for comprehensive reform that includes contribution and expenditure limitations in Vermont electoral campaigns. The General Assembly finds that:

(1) Election campaigns for statewide and state legislative offices are becoming too expensive. As a result many Vermonters are financially unable to seek election to public office and candidates for statewide offices are spending inordinate amounts of time raising campaign funds.

(2) Some candidates and elected officials, particularly when time is limited, respond and give access to contributors who make large contributions in preference to those who make small or no contributions.

(3) In the context of Vermont, contributions larger than the amounts specified in this act are considered by the legislature, candidates and elected officials to be large contributions.

(4) Robust debate of issues, candidate interaction with the electorate, and public involvement and confidence in the electoral process have decreased as campaign expenditures have increased.

(5) Increasing campaign expenditures require candidates to seek and rely on a smaller number of larger contributors, often outside the state, rather than a large number of small contributors.

(6) In the context of Vermont, contributions scaled in proportion to the size of the electoral district of the office and up to the amounts specified in this act adequately allow contributors to express their opinions, level of support and their affiliations.

(7) In the context of Vermont, candidates can raise sufficient monies to fund effective campaigns from contributions no larger than the amounts specified in this act.

(8) Limiting large contributions, particularly from out-of-state political committees or corporations, and limiting campaign expenditures will encourage direct and small group contact between candidates and the electorate and will encourage the personal involvement of a large number of citizens in campaigns, both of which are crucial to public confidence and the robust debate of issues.

(9) Large contributions and large expenditures by persons or committees, other than the candidate and particularly from out-of-state political committees or corporations, reduce public confidence in the electoral process and increase the appearance that candidates and elected officials will not act in the best interests of Vermont citizens.

(10) Citizen interest, participation and confidence in the electoral process is lessened by excessively long and expensive campaigns.

(11) Public financing of campaigns, conditioned on an appropriate number of qualifying contributions, will increase .citizen participation and will limit the time spent soliciting contributions, and will reduce the need of elected officials to respond to, and provide access to, contributors. As a result candidates will be freed to devote more time and energy to debate of the issues and elected officials will be able to spend more time responding to constituents and to performing their official duties.

(12) Public financing of campaigns, coupled with generally applicable contribution and expenditure limitations, will level the financial playing field among candidates and provide resources to independent candidates, both of which will increase the debate of issues and ideas.

(13) In Vermont, campaign expenditures by persons who are not candidates have been increasing and public confidence is eroded when substantial amounts of soft money are expended, particularly during the final days of a campaign.

(14) Identification of persons who publish political advertisements assists in enforcement of the contribution and expenditure limitations established by this act.

(15) Because it is essential for all candidates to have their names and positions on issues known to the electorate and because incumbents have a substantial advantage in these areas, public grants and campaign expenditures must be reduced for incumbents.

1997 Vt. Laws P.A. 64 (H. 28). On June 26, 1997, Vermont’s Governor signed Act 64 into law.

B. Procedural History

The current suit was consolidated from three separate civil actions. On May 18, 1999, Marcella Landell, Donald R. Bru-nelle, and the Vermont Right to Life Committee, Inc., sued Vermont’s Attorney General, Secretary of State and fourteen state’s attorneys (“Vermont”). On August 13, 1999, Neil Randall, George Kuusela, Steve Howard, Jeffrey A. Nelson, John Patch, and the Vermont Libertarian Party also brought suit, as did the Vermont Republican State Committee on February 15, 2000. The remaining defendants, including the Vermont Public Interest Research Group, the League of Women Voters of Vermont, and numerous members of Vermont’s General Assembly (collectively “Defendant-Intervenors”), successfully intervened in the consolidated action.⁴

Plaintiffs argued that the challenged provisions unconstitutionally infringe them First Amendment rights to free speech and political association.⁵ The District Court held a ten-day bench trial between May 8, 2000 and June 2, 2000. An array of former and current public office holders, private citizens, and electoral experts testified about Vermont’s interest in campaign finance legislation, the history of elections and campaign finance reform in Vermont, the cost of campaigning in Vermont, and the likely effect of Act 64’s challenged provisions on Vermont races, candidates and political actors. As we discuss in more detail below, the ten-day bench trial resulted in the District Court’s upholding most of the challenged provisions, but striking down Act 64’s expenditure limitations, its limitations on contributions by parties to candidates, and its restriction on contributions from out-of-state sources. Vermont and the other defendant-appellants timely appeal from the District Court’s order holding those portions of Act 64 unconstitutional. Vermont is joined by amici, the Brennan Center for Justice at New York University School of Law and the States of Colorado, Connecticut, Maryland, New York, and Oklahoma. The plaintiffs have cross-appealed, contending that the District Court should have also enjoined the enforcement of the other disputed provisions of the Act.

C. The District Court’s Decision

After receiving post-trial submissions, the District Court issued an opinion containing its findings of fact and conclusions of law. See Landell v. Sorrell, 118 F.Supp.2d 459 (D.Vt.2000). First, the District Court held that the plaintiffs have standing to challenge the subject provisions of the Act. Id. at 475. As to the merits, although the District Court found that Vermont had generally demonstrated several compelling justifications for Act 64’s comprehensive reform of the campaign finance system, the court concluded that some of Act 64’s provisions violated the First Amendment. With the exception of the expenditure limitations, the District Court applied the standard of review of “exacting scrutiny,” inquiring whether the provision is narrowly tailored to serve a sufficiently important governmental interest. With regard to the expenditure limits, the District Court interpreted Buckley v. Valeo, 424 U.S. 1, 96 S.Ct. 612, 46 L.Ed.2d 659 (1976) (per curiam), as forbidding such limitations per se and held that any contrary decision would violate the doctrine of stare decisis. 118 F.Supp.2d at 483. The District Court rejected the expenditure limitations despite its findings that Vermont had established several compelling interests in their favor, namely: (1) freeing office holders from the requirements of excessive fundraising so that they can perform their duties; (2) preserving faith in democracy; (3) protecting access to the political arena for those unable to access large sums of money; and (4) diminishing the importance of repetitive 30-second commercials. Id. at 482-83. Despite holding that the expenditure limitations are illegal under Buckley, the District Court did find that the expenditure limits would permit effective campaigning. Id. at 471-72.

The District Court upheld the provisions imposing limitations on amounts that individuals may contribute to political campaigns, Vt. Stat. Ann. tit. 17, §§ 2805(a)-(b). The District Court found that the Vermont provision, like the statutory provision upheld in Buckley, served the governmental interest in preventing actual and perceived corruption in the political system. Id. at 476-79. As evidence of the existence of such an interest, the District Court relied on citizen polls, comments by public officials, and media accounts of citizen concern with the state of the political system, as well as direct testimony from citizens regarding their views of the political system. Id. at 465-70. The evidence indicated that the current financing scheme eroded public confidence in the democratic system and contributed to a waning public interest in elections. Id. Finally, the evidence supported the public’s perception that large contributions won actual influence over the legislative process. Again, the District Court relied not only on trial testimony, but also on studies showing how the pressure to raise money made legislative initiatives less likely to succeed if contrary to the wishes of well-organized interest groups who frequently contribute to candidates. Id.

The District Court further analyzed the amounts of the contribution limitations, and held that they were narrowly tailored to serve this anti-corruption purpose. . In support of the narrow-tailoring conclusion, the court relied upon the cost of previous elections in Vermont, the size of Vermont electoral districts and the corresponding cost-per-voter, the effect of the limitations on the Burlington mayoral election held after the passage of Act 64, the widely-held public view that donations in excess of the Act’s limitations were suspicious, and the fact that the limitation did not inhibit “effective campaigning.” Id. at 470-72, 476-80.

The District Court rejected the contention that PACs merit special treatment; it thus upheld the restrictions on contributions by and to PACs pursuant to Act 64. See Vt. Stat. Ann. tit. 17, §§ 2805(a)-(b). If contributions by individuals may be restricted, the court reasoned, then so too may gifts by individuals to associations that in turn give funds to candidates. The District Court reasoned that Vermont has the same anti-corruption interest in limiting PAC contributions as those by individuals. The contribution limit closes a loophole which individuals could exploit to evade individual contribution limitations. Id. at 488-89.

The District Court held, however, that political parties deserve greater freedom in their ability to make contributions to political candidates. Although the District Court upheld the $2000 limitation on contributions to political parties pursuant to Vt. Stat. Ann. tit.' 17, § 2805(a), it struck down the provision limiting contributions to candidates insofar as it applies to those candidate’s own political parties pursuant to Vt. Stat. Ann. tit. 17, § 2805(b). Id. at 486-87. Regarding contributions to political parties, the court relied on Vermont’s anti-corruption interest, noting that unrestrained contributions to parties provided a loophole to individuals wishing to evade restrictions on direct contributions. Quoting Nixon v. Shrink Mo. Gov’t PAC, 528 U.S. 377, 397, 120 S.Ct. 897, 145 L.Ed.2d 886 (2000)(Shrink), the District Court found that the limit imposed by the statute is not “so radical in effect as to render political association ineffective, drive the sound of [a political party’s] voice below the level of notice, and render contributions pointless.” 118 F.Supp.2d at 485. Moreover, the District Court found that, given Vermont’s electoral situation, the $2000 limit did not inhibit the strength of political parties. The court relied on the evidence specifically concerning Vermont campaigns and politics, a comparison of limits on contributions to candidates in other jurisdictions, and the ability of the Republican Party to raise substantial sums while subject to Act 64’s limitations. Id. at 484-86. The District Court, however, did not address the constitutionality of transfers of money to state and local parties from the national affiliated party, which are apparently subject to the $2000 limitation.

The District Court held that Vermont’s limits on how much a political party could give to its.own candidates for various state offices ($400, $300; and $200, respectively) were unconstitutionally low. Id. at 487. The court recognized that the anti-corruption interest may justify some limitations, given that' corruption may “filter[ ] through the party machine.” Id. at 486. But according to the District Court, those limitations must be balanced against the special role political parties play in the American electoral system. Without much factual discussion, the court concluded that the limits would reduce the party’s voice to a whisper — since political parties speak through their candidates and the restrictions were too stringent even for the small scale of Vermont’s electoral races. Id. at 487.

The District Court also upheld the treatment of state and local parties as a single entity for the purpose of calculating the contribution limitations pursuant to Vt. Stat. Ann. tit. 17, §§ 2801(5) & 2301-20. The court relied on a number of factors, including the fact that notwithstanding its adamant assertions, the defendant Vermont Republican State Committee had never acted as a loose confederation of entities in the conduct of the litigation. Id. at 487-88.

The District Court upheld the provision of Act 64 that treats third party expenditures “intentionally facilitated by, solicited by or approved by the candidate or the candidate’s political committee” as contributions to the candidate pursuant to the Act. See Vt. Stat. Ann. tit. 17, § 2809(a) & (c). The purpose of the provision is to close a loophole which would otherwise permit evasion of the legitimate contribution limitations by engaging in coordinated expenditures. The District Court further upheld the provision establishing a rebut-table presumption that any third party expenditure benefiting six or fewer candidates is a related expenditure. See id. § 2809(d). The court explained that the presumption is a guideline to assist in compliance, and that since Vermont’s Secretary of State has determined that the presumption is rebuttable, it does not unduly chill otherwise protected speech activity. Id. at 492. Although the District Court upheld the provision treating related expenditures as contributions to candidates, it struck down the provision treating related expenditures as expenditures by candidates. Vt. Stat. Ann. tit. 17 § 2809(a) & (b).

The District Court found unconstitutional the provision that caps out-of-state funds at 25 percent of total contributions received by a candidate, political party, or PAC pursuant to Vt. Stat. Ann. tit. 17, § 2805(c). The court found that the factual record did not establish any legitimate governmental interest in limiting such contributions. Id. at 483-84. Instead, the record only supported an inference that such contributions raise the risk of corruption when they are large—a problem solved by the contribution limits. The fact that a donor is a resident of another state is not an important factor in either increasing the risk of corruption or the public’s perception of corruption. Moreover, the mechanics of the ban indicated a lack of proper tailoring because it acts as a complete bar to contributions by some would-be contributors.

Finally, the court held that, under Vermont law, the unconstitutional provisions may be severed from the rest of Act 64. Id. at 492-93.

DISCUSSION

As a threshold matter, the defendants have challenged the plaintiffs’ standing to assert this facial challenge to Act 64’s expenditure and contribution limitations. In order to present a “case or controversy” within the meaning of Article III of the Constitution, the plaintiffs seeking relief must have a sufficient “personal stake in the outcome of the controversy.” Buckley v. Valeo, 424 U.S. 1, 11, 96 S.Ct. 612, 46 L.Ed.2d 659 (1976) (internal quotation marks omitted). The District Court provided careful analysis demonstrating that each of the challenged provisions arguably affects the First Amendment rights of one or more of the plaintiffs. See Landell v. Sorrell, 118 F.Supp.2d 459, 474-76 (D.Vt.2000). For the reasons set forth by the District Court, we uphold its determination that the plaintiffs have standing to assert their challenge to Act 64’s expenditure and contribution limits.

Although we review the District Court’s factual findings for clear error pursuant to Rule 52(a) of the Federal Rules of Civil Procedure, see Bose Corp. v. Consumers Union of U.S., Inc., 466 U.S. 485, 498, 104 S.Ct. 1949, 80 L.Ed.2d 502 (1984), the District Court’s legal conclusions regarding the campaign finance reform legislation are subject to de novo review. Indeed, the breadth of review of factual issués is greater in cases raising First Amendment issues: “an appellate court has an obligation to ‘make an independent examination of the whole record’ in order to make sure that ‘the judgment does not constitute a forbidden intrusion on the field of free expression.’ ” Id. at 499, 104 S.Ct. 1949 (quoting New York Times Co. v. Sullivan, 376 U.S. 254, 284-286, 84 S.Ct. 710, 11 L.Ed.2d 686 (1964)). The appellate court must also be vigilant for errors of law that “may infect a so-called mixed finding of law and fact, or a finding of fact that is predicated on a misunderstanding of the governing rule of law.” Bose Corp., 466 U.S. at 501, 466 U.S. 485.

In reviewing campaign finance regulations, “the level of scrutiny is based on the importance of the political activity at issue to effective speech or political association.” Federal Election Comm’n v. Beaumont, 539 U.S. 146, 161, 123 S.Ct. 2200, 156 L.Ed.2d 179 (2003) (internal quotation marks omitted). Campaign contributions advance political association by allowing one to affiliate with a political candidate, and “enabling] like-minded persons to pool their resources in furtherance of common political' goals.” Buckley, 424 U.S. at 22, 96 S.Ct.' 612. However, restrictions on contributions have been treated as merely “marginal” speech restrictions because contributions “lie closer to the edges than to the core of political expression.” Beaumont, 539 U.S. at 161, 123 S.Ct. 2200. As a result, contribution limits pass muster if they are “closely drawn to match a sufficiently important interest.” Id. at 162, 123 S.Ct. 2200 (citations and internal quotation marks omitted). And, as the Supreme Court recently observed, its cases “have made clear that the prevention of corruption or its appearance constitutes a sufficiently important interest to justify political contribution limits.” McConnell, 540 U.S. at-, 124 S.Ct. at 660; see also id. at-, 124 S.Ct. at 657 n. 40 (explaining that since Buckley, the Court has “consistently applied less rigorous scrutiny to contribution restrictions aimed at the prevention of corruption and the appearance of corruption”) (collecting cases).

However, “limits on political expenditures deserve closer scrutiny than restrictions on political contributions'.” Federal Election Comm’n v. Colorado Republican Federal Campaign Comm., 533 U.S. 431, 440, 121 S.Ct. 2351, 150 L.Ed.2d 461 (2001) (Colorado Republican II); see also McConnell, 540 U.S. at -, 124 S.Ct. at 655. The Supreme Court has treated limits on campaign spending as a direct restraint on speech, and thus, expenditure limits must be narrowly tailored to serve a compelling state interest. See Austin v. Michigan Chamber of Commerce, 494 U.S. 652, 657, 110 S.Ct. 1391, 108 L.Ed.2d 652 (1990) (addressing corporate expenditures).

With these standards in mind, we review each of the challenged provisions in turn.

I. Act 64’s Expenditure Limitations

A. The Rule of Buckley

Buckley v. Valeo remains the seminal case governing the constitutional review of campaign finance reform efforts, including expenditure limitations. 424 U.S. 1, 96 S.Ct. 612, 46 L.Ed.2d 659 (1976). The Buckley Court considered and reject-? ed a variety of expenditure limitations, in-eluding a ceiling on independent, campaign-related expenditures, a ceiling on a candidate’s use of personal or family resources, and a ceiling on a candidate’s campaign expenditures. Like the federal statute reviewed in Buckley, Act 64 limits the total amount of campaign funds that a candidate may spend.

Although the clear language of Buckley requires that courts should review expenditure limits with exacting scrutiny, the District Court in this case (and it is by no means alone) apparently felt that Buckley categorically prohibits expenditure limitations. See, e.g., Homans v. City of Albuquerque, 366 F.3d 900, 914-21 (10th Cir.2004); Kruse v. City of Cincinnati, 142 F.3d 907, 918-19 (6th Cir.), cert. denied 525 U.S. 1001, 119 S.Ct. 511, 142 L.Ed.2d 424 (1998); see also post at 151, 152, 159, 172, 185 (Winter, dissenting). We disagree. The Buckley Court’s rejection of particular federal campaign expenditure limitations was rooted in Congress’ purported reasons for such legislation and the failures of those interests to demonstrate any need for expenditure limits. 424 U.S. at 55-58, 96 S.Ct. 612. Ultimately, the Court concluded that the federal government had failed to assert any sufficiently important interest that its expenditure limitations served. See id. at 55, 96 S.Ct. 612.

Examining the federal government’s interest in eliminating corruption from federal elections, the Buckley Court concluded that the government’s asserted rationale only applied to large contributions — that is, eliminating large contributions fully satisfied the government’s anti-corruption interest. See id. at 56-57, 96 S.Ct. 612. The federal government claimed that expenditure limitations were necessary to make contribution limitations easier to enforce, arguing that when candidates cannot spend large quantities of money, they have a weaker incentive to accept illegally large contributions. But the Court concluded that the contribution limitations promised to be sufficiently effective on their own. See id. Based on the Court’s review of the record, “[tjhere [was] no indication that the substantial criminal penalties” attached to violations of contribution limits, as well as the “political repercussion of such violations,” would not suffice to realize this anti-corruption interest. Id.

Nor was the Court persuaded that the federal government had a sufficient interest in utilizing expenditure limitations to equalize the financial resources of candidates competing for office. See id. at 56-57, 96 S.Ct. 612. The contribution limits would assure that any difference in resources “varfies] with the size and intensity of the candidate’s support.” Id. at 56, 96 S.Ct. 612. Finally, the Court addressed the argument that expenditure limitations served the federal government’s interest “in reducing the allegedly skyrocketing costs of political campaigns.” Id. at 57, 96 S.Ct. 612. The Court rejected the idea that the state had a sufficient interest in setting the appropriate scope of the “quantity and range of debate on public issues in a political campaign.” Id. In other words, Buckley held that large campaign expenditures, in and of themselves, are not inherently suspect.

We conclude, then, that Vermont cannot sustain Act 64 by asserting a need to control excessive campaign spending per se. But critically, the Buckley Court did not conclude that the Constitution would always prohibit expenditure limits, regardless of the reasons asserted and the record supporting the limitations. It simply held that based on the record before it, “[n]o governmental interest that has been suggested is sufficient to justify” the federal expenditure limits. Id. at 55, 96 S.Ct. 612. Accordingly, after Buckley, there remains the possibility that a legislature could identify a sufficiently strong interest, and develop a supporting record, such that some expenditure limits could survive constitutional review.

We are not alone in concluding that Buckley does not permanently foreclose any consideration of campaign expenditure limitations. In Shrink, Justices Breyer, Ginsburg and Stevens all recognized that our post-Buckley experiences with campaign finance have demonstrated that we need a flexible approach to the constitutional review of campaign finance rules. Justice Breyer, who was joined by Justice Ginsburg, concluded that courts must resist a static reading of Buckley's mandate, which may require reinterpretation in light of subsequent experience, including a legislature’s “political judgment that unlimited spending threatens the integrity of the electoral process.” 528 U.S. at 403-04, 120 S.Ct. 897 (Breyer, J., concurring). Legislatures may protect the electoral process not only from quid pro quo corruption, but also from the threat that campaign funding may pose to the “integrity of the electoral process.” Id. at 401, 120 S.Ct. 897. Justice Stevens also articulated the need for “a fresh reexamination” of Buckley, and concluded that “Money is property; it is not speech.” Id. at 398, 120 S.Ct. 897 (Stevens, J., concurring). And although Justice Kennedy argued from a different perspective that the post-Buckley experience requires a wholesale abandonment of the approach adopted in Buckley, he too left open the possibility that “Congress, or a state legislature, might devise a system in which there are some limits on both expenditures and contributions thus permitting officeholders to concentrate their time and effort on official duties rather than on fundraising.” Shrink, 528 U.S. at 409, 120 S.Ct. 897 (Kennedy, J., dissenting); see also McConnell, 540 U.S. at -, 124 S.Ct. at 745 (Kennedy, J., concurring in part and dissenting in part) (indicating by implication that Buckley did not make expenditure limits per se invalid).⁶

Indeed, some judges have noted that reconsideration might be required were a court faced with compelling evidence that unlimited expenditures posed great dangers to the very political process that Buckley sought to safeguard. Justices Stevens and Ginsburg have supported the constitutionality of spending limits on political parties for, among other reasons, the likelihood that such limits would improve, rather than inhibit, a flourishing political system:

It is quite wrong to assume that the net effect of limits on contributions and expenditures — which tend to protect equal access to the political arena, to free candidates and their staffs from the interminable burden of fundraising, and to diminish the importance of repetitive 30-second commercials — will be adverse to the interest in informed debate protected by the First Amendment.

Colorado Republican Federal Campaign Comm. v. Federal Election Comm’n, 518 U.S. 604, 649-50, 116 S.Ct. 2309, 135 L.Ed.2d 795 (1996) (Colorado Republican I) (Stevens, J., dissenting). In part, they reached this conclusion because of the comparative competency of the different branches of government: “Congress surely has both wisdom and experience in these matters that is far superior to ours.” Id. at 650, 116 S.Ct. 2309. Moreover, one judge sitting on the Sixth Circuit has pointed out that Buckley was “decided on a slender factual record” and that a fuller record might satisfy the constitutional requirement that expenditure limits be narrowly tailored to a compelling interest. Kruse, 142 F.3d at 919 (Cohn, J., concurring); Cf. Laurence H. Tribe, AMERICAN Constitutional Law § 13-27, at 1133 n.l (2d ed. 1988) (“One consequence of th[e] expedited review [in Buckley] was that the Supreme Court, working in a factual vacuum, was forced to indulge in more than a little empirical speculation about such issues as the circumvention of expenditure limits and the impact of those limits on campaign speech.”); Burt Neuborne, One Dollar-One Vote: A Preface to Debating Campaign Finance Reform, 37 WashbuRN L.J. 1, 30 (1997) (“Since the Buckley Court’s judgment was made without the benefit of a factual record, critics have argued that it is time for a factually based study of the potential for corruption inherent in large, independent expenditures.”); David R. Lagasse, Note, Undue Influence: Corporate Political Speech, Power and the Initiative Process, 61 BROOK. L. Rev. 1347, 1357 (1995) (“The Supreme Court granted certiorari in Buckley v. Valeo without either party to the action having the opportunity to develop a strong factual record on which the Court could base its ultimate decision. Thus, the Court faced the issue of Congress’s power to regulate campaign expenditures purely on theoretical grounds, without the benefit of developing an adequate factual record.”) (citing Bob Woodward & Scott Armstrong, The BRETHREN 469-70 (1979)).

The academic literature also contains persuasive analyses that our post -Buckley understanding of campaign finance requires a careful evaluation of the evidence in support of expenditure limits. See, e.g., Richard Briffault, Nixon v. Shrink Missouri Government PAC. The Beginning of the End of the Buckley Era?, 85 Minn. L. Rev. 1729, 1765-69 (2001) (arguing that fair and competitive elections may require some form of expenditure limitations); Vincent Blasi, Free Speech and the Widening Gyre of Fund-Raising: Why Campaign Spending Limits May Not Violate the First Amendment After All, 94 Colum. L. Rev. 1281, 1288-89 (1994) (noting that changed circumstances and never-before considered governmental interests, including the protection of candidates’ time, might be sufficiently compelling to support expenditure limits).

Although we recognize that there is considerable dissatisfaction with Buckley’s, approach, we still premise our conclusions on the assumption that Buckley continues to govern the constitutional review of campaign finance laws. However, we do not accept an unyielding interpretation of Buckley that expenditure limits are per se unconstitutional, because such a static approach to Buckley’s import would require us to ignore not only Buckley’s own language, but also over three decades of experience as to how the campaign funds race has affected • public confidence and representative democracy.⁷ In sum, like the federal expenditure limitations considered in Buckley, 'Act 64’s expenditure limitations rise or fall on whether they have been narrowly tailored to a compelling governmental interest. It is to that question that we now turn.

B. The Requisite Level of Scrutiny

As a regulation of the amount that a candidate can spend on speech made “for the purpose of influencing an election,” Vermont’s expenditure limits are a content-based restriction on speech. See Burson v. Freeman, 504 U.S. 191, 197, 112 S.Ct. 1846, 119 L.Ed.2d 5 (1992) (treating election provision as content-based because “whether individuals may exercise their free speech rights ... depends entirely on whether their speech is related to a political campaign”).⁸ “Content-based regulations are presumptively invalid,” R.A.V. v. St. Paul, 505 U.S. 377, 382, 112 S.Ct. 2538, 120 L.Ed.2d 305 (1992), and the government bears the burden of rebutting that presumption. United States v. Playboy Entertainment Group, Inc., 529 U.S. 803, 817, 120 S.Ct. 1878, 146 L.Ed.2d 865 (2000). To uphold a content-based restriction on speech, the government must prove the existence of a compelling state interest to support the restriction, and that the restriction is narrowly tailored to advance that interest. ■

In the context of expenditure limits, then, the level of scrutiny applied is akin to the “strict scrutiny” standard frequently employed in the equal protection context, in terms of the required degree of “fit” between means and ends. Cf. Guido Calabresi, Antidiscrimination and Constitutional Accountability (What the Bork-Brennan Debate Ignores), 105 HaRV. L. Rev. 80, 112-13 n.94 (citing cases) (noting that, traditionally, judicial review has been at its strongest in protecting against infringement on First Amendment rights). Our application of this standard is informed both by the particular First Amendment right implicated by the challenged restrictions, as well as by the degree of deference owed to the supporting legislative findings.

Turning to the first of these factors, there is no doubt that “[pjolitical speech is the primary object of First Amendment protection.” Shrink, 528 U.S. at 410-11, 120 S.Ct. 897 (Thomas, J., dissenting). Moreover, in our representative democracy, the free exchange of political information “should receive the most protection when it matters the most—during campaigns for elective office.” Id. at 411, 120 S.Ct. 897. However, the precise object of First Amendment protection in this case, for most plaintiffs, is the ability to spend money on political speech—not the speech itself. See Shrink, 528 U.S. at 400, 120 S.Ct. 897 (Breyer, J., concurring) (money is not speech; it “enables speech”). To be sure, the Supreme Court has consistently held that the expenditure of money is so critical in enabling political speech in today’s mass society, that it should receive the same First Amendment protection as the speech itself. We do not question this proposition—and indeed apply it in this case—but, particularly in light of at least one Supreme Court Justice’s willingness to rethink the money equals speech equation (J. Stevens concurring in Shrink, 528 U.S. at 398, 120 S.Ct. 897), think it important to define the protected interest as precisely as possible.

Although most of the plaintiffs are persons or organizations that want to spend money on speech, plaintiff Marcella Landell is a voter who wants to receive political speech. Her First Amendment right to receive such speech is the equivalent of the right of the speakers. See Virginia State Board of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U.S. 748, 756, 96 S.Ct. 1817, 48 L.Ed.2d 346 (1976) (the First Amendment protection afforded is to “the communication, to its source and to its recipients both”). As Landell describes her interest in her brief, she “does not wish her ability to cast a wise and informed vote to be restricted by the State of Vermont imposing a direct barrier on the amount of candidate speech she may receive.” Restrictions of political speech that “hamstring[ ] voters seeking to inform themselves about the candidates and the campaign issues” are unconstitutional. Eu v. San Francisco County Democratic Central Comm., 489 U.S. 214, 223, 109 S.Ct. 1013, 103 L.Ed.2d 271 (1989). Plaintiffs argue that this high level of protection, as applied in Buckley, dictates that the expenditure limit provision must automatically be struck down.

On the other hand, Vermont appears to argue that deference to the legislature—on whether the interests asserted in favor of expenditure limits are compelling, and whether expenditure limits are necessary to achieve these goals—is warranted. Vermont cites several Supreme Court cases in support of its view of legislative deference, including Federal Election Comm’n v. National Right to Work Comm., 459 U.S. 197, 210, 103 S.Ct. 552, 74 L.Ed.2d 364 (1982) (it is improper to “second-guess a legislative determination as to the need for prophylactic measures where corruption is the evil feared”); Turner Broadcasting System, Inc. v. FCC, 512 U.S. 622, 665, 114 S.Ct. 2445, 129 L.Ed.2d 497 (1994) (Turner T) (“courts must accord substantial deference to the predictive judgments” of the legislature); Turner Broadcasting System, Inc. v. FCC, 520 U.S. 180, 196, 117 S.Ct. 1174, 137 L.Ed.2d 369 (1997) (Turner II) (“We owe Congress’ findings an additional measure of deference out of respect for its authority to exercise the legislative power.”); and Walters v. National Association of Radiation Survivors, 473 U.S. 305, 330-31 n. 12, 105 S.Ct. 3180, 87 L.Ed.2d 220 (1985) (Congress’ factual findings are entitled to “a great deal of deference, inasmuch as Congress is an institution better equipped to amass and evaluate the vast amounts of data bearing on” an issue). Indeed, the District Court concluded that “[ajlthough legislative findings are not entirely isolated from review,” it was “required to exercise considerable deference to such findings.” 118 F.Supp.2d at 476 (citing Turner II). Accordingly, the court adopted the fifteen official findings excerpted swpra and in the District Court opinion, but made clear that it was also considering the other evidence presented at trial. Id. at 468-74.

As to plaintiffs’ position, we disagree that the high level of protection accorded political speech or the money enabling it dictates that the provision must automatically be struck down. Cf. McConnell, 540 U.S. at -, 124 S.Ct. at 706 (“Many years ago we observed that ‘[t]o say that Congress is without power to pass appropriate legislation to safeguard ... an election from the improper use of money to influence the result is to deny to the nation in a vital particular the power of self protection.’ ”) (quoting Burroughs v. United States, 290 U.S. 534, 545, 54 S.Ct. 287, 78 L.Ed. 484 (1934)); Storer v. Brown, 415 U.S. 724, 729-30, 94 S.Ct. 1274, 39 L.Ed.2d 714 (1974) (compelling' interest in the integrity and stability of the election process means that “every substantial restriction on the right to vote or to associate” should not automatically be invalidated). In our view, this level of protection is the starting point, not the endpoint, for scrutiny of Vermont’s expenditure limits.

Indeed, the Supreme Court has been clear in its rejection of the view that “strict scrutiny is ‘strict in theory, but fatal in fact.’ ” Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 237, 115 S.Ct. 2097, 132 L.Ed.2d 158 (1995) (quoting Fullilove v. Klutznick, 448 U.S. 448, 519, 100 S.Ct. 2758, 65 L.Ed.2d 902 (1980) (Marshall, J., concurring)) (explaining that “[w]hen race-based action is necessary to further a compelling interest, such action is within constitutional constraints if it satisfies the ‘narrow tailoring’ test this Court has set out in previous cases”). This observation has proven true in the First Amendment context, as the Supreme Court has validated a number of electoral regulations against First Amendment challenge even while applying strict scrutiny. See, e.g., Burson v. Freeman, 504 U.S. 191, 112 S.Ct. 1846, 119 L.Ed.2d 5 (1992) (plurality opinion) (upholding state ban on electioneering activity near polling places); Austin v. Michigan Chamber of Commerce, 494 U.S. 652, 110 S.Ct. 1391, 108 L.Ed.2d 652 (1990) (upholding statute restricting independent expenditures by corporations on campaigns). Careful analysis is particularly important, in applying strict scrutiny, where, as Justice Breyer has put it, “a law significantly implicates competing constitutionally protected interests in complex ways.” Shrink, 528 U.S. at 402, 120 S.Ct. 897 (Breyer, J., concurring). As we will explain, this is a case where “constitutionally protected interests lie on both sides of the legal equation,” preventing a simple equation of strict scrutiny with constitutional infirmity. Id. at 400, 120 S.Ct. 897; see also, e.g., Burson, 504 U.S. at 199, 112 S.Ct. 1846 (plurality opinion) (recognizing compelling interest in preserving integrity of electoral process); id. at 213, 112 S.Ct. 1846 (Kennedy, J., concurring) (“[TJhere is a narrow area in which the First Amendment permits freedom of expression to yield to the extent necessary for the accommodation of another ■ constitutional right.”); Storer, 415 U.S. at 736, 94 S.Ct. 1274 (allowing some restrictions on ballot access in order to further the “State’s interest in the stability of its political system”).

Nor should we adopt total legislative deference as the appropriate level of scrutiny. Deference to legislative findings may well be warranted on certain issues relating to the constitutionality of election-related laws, such as the precise level of contribution limits, as in Buckley, 424 U.S. at 30, 96 S.Ct. 612, or whether 100 feet, as opposed to 50 or 75- feet, is an adequate radius surrounding a polling place to ban electioneering, as in Burson, 504 U.S. at 209-10, 112 S.Ct. 1846. Some degree of deference on the issue of whether there are state interests that justify legislative changes to the State’s electoral system may also be appropriate. See, e.g., Federal Election Comm’n v. Beaumont, 539 U.S. 146, 155, 123 S.Ct. 2200, 156 L.Ed.2d 179 (2003) (“[Djeference to legislative choice is warranted particularly when Congress regulates campaign contributions, carrying as they do a plain threat to political integrity and a plain warrant to counter the appearance and reality of corruption and the misuse of corporate advantages.”). But total deference is not warranted on the core questions of whether those interests are truly compelling enough, in a constitutional sense, to justify the expenditure limits, and whether this regulation places an undue burden on the First Amendment rights of those who bring this challenge. See, e.g., Metromedia, Inc. v. City of San Diego, 453 U.S. 490, 519, 101 S.Ct. 2882, 69 L.Ed.2d 800 (1981) (plurality opinion) (“[I]t has been this Court’s consistent position that democracy stands on a stronger footing when courts protect First Amendment interests against legislative intrusion, rather than deferring to merely rational legislative judgment in this area.”); Schneider v. State, 308 U.S. 147, 161, 60 S.Ct. 146, 84 L.Ed. 155 (1939) (“This court has characterized the freedom of speech and that of the press as fundamental personal rights and liberties.... [T]he delicate and difficult task falls upon the courts ... to appraise the substantiality of the reasons advanced in support of the regulation of the free enjoyment of the rights.”).

We read the District Court opinion as consistent with this view. It gives “considerable deference” to the legislative findings on the need for the law only, but not to the legislature’s assessment of whether its solution is narrowly tailored. Cf. Regents of University of California v. Bakke, 438 U.S. 265, 299, 98 S.Ct. 2733, 57 L.Ed.2d 750 (1978) (Powell, J.) (“Political judgments regarding the necessity for the particular classification may be weighed in the constitutional balance, but the standard of justification will remain constant”), quoted in Adarand Constructors, Inc. v. Pena, 515 U.S. 200, 224-25, 115 S.Ct. 2097, 132 L.Ed.2d 158 (1995). This approach is consistent with Justice Breyer’s concurrence in Shrink, where he indicated that the Court should “defer to [the Missouri legislature’s] political judgment that unlimited spending threatens the integrity of the electoral process,” but not with respect to whether “its solution, by imposing too low a contribution limit, significantly increases the reputation-related or media-related advantages of incumbency and thereby insulates legislators from effective electoral challenge.” 528 U.S. at 403-04, 120 S.Ct. 897.

Although we bear in mind Justice Breyer’s observations in Shrink, we cannot adopt his conclusion, in light of the extensive Supreme Court precedent to the contrary, that the interests must be balanced here, and that there is therefore “no place” for a “strong presumption against constitutionality of the sort often thought to accompany the words ‘strict scrutiny.’ ” Id. at 400, 120 S.Ct. 897. Such a presumption is proper, at least until the Supreme Court tells us otherwise, and it means that the burden of persuasion at trial was on the State to defend Act 64 — i.e., to establish that there was a compelling state interest to support the expenditure limit provision and that the provision was narrowly tailored to advance that interest. See Bur-son, 504 U.S. at 226, 112 S.Ct. 1846 (Stevens, J., dissenting, joined by O’Connor and Souter) (noting that “a core premise of strict scrutiny” is that “the heavy burden of justification is on the State”). But this burden does not excuse the courts from actually applying the scrutiny that the First Amendment demands, and the State of Vermont deserves.

Therefore, although we do not question the validity of the factual findings developed by the legislature in support of Act 64,⁹ our system of judicial review provides plaintiffs the opportunity to present competing evidence, assigns to the District Court the responsibility for making findings of fact and conclusions of law after weighing the evidence, and leaves to the Court of Appeals the independent responsibility to assess the legal significance of these factual findings. This responsibility is particularly important here, where, as plaintiffs claim, complete deference to the legislature could “risk such constitutional evils as permitting incumbents to insulate themselves from effective electoral challenge.” Shrink, 528 U.S. at 402, 120 S.Ct. 897 (Breyer, J., concurring).

Put differently, this level of scrutiny is a serious barrier for expenditure limits, but it is not impenetrable. Rather, an independent court must be convinced that the legislature was serving the people’s interest and not its own. See, e.g., Burson, 504 U.S. at 213, 112 S.Ct. 1846 (Kennedy, J., concurring) (discussing the use of the compelling-interest test as “one analytical device to detect, in an objective way, whether the asserted justification is in fact an accurate description of the purpose and effect of the law”), quoted in R.A.V. v. City of St. Paul, 505 U.S. 377, 395, 112 S.Ct. 2538, 120 L.Ed.2d 305 (1992); Cf. Croson, 488 U.S. at 493, 109 S.Ct. 706 (noting in the equal protection context that “the purpose of strict scrutiny is to ‘smoke out’ illegitimate uses of race by assuring that the legislative body is pursuing a goal important enough to warrant use of a highly suspect tool,” with the narrow tailoring analysis helping to ensure that there is “little or no possibility that the motive for the classification was illegitimate”).

C. Compelling Interests

In Shrink, the Court indicated that “[t]he quantum of empirical evidence needed to satisfy heightened judicial scrutiny of legislative judgments will vary up or down with the novelty and plausibility of the justification raised.” 528 U.S. at 391, 120 S.Ct. 897. For example, the Court in Shrink accepted a relatively minimal evi-dentiary showing of Missouri’s interest in preventing corruption or the appearance thereof, because in its view, there was “little reason to doubt that sometimes large contributions will work actual corruption of our political system, and no reason to question the existence of a corresponding suspicion among voters.” Id. at 395, 120 S.Ct. 897; see also McConnell, 540 U.S. at-, 124 S.Ct. at 661 (“The idea that large contributions to a national party can corrupt or, at the very least, create the appearance of corruption of federal candidates and officeholders is neither novel nor implausible.”). Similarly, the Shrink Court relied in large part on Buckley’s conclusion on the fit between contribution limits and the anti-corruption interest, to decide that the contribution limits in Missouri were sufficiently tailored and not “so different in kind as to raise essentially a new issue about the adequacy of the Missouri statute’s tailoring to serve its purposes.” 528 U.S. at 395, 120 S.Ct. 897.

With Shrink’s guidance on “the quantum of empirical evidence needed” in mind, we turn then to the interests asserted by Vermont in support of Act 64’s expenditure limits to assess whether any of the interests might be sufficiently compelling to support this regulation of political speech. Vermont offers five interests that it argues are sufficiently compelling to support the spending limits on candidates: (1) “avoiding the reality and appearance of corruption in elective politics and government”; (2) “assuring] that candidates and officeholders will spend less time fund-raising and more time interacting with voters and performing official duties”; (3) promoting “electoral competition and in protecting equal access to political participation”; (4) “bolsterfing] voter interest and engagement in elective politics”; and (5) “enhancing] the quality of political debate and voters’ understanding of the issues.”

Defendants-intervenors appear to rely primarily on the first two of these interests to support the spending limits — describing those interests as (1) deterring corruption and the appearance of corruption; and (2) permitting candidates and officeholders to spend less time fund-raising and more time interacting with voters and performing duties. Defendants-inter-venors also argue that the third interest asserted by the State — '“protecting political equality” — should be recognized as an “additional basis” to support the spending limits on candidates.¹⁰

We now consider the interests asserted by the defendants.

1. Anti-Corruption

The Supreme Court has recently clarified that the anti-corruption interest, in the campaign finance context, is “not confined to bribery of public officials, but extend[s] to the broader threat from politicians too compliant with the wishes of large contributors.” Shrink, 528 U.S. at 389, 120 S.Ct. 897; see also McConnell, 540 U.S. at-, 124 S.Ct. at 660. Moreover, the Shrink Court reiterated that, in addition to the actual influence of campaign contributions on politicians’ behavior, the perception of corruption was an important part of this compelling state interest because it “could jeopardize the willingness of voters to take part in democratic governance.” Shrink, 528 U.S. at 390, 120 S.Ct. 897 (citing United States v. Mississippi Valley Generating Co., 364 U.S. 520, 562, 81 S.Ct. 294, 5 L.Ed.2d 268 (1961) (democracy works “only if the people have faith in those who govern”)).

In terms of “the quantum of empirical evidence needed,” we note that although the interest in avoiding corruption and the appearance thereof is well-established as sufficiently important in the context of contribution limits, the rejection in Buckley of the anti-corruption interest as a constitutional justification for spending limits dictates the need for considerable evidence to demonstrate that unlimited spending is part of the corruption problem, and that spending limits are a necessary and plausible solution.

In this case, the District Court found that Vermont had proven that the reality and perception of corruption in its political system was a legitimate concern. Specifically, it found that “[e]vidence at trial overwhelmingly demonstrated that the Vermont public is suspicious about the effect of big-money influence over politics,” and “it appears they have reason to feel that way,” 118 F.Supp.2d at 468, concluding that “[t]he record suggested that large contributors often have an undue influence over the legislative agenda.” Id. In light of Shrink, this “undue influence” over the legislative agenda is properly considered part of the anti-corruption interest. See 528 U.S. at 389, 120 S.Ct. 897.¹¹ For the reasons that follow, our independent review of the evidence supports these findings by the District Court.

First, citizens in Vermont have consistently demonstrated a belief that the attention of their public representatives may be available for a price. As a result, public faith in the democratic system has declined. The General Assembly described the effects of a need to raise ever growing amounts of funds: “Robust debate of issues, candidate interaction with the electorate, and public involvement and confidence in the electoral process have decreased as campaign expenditures have increased.” 1997 Vt. Laws P.A. 64 (H. 28) (finding No. 4). “Large contributions and large expenditures by persons or eommit-tees, other than the candidate and particularly from out-of-state political committees or corporations, reduce public confidence in the electoral process and increase the appearance that candidates and elected officials will not act in the best interests of Vermont citizens.” Id. (finding No. 9). At trial, one expert witness, Celinda C. Lake, concluded that “[vjoters are extremely concerned about the influence of special interests in the political process.” In fact, according to polling data, nearly 75 percent of Vermont voters say that ordinary voters do not have enough influence over Vermont politics and government, and more than two thirds believe that “large corporations and wealthy individuals have too much influence.” (expert report of Ce-linda C. Lake).

Testimony by Vermont’s elected officials revealed that this disenchantment and loss of public faith played a critical role in their belief that expenditure limits are necessary. One sponsor of Act 64, Representative Karen Kitzmiller, presented the Vermont House of Representatives with evidence showing that 94 percent of Vermonters believe that too much money is spent in politics, and 76 percent believe that ending private contributions would “reduce the power of special interest groups.” Another state legislator, Gordon Bristol, testified at trial about his concern “about the regular guy on the street, and I think if they feel that candidates are spending a modest amount of money, that they are going to get candidates in there who are representing issues and not a special interest .... ” According to another legislator, citizens have reported that they do not vote because “ ‘[a]ll the big money controls everybody in Montpelier anyways.’ ... They think it’s all wrapped up and that the special interests control it and, quite frankly, they aren’t that wrong.” (testimony of Elizabeth Ready). Another legislator said: “[I]t’s the monied interests that control the process, and that cynicism ... it keeps people from participating, from engaging .... ” (testimony of Donald Hooper).

Second, because of the limited number of campaign contributors and the constant concern of being outspent, candidates and elected officials are significantly influenced in deciding positions on issues by a belief that they are unable to oppose too many special interests, no matter how unpopular, because they will be cut off from funds. The General Assembly described the effects of a need to raise ever growing amounts of funds: “Increasing campaign expenditures require candidates to seek and rely on a smaller number of larger contributors, often outside the state, rather than a large number of small contributors.” 1997 Vt. Laws P.A. 64 (H.28) (finding No. 5). If legislation alienates one major special interest group, officials are reluctant to alienate others because the number of entities and people making political contributions is finite and small. One state legislator admitted that, when considering a piece of legislation, “You have to initially consider it as whether or not you want to risk losing the financial support or, in the worst case, having that financial support go to a primary opponent or to a person who opposes you in a general election.” (testimony of Peter Smith). One candidate recalled being told by another lawmaker: “We’ve already lost the drug money [because of the pharmacy bill], and I don’t need to lose the food manufacture money too. So I’m not going to sign the bill.” (testimony of Cheryl Rivers).

Third, and perhaps most perniciously, the demands of fundraising also affect the behavior of elected officials in the context of agenda-setting, since officials pay attention to which contributor “wants what to happen in terms of language of the bill, in terms of calendaring the bill, in terms of writing the rules.” (testimony of Peter Smith). That same witness also noted that a crucial part of any deliberation on a bill involves speculation about the reaction of contributors because they control the money: politicians are forever asking “what’s the industry position, what’s the union position, what’s — you know, and what they’re talking about is where [is] the money behind the issue, what does the money want, where is the conflict between and among the power brokers.” Senator Rivers testified that campaign contributors, by virtue of their role as contributors, can dominate the attention of party leadership or a committee chair, and thereby influence the legislature’s agenda. In her words, “there is kind of an atmosphere that is created that there is [an] assumption that phone calls [of contributors] will get taken and [their] policy issues will be considered.” Another senator, Elizabeth Ready, recognized that “there is an agenda out there that is pretty much set by folks that are not elected.”¹² Candidates, often with great reluctance, accept the bargain with contributors so that they do not lose large sources of potential fundraising for the “arms race” in which they feel compelled to participate.

The evidence at trial established that candidates for public office rely on special interests for financial support, produced directly or by way of “bundling” smaller contributions from a particular company or industry.¹³ The Buckley Court seemed to assume that many small contributions could not raise the specter of corruption. “If a senatorial candidate can raise $1 from each voter, what evil is exacerbated by allowing that candidate to use all that money for political communication?” 424 U.S. at 56 n. 64, 96 S.Ct. 612 (internal quotation marks omitted). But the reality of campaign financing in Vermont is a far cry from this idyllic vision of political fund-raising, in large part because not every voter has the financial ability to participate by giving campaign contributions. “[T]he average Vermonter has been, to some degree, disenfranchised because the average Vermonter cannot afford the price of admission.” Senate J. of the State of Vt., at 1338 (Biennial Session, 1997)'(statement of William T. Doyle).

Vermont has a compelling interest in safeguarding its political process from such contributor dominance, because it corrupts the process for achieving accessibility and accountability of state officials and candidates. The evidence at trial demonstrated that money — and the special interests that wield it — has a great influence on candidate behavior in Vermont, at the expense of the electorate as a whole, since candidates depend on it in order to run for office. Where access and influence can be bought, citizens are less willing to believe that the political system represents the electorate, exacerbating cynicism and weakening the legitimacy of government power. See Jacobus v. Alaska, 338 F.3d 1095, 1113 (9th Cir.2003) (describing the phenomenon of “access-peddling” and explaining that it “creates a danger of corruption and the appearance of corruption”). The accessibility and accountability of public officials — and the public’s faith that Vermont’s government is accessible and accountable — are fundamental to any democratic system.

In our view, such influence of campaign contributors is pernicious because it is bought. Certain private citizens and organizations should not be given greater access to public office holders — and thus greater influence — on account of those citizens’ ability and willingness to pay for candidates’ campaigns. Even with contribution limits, the arms race mentality has made candidates beholden to financial constituencies that contribute to them, and candidates must give them special attention because the contributors will pay for their campaigns. Quid pro quo corruption is troubling not because certain citizens are victorious in the legislative process, but because they achieve the victory by paying public officials for it.

In short, we believe, based on the District Court’s findings and our own independent review of the record, that Vermont has proven the strength of this interest, and its relationship to unlimited campaign spending. And we believe that the factual record developed by Vermont in support of the anti-corruption interest, through the legislative process and at trial, may be sufficient to distinguish Buckley. Cf Colorado Republican I, 518 U.S. at 617-18, 116 S.Ct. 2309 (plurality opinion) (indicating that “the lack of coordination between the candidate and the source of the expenditure ... prevents us from assuming, absent convincing evidence to the contrary, that a limitation on political parties’ independent expenditures is necessary to combat a substantial danger of corruption of the electoral system.”) (emphasis added); Planned Parenthood of Southeastern Pennsylvania v. Casey, 505 U.S. 833, 863-64, 112 S.Ct. 2791, 120 L.Ed.2d 674 (1992) (citing West Coast Hotel and Broum v. Board of Education as examples of “applications of constitutional principle to facts as they had not been seen by the Court before”). Nonetheless, given Buckley’s holding rejecting the anti-corruption interest as inadequate to support the expenditure limits at issue in that case, we are reluctant to conclude that the same general interest, standing alone, is sufficiently compelling to support Act 64’s expenditure limits. See Buckley, 424 U.S. at 46-48, 96 S.Ct. 612; see also McConnell, 540 U.S. at -, 124 S.Ct. at 647. We turn then to the second interest asserted by defendants.

2. Time Protection

Vermont also submits that it has a compelling interest in “assuring] that candidates and officeholders will spend less time fundraising and more time interacting with voters and performing official duties.” Indeed, the District Court found that “the need to solicit money from large donors at times turns legislators away from their official duties.” 118 F.Supp.2d at 468. The District Court also indicated that the State proved that this concern exists, and that Vermont’s expenditure limits addressed this interest, among others.' Id. at 482-83.

Again, we are mindful of Shrink’s guidance that “[t]he quantum of empirical evidence needed to satisfy heightened judicial scrutiny of legislative judgments will vary up or down with the novelty and plausibility of the justification raised.”¹⁴ 528 U.S. at 391, 120 S.Ct. 897. On this score, the “time protection” rationale has been recognized as compelling, although not in the context of candidate spending limits. Indeed, the Buckley Court considered this interest in assessing, and deemed it sufficiently important to support, the public financing scheme for Presidential election campaigns — a provision it upheld. See 424 U.S. at 96, 96 S.Ct. 612 (“Congress properly regarded public financing as an appropriate means of relieving major-party Presidential candidates from the rigors of soliciting private contributions”) (citing Senate Rep. No. 93-689); 424 U.S. at 91, 96 S.Ct. 612 (“Congress was legislating for the ‘general welfare’ ... to free candidates from the rigors of fundraising.”). See also Republican Nat’l Committee v. Federal Election Comm’n, 487 F.Supp. 280, 284-86 (S.D.N.Y.) (three-judge District Court) (upholding constitutionality of expenditure limits as condition of accepting presidential public financing in part on ground that it would “give candidates the opportunity to lessen the ‘great drain on (their) time and energies’ required by fundraising ‘at the expense of providing competitive debate of the issues for the electorate’”) (quoting Senate Rep. No. 93-689), ajfd mem., 445 U.S. 955, 100 S.Ct. 1639, 64 L.Ed.2d 231 (1980).

Moreover, other circuits have more recently recognized the compelling nature of the time-protection interest in similar contexts. See Rosenstiel v. Rodriguez, 101 F.3d 1544, 1553 (8th Cir.1996), cert denied, 520 U.S. 1229, 117 S.Ct. 1820, 137 L.Ed.2d 1028 (1997) (upholding Minnesota’s voluntary public financing scheme because the government has a compelling interest in reducing “the time candidates spend raising campaign contributions, thereby increasing the time available for discussion of the issues and campaigning”); Vote Choice, Inc. v. DiStefano, 4 F.3d 26, 39 (1st Cir.1993) (holding that statute survives exacting scrutiny because Rhode Island has “a valid interest in having candidates accept public financing because such programs ‘facilitate communication by candidates with the electorate’ [and] free candidates from the pressures of fundrais-ing.”) (quoting Buckley, 424 U.S. at 91, 96 S.Ct. 612). Indeed, the Rosenstiel court determined that it is “well settled” that this interest is compelling. 101 F.3d at 1553 (collecting cases).

The Buckley Court, in determining that the expenditure limits in that case were unconstitutional, alluded to this time-protection interest only in passing. 424 U.S. at 91, 96, 96 S.Ct. 612 (mentioning generally Congress’ desire to relieve political candidates from the “rigors” of soliciting and fundraising); see also Blasi, supra, at 1285-86 & n. 15 (“[D]uring the public and legislative debates that led to the passage in 1974 of mandatory spending limits for congressional races, and during the Buckley litigation which resulted in the invalidation of those limits, candidate time protection was almost wholly ignored as a justification for campaign spending limits.”)- Only Justice White, concurring in part and dissenting in part, observed that imposing “expenditure ceilings” would “ease the candidate’s understandable obsession with fundraising, and so free him and his staff to communicate in more places and ways unconnected with the fundraising function.” Buckley, 424 U.S. at 264-65, 96 S.Ct. 612 (“There is nothing objectionable — indeed it seems to me to be a weighty interest in favor of the provision — in the attempt to insulate the political expression of federal candidates from the influence inevitably exerted by the endless job of raising increasingly large sums of money.”). One commentator explains that “candidate time protection was not at the center of either the reform agenda or the constitutional analysis” because Buckley was decided “[b]efore the advent of pervasive war chests and candidate-PAC merchandizing bazaars.”¹⁵ Bla-si, supra, at 1287.

Plaintiffs argue that this interest is no different than the goal of reducing the “skyrocketing costs of political campaigns,” rejected by Buckley as an insufficiently compelling interest to support expenditure limits. 424 U.S. at 57, 96 S.Ct. 612. The Sixth Circuit agreed in Kruse, reasoning that “[t]he need to spend a large amount of time fundraising is a direct outgrowth of high costs of campaigns. However, because the government cannot constitutionally limit the cost of campaigns, the need to spend time raising money, which admittedly detracts an officeholder from doing her job, cannot serve as a basis for limiting campaign spending.” 142 F.3d at 916-17. We are unpersuaded by this reasoning.

Indeed, we think the language of Buckley, as well as an examination of the Buckley briefs, oral argument, and subsequent commentary from judges and scholars, precludes such an interpretation. In its discussion of the federal campaign expenditure ceilings at issue in Buckley, the Buckley Court explained that the limits “appear to be designed primarily to serve the governmental interests in reducing the allegedly skyrocketing costs of political campaigns,” and cited the statistics put forward by appellees and appellants on how the percentage increase in campaign spending in recent years compared to the rise in the consumer price index, gross national product, and total expenditures for commercial advertising over the same time period. 424 U.S. at 57, 96 S.Ct. 612. The Court concluded that, regardless of the import of such statistics, “the mere growth in the cost of federal election campaigns in and of itself provides no basis for governmental restrictions on the quantity of campaign spending and the resulting limitation on the scope of federal campaigns.” Id. (emphasis added). Particularly in light of the recent statements of three Justices indicating that this “time protection” rationale may be a compelling interest, see supra at 108 (quoting Shrink, 528 U.S. at 409, 120 S.Ct. 897 (Kennedy, J., dissenting)); and Colorado Republican I, 518 U.S. at 649-50, 116 S.Ct. 2309 (Stevens, J., joined by Ginsburg, J., dissenting), we see no reason to read Buckley more broadly than its language indicates.

At trial, Vermont presented powerful evidence concerning the time pressures which the prospect of unlimited expenditures places on candidates for office. In particular, there is strong evidence that unlimited expenditures have compelled candidates to engage in lengthy fundrais-ing in order to preempt the possibility that their political opponents may develop substantially larger campaign war chests. The Vermont General Assembly found that such fundraising by candidates requires an “inordinate[ ] amount of time.” 1997 Vt. Laws P.A. 64 (H. 28) (finding No. 1). The large, and growing, campaign war chests in Vermont have created strong pressures on elected officials to ensure that they can raise funds comparable to any opponent. One witness, former State Senator and Lieutenant Governor Peter Smith, described the “stampede or nuclear arms race mentality that we currently have, which is just keep building the bank because you never know what’s going to happen.” Under the current system, Vermont candidates feel like “you had two races you were running. The first was for the money .... ” (testimony of Donald Hooper).

Although there may be no inherent problem with candidates competing to raise large quantities of funds, the evidence in Vermont is clear that the pressure to raise large sums of money greatly affects the way candidates and elected officials spend their time. Special interests, well placed to take advantage of candidates’ fear of losing this fundraising war, dominate candidates’ time and thereby have been able to exercise substantial control over the information that passes to candidates. They do this by increasingly consuming the opportunities candidates have for meeting with constituent groups and forcing candidates to choose contributors over private citizens who make small or no contributions. This command of available time, inherent in endless fund-raising, drastically reduces opportunities that candidates have to meet with noncontributing citizens.¹⁶

Legislators explained at trial that officials are more likely to return donors’ phone calls. “If I have only got an hour at night when I get home to return calls, I am much more likely to return [a donor’s] call then I would [a non-donor’s] .... [W]hen you only have a few minutes to talk, there are certain people that get access.” (testimony of Elizabeth Ready). A former candidate for Congress and current lobbyist in Vermont, Anthony Pollina, described the process:

[Candidates and policymakers ... can only talk to so many people in a day. They can only respond to so many phone calls. The governor can only have so many meetings in a day. And if in fact large contributors are using their contributions to buy access to the governor or other policymakers ... then that means that the policymaker, the governor and others are not spending their time talking to other people who have not provided other large contributions....

Nor is this just a theoretical concern. One widely reported case involved the differing access that state officials granted to interested groups as the state government considered whether to label milk produced using genetically engineered hormones. Major dairy companies, who in the past had been contributors, were able to arrange meetings with critical state leaders, whereas local farmer organizations that lacked importance as contributors could not arrange similar meetings.

By giving money, contributors “haven’t bought the person, but they have certainly bought a piece of that time there where they have that person’s attention.” (testimony of Elizabeth Ready). Even if candidates receive valuable information during every hour spent fundraising, their time is being controlled by those with campaign cash, and this effect is corrosive. The Vermont legislature considered one article in the Burlington Free Press stating that “[m]oney not only threatens to corrupt the process, it sabotages the political dialogue as well. Candidates spend too much time begging for dollars and too little time talking issues .... ” See Democratic Process Relies on Reform, Burlington Free Press, Oct. 6,1997 at 6A.

Public officials testified at trial that the financial necessity imposed by fundraising, and bred by the “arms race” mentality in campaigns with unlimited spending, requires that elected officials spend time with donors rather than on their official duties. One state Senator testified that legislators have to spend time at party fundraising events to give donors access to elected officials, (testimony of Cheryl Rivers). Another Senator explained how spending limits would affect her time:

If I can go out and raise what I have to raise and know that those limits are in place, I can spend the whole rest of my campaign, once I have raised that money, out with the public, okay. I can go door-to-door. I can go around to local events. I can go to the county fairs. I can have a little booth, you know, and be talking to people. I am not going to be locked away, you know, in the Democratic Party somewhere or in my own office somewhere making fundraising calls, (testimony of Elizabeth Ready).

Simply put, every hour spent drumming up financial contributions is an hour that cannot be spent independently studying legislative proposals or meeting with constituents who may not be likely donors. And the public understands this reality: at trial, Vermont presented survey data that 85% of Vermonters are concerned that “political fundraising took away time from important government business.” (testimony of Celinda C. Lake).

Indeed, although we do not balance interests, the fact that this time-protection interest is itself fundamental to our representative democracy, and related to First Amendment values, cannot be ignored. As' one First Amendment scholar put it, the quality of democratic representation suffers “when legislators continually concerned about re-election are not able to spend the greater part of their workday on matters of constituent service, information gathering, political and policy analysis, debating and compromising with fellow representatives, and the public dissemination of views.” Blasi, supra, at 1282-83.

Unfortunately, without spending limits, the contribution limits would exacerbate the time problem. A lobbyist who supports Act 64 noted that contribution limits coupled with unlimited expenditures would require that candidates “continue to spend more time and energy raising those smaller contributions to see who could raise the most money and outspend their opponent and therefore win the race. So the spending limits, tied to the contribution limits, create a situation where the candidates simply don’t have to spend as much time and energy raising money.... [The limits] change the way campaigns are run, in a sense, and make them more people oriented or voter oriented as opposed to fundraising oriented .... ” (testimony of Anthony Pollina). Cf. McConnell, 540 U.S. at-, 124 S.Ct. at 656 (“The ‘overall effect’ of dollar limits on contributions is ‘merely to require candidates and political committees to raise funds from a greater number of persons.’”) (quoting Buckley, 424 U.S. at 21-22, 96 S.Ct. 612). That same lobbyist also explained- that with contribution limits alone, “the unfortunate thing is that candidates would feel compelled to look for those other sources because they would still be trying to outspend ... their opponents, and that would cause them to then spend more time and more energy into looking for those other sources of funding. It might then encourage the bundling practices that were referred to earlier, and ... it would not address the problem that we are hoping to address.” (testimony of Anthony Pollina).¹⁷

In sum, our independent review of the evidence adduced at trial supports the District Court findings that “the Vermont public perceives, legitimately, that candidates frequently spend an excessive amount of time fundraising and not enough time interacting with voters,” and that “the need to solicit money from large donors at times turns legislators away from their official duties.” 118 F.Supp.2d at 468, 470. So long as the danger remains that a political opponent might severely outstrip a candidate’s financial resources, candidates have continued to feel it necessary to raise ever larger sums of money. For elected officials, this will mean giving more time to contributors over non-contributors, and expending more effort on relatively generous contributors over less important ones.

3. Conclusion: Two Compelling Interests

Faced with this evidence and the resulting findings of the District Court, we conclude that Vermont has established at least two interests in maintaining campaign expenditure limits: preventing the reality and appearance of corruption, and protecting the time of candidates and elected officials. In this case, Vermont’s well-documented interest in time-protection is particularly compelling when considered in tandem with the State’s firmly-rooted interest in preventing corruption (or the appearance thereof). Cf. Miller v. Johnson, 515 U.S. 900, 921, 115 S.Ct.. 2475, 132 L.Ed.2d 762 (1995) (leaving open question whether “compliance with the [Voting Rights] Act, standing alone, can provide a compelling interest independent of any interest in remedying past discrimination”).¹⁸ Regardless of whether one finds Vermont’s justifications novel, the quantum of evidence demonstrating the depth of the problem in Vermont campaigns is great. The drive for campaign funds has created a situation where candidate time is effectively for sale. As a democracy, Vermont has a compelling interest in ensuring that its representatives’ time is not available only — or mostly — to the people who are willing and able to pay for it. Fundamentally, Vermont has shown that, without expenditure limits, its elected officials have been forced to provide privileged access to contributors in exchange for campaign money. Vermont’s interest in ending this state of affairs is compelling: the basic democratic requirements of accessibility, and thus accountability, are imperiled when the time of public officials is dominated by those who pay for such access with campaign contributions.

Because we conclude that Vermont has established two interests that, taken together, are sufficiently compelling to support its expenditure limits, we need not consider the other interests asserted by the State. Specifically, we need not consider whether the interest in encouraging electoral competition and protecting the ability of non-wealthy Vermonters to run for state office in Vermont is sufficiently compelling. And we do not need to reach the question of whether Vermont has sufficiently compelling, independent interests in (1) bolstering voter interest and engagement in elective politics; and (2) encouraging public debates and other forms of meaningful constituent contact in place of the growing reliance on 30-second commercials. We do note that the first of these interests is properly considered part of the anti-corruption interest, and the second relates to the time-protection rationale.

D. Narrow Tailoring

Our analysis next requires a determination as to whether the particular limits are narrowly tailored to serve the compelling interests offered. Because mandatory expenditure limits are so rare, and the Supreme Court and federal courts of appeals that have considered the constitutionality of expenditure limits have found no compelling interests sufficient to support them, no court has reached the narrow tailoring question in this context. Thus, we are in largely uncharted waters.

Plaintiffs argue that even if there is a compelling interest to support Act 64’s spending limits, the spending limits are not narrowly tailored. They argue that the spending limits are too serious an impingement on First Amendment rights, without significantly advancing the interests asserted by the State. Because the District Court held that mandatory spending limits are per se unconstitutional under Buckley, it never fully reached the narrow tailoring inquiry, although it did address the subsidiary question of whether candidates could run effective campaigns under the rubric of “narrow tailoring.” 118 F.Supp.2d at 470-72.

The parties present the narrow tailoring issue as whether the expenditure limits are sufficiently high to allow candidates to run effective campaigns. Indeed, in our initial consideration of this case, we assumed that the parties were correct in focusing the narrow tailoring question on the ability of a candidate to run an “effective campaign.” We now believe, however, that the narrow tailoring inquiry is broader, and that answering the question requires remand to the District Court. We write further to explain the nature of the narrow tailoring inquiry required.

The narrow tailoring inquiry examines the “fit” between means and ends. Here, the question is whether mandatory spending limits will significantly advance the State’s time-protection and anti-corruption interests, without severely burdening the First Amendment rights of the plaintiffs. ‘Where at all possible, government must curtail speech only to the degree necessary to meet the particular problem at hand, and must avoid infringing on speech that does not pose the danger that has prompted regulation.” Federal Election Comm’n v. Massachusetts Citizens for Life, Inc., 479 U.S. 238, 265, 107 S.Ct. 616, 93 L.Ed.2d 539 (1986).

In order to satisfy the “narrow tailoring” standard, the government must also prove that the mechanism chosen is the least restrictive means of advancing that interest. See, e.g., Playboy Entertainment Group, Inc., 529 U.S. at 816, 120 S.Ct. 1878 (“When a plausible, less restrictive alternative is offered to a content-based speech restriction, it is the Government’s obligation to prove that the alternative will be ineffective to achieve its goals.”); Universal City Studios, Inc. v. Corley, 273 F.3d 429, 450 (2d Cir.2001) (“Content-based restrictions are permissible only if they serve compelling state interests and do so by the least restrictive means available.”).¹⁹ When the First Amendment demands strict scrutiny, “[i]f a less restrictive alternative would serve the Government’s purpose, the legislature must use that alternative.” Playboy Entertainment Group, 529 U.S. at 813, 120 S.Ct. 1878; Sable Communications of Cal., Inc., v. FCC, 492 U.S. 115, 126, 109 S.Ct. 2829, 106 L.Ed.2d 93 (1989) (“The Government may ... regulate the content of constitutionally protected speech in order to promote a compelling interest if it chooses the least restrictive means to further the articulated interest.”); see also Boos v. Barry, 485 U.S. 312, 329, 108 S.Ct. 1157, 99 L.Ed.2d 333 (1988) (concluding that government regulation at issue was not narrowly tailored because “a less restrictive alternative is readily available”); California Democratic Party v. Jones, 530 U.S. 567, 585-86, 120 S.Ct. 2402, 147 L.Ed.2d 502 (2000) (observing, in dicta, that California’s “blanket” partisan primary system was not a narrowly tailored means of furthering asserted state interests because “a nonpartisan blanket primary” would advance the same interests “without severely burdening a political party’s First Amendment right of association”).

Accordingly, answering the narrow tailoring question requires addressing three different issues: (1) the extent to which the State’s interests are advanced by the regulation; (2) the extent to which candidates can conduct “effective advocacy” under the limits; and (3) whether the government has proven the absence of less restrictive alternatives that are as effective in advancing its compelling interests, while impinging less on First Amendment rights.

1. Are Vermont’s time-protection and anti-corruption interests advanced by campaign spending caps?

Plaintiffs argue that the spending limits do not actually advance the interests asserted by the State because the limits are set at the equivalent of current levels of spending, and when considered in combination with the contribution limits, actually force candidates and elected officials to spend more time and attention on fund-raising, not less. For the reasons that follow, we disagree and conclude that the spending limits are likely to advance both the time-protection and anti-corruption interests asserted by the State.

First, we do note the apparent tension between seeking to reform the political process by imposing expenditure limits, yet setting limits based on current candidate expenditure patterns in an effort to approximate the spending needs of such candidates. We believe, however, that this tension is more apparent than real. Indeed, plaintiffs’ argument misunderstands the driving force behind the spending limits. The evidence at trial, including the evidence from legislative hearings, indicated the widespread presence of an “arms race” mentality. The record in Vermont demonstrates that often it is this potential of being vastly outspent that creates powerful and deleterious pressures to raise funds. The significance of the spending cap lies not in reducing the amount of money spent on campaigns, but rather in eliminating this potential of being vastly outspent that leads to the “arms race” mentality among candidates and elected officials.

Limiting the arms race promises to have a direct impact on the time of candidates and elected officials. Indeed, the witnesses’ testimony at trial supported the idea that reducing the arms race mentality, spending limits would allow candidates and elected officials to focus more time on issues. One elected official shared her sense of how spending limits will liberate public officials: “[The spending limit] lessens the pressure.... I am not going to be locked away ... in the Democratic Party somewhere or in my own office somewhere making fundraising calls.” (testimony of Elizabeth Ready). Another State Senator, and a sponsor of Act 64, testified that “I would hope that it’s going to give folks running for office more of an opportunity to go out and engage the voters on the issues.” (testimony of Cheryl Rivers). And William T. Doyle, another senator, testified that without the need to raise such large sums of money “there will be increased time for real debate ... candidates will be able to concentrate more on issues rather than raising public money.”

The “arms race” mentality — and its effect on the behavior of candidates and elected officials — is also quite relevant to the anti-corruption interest, and helps explain why the spending caps also address this “threat from politicians too compliant with the wishes of large contributors.” Shrink, 528 U.S. at 389, 120 S.Ct. 897. The evidence presented at trial indicated that the agenda of candidates and elected officials is affected by the perceived need to raise increasing amounts of funds. Because the sources of campaign money are necessarily limited, candidates are reluctant to alienate potential fundraising constituencies. This affects what issues are put on the agenda, what issues are taken off, and how certain issues are addressed. With spending caps, this calculus changes to a certain extent. For example, with a limit on how much money can be spent, elected officials testified that they would be more willing to take a position which a particular industry opposed, (testimony of State Sen. Cheryl Rivers; testimony of former Congressman and Lt. Governor Peter Smith).

Plaintiffs also argue that Act 64’s expenditure limits do not advance the interest in reducing the time dedicated to fundraising because Act 64, as a whole, actually forces candidates to devote more time to fund-raising, not less. Defendants essentially do not dispute that lower contribution limits increase the amount of time that candidates must spend on fundraising. Instead they argue that this makes the interest in time-protection more compelling, not less, with respect to expenditure limits. In essence, plaintiffs argue that rather than address both the anti-corruption interest with contribution limits, and the anti-corruption and time-protection interests with expenditure limits, the State of Vermont must address just one interest, or else resign itself to the state of affairs post-Buckley. After Buckley, when the Court upheld the contribution limits but not spending limits, Congress’ regulatory scheme fell prey to precisely this problem: candidates have been forced to spend increasing amounts of time fundraising under a regime with contribution limits but no spending limits. We reject the notion that Vermont cannot try to address both interests — anti-corruption and time-protection — at once.

Finally, plaintiffs argue, as the Buckley plaintiffs did for contribution limits, that the “limitations work such an invidious discrimination between incumbents and challengers that the statutory provisions must be declared unconstitutional on their face.” 424 U.S. at 30-31, 96 S.Ct. 612. We agree with plaintiffs — and with our dissenting colleague, see post — that election laws, written by legislators who are, at least in part, necessarily self-interested, must be scrutinized for indications that the limits unduly benefit incumbents or otherwise create dangerous distortions of the electoral system. See Shrink, 528 U.S. at 402, 120 S.Ct. 897 (Breyer, J., concurring) (noting the need for courts to scrutinize legislative judgments that “risk such constitutional evils as, say, permitting incumbents to insulate themselves from effective electoral challenge.”). It should be recognized, however, that a legislature’s inaction may maintain such barriers more easily than reforms create them, and review of legislation should not amount to a presumption against, the fairness of spending limits simply because elected officials have an interest in the reforms they are enacting. See Frank I. Michelman, The Constitutional Question, 24 HáRV. J. L. & Pub. Pol’y 17, 22 (2000).

Indeed, there is considerable evidence in Act 64 itself that incumbent protection was not the legislature’s motive. Act 64 permits challengers to outspend incumbents, partially neutralizing the advantages that incumbents often enjoy from free media exposure. Specifically, incumbent candidates for statewide office may only spend 85 percent of the amount permitted challengers. See Vt. Stat. Ann. tit. 17, § 2805a(c). Incumbents in the General Assembly may spend 90 percent. See id. These are rare types of provisions in state campaign finance laws; if the legislature wanted to achieve incumbent protection, it seems unlikely that they would have inserted such provisions. Moreover, defendants presented evidence at trial that the disparity between challenger and incumbent spending is what most frequently disadvantages challengers, and reduces electoral competition — a disparity that spending limits would inevitably reduce, (testimony of Donald Gross). In short, plaintiffs have not demonstrated that challengers will be disproportionately harmed by the spending limits. Like the Buckley court, we see no evidence in the record sufficient to overcome the presumption that “a court should generally be hesitant to invalidate legislation which on its face imposes evenhanded restrictions.” 424 U.S. at 31, 96 S.Ct. 612.

In sum, because Vermont has demonstrated that the time-protection and anti-corruption interests are advanced, and plaintiffs have not succeeded in demonstrating impermissible legislative motives, we conclude that the State has met its burden on this aspect of narrow tailoring— that the spending limits actually advance these asserted interests.

2. Do spending limits at these levels allow for “effective advocacy”?

We must then turn to the question of whether the spending limits prevent “effective advocacy” by limiting the ability of candidates to communicate adequately with voters, and the ability of voters to receive the information they need to make a choice on election day. This “effective advocacy” requirement is drawn from the caselaw on whether contribution limits are sufficiently high. See McConnell, 540 U.S. at-, 124 S.Ct. at 655-56 (“Because the communicative value of large contributions inheres mainly in their ability to facilitate the speech of their recipients, we have said that contribution limits impose serious burdens on free speech only if they are so low as to ‘preven[t] candidates and political committees from amassing the resources necessary for effective advocacy.’ ”) (quoting Buckley, 424 U.S. at 21, 96 S.Ct. 612); Shrink, 528 U.S. at 395-96, 120 S.Ct. 897 (same). Although the concept of “effective advocacy” originated with regard to the freedom of association rights rooted in the First Amendment, see NAACP v. Alabama, 357 U.S. 449, 459-60, 78 S.Ct. 1163, 2 L.Ed.2d 1488 (1958), the Supreme Court has also used this concept in assessing candidates’ claims that campaign finance regulations place too great a burden on their First Amendment speech rights. In Shrink, for example, one of the plaintiffs was a candidate for statewide office who argued that Missouri’s contribution limits prevented him from “amassing the resources necessary for effective advocacy.” 528 U.S. at 396, 120 S.Ct. 897 (quoting Buckley). We believe that the use of this “effective advocacy” standard is appropriate here as a threshold consideration in assessing whether the expenditure limits are “narrowly tailored,” as the parties have argued.²⁰

Discussing the nature of this “effective advocacy” analysis in the context of contribution limitations, the Court asked in part whether the limitation was “so radical in effect” as to “drive the sound of a candidate’s voice below the level of notice.” Shrink, 528 U.S. at 397, 120 S.Ct. 897; see also McConnell, 540 U.S. at -, 124 S.Ct. at 677. The nature of the “effective advocacy” requirement, then, is that of a constitutional minimum; as long as the regulation does not “drive the sound of a candidate’s voice below the level of notice,” based on evidence from past campaigns, then the First Amendment is not violated on this ground.

Although the District Court never reached the legal issue of narrow tailoring, it did make findings as to whether “effective campaigns” could be run under the limits. The District Court found that Vermont’s expenditure limitations reflect the actual cost of running for office in Vermont, would not cause a revolutionary change in campaign spending, and would leave candidates fully capable of conducting effective campaigns. 118 F.Supp.2d at 472. These conclusions are subject to a mixed standard of review, consistent with Rule 52(a) of the Federal Rules of Civil Procedure but also bearing in mind the obligation in First Amendment cases for appellate courts to make an independent examination of the record as a whole. See Bose Corp. v. Consumers Union of the United States, Inc., 466 U.S. 485, 499, 104 S.Ct. 1949, 80 L.Ed.2d 502 (1984); Harte-Hanks Communications, Inc. v. Connaughton, 491 U.S. 657, 688, 109 S.Ct. 2678, 105 L.Ed.2d 562 (1989); Ezekwo v. New York City Health & Hospitals Corp., 940 F.2d 775, 780 (2d Cir.), cert. denied, 502 U.S. 1013, 112 S.Ct. 657, 116 L.Ed.2d 749 (1991).

Based on the data presented at trial through expert witnesses, the District Court found that the average spending in Vermont House district races during the three election cycles preceding the District Court’s opinion was almost uniformly below the limits set pursuant to Act 64. 118 F.Supp.2d at 471. Similarly, multi-member Senate districts all involved average spending below that permitted pursuant to Act 64, with average spending exceeding the Act’s expenditure limits only in single-member Senate districts. Id. In addition to reflecting the actual expenditures in Vermont elections, the District Court found that Act 64’s expenditure limits are also appropriate given the costs of running for office in Vermont. The District Court credited the testimony of a number of fact witnesses who testified to the details of previous campaigns they had run, including a Senate challenger in Chit-tenden County and a former Senate candidate in Rutland County, “confirm[ing] that fully effective campaigns for the Vermont Senate can be run under the limits established by Act 64.” Id. at 472. The court noted that Vermont candidates for legislative office frequently use low-cost campaigning methods, such as community debates, door-to-door campaigning, town barbecues and suppers, advertising placards and the issuance of press releases. Id. Legislative candidates rarely hire campaign staff or purchase expensive mass media. Id. Indeed, the evidence at trial indicated that most Vermont House and Senate candidates do not use television advertising primarily because the lack of congruence between media markets and district boundaries render such advertising an inefficient and ineffective way to communicate with voters. (Ex. AA, Lan-dell Admission # 49; Ex. BB, Randall Admission # 84, # 85; testimony of Neil Randall; testimony of Toby Young; Ex. U-l, Expert Report of Anthony Gierzyn-ski, at 8.) Nonetheless, candidates are able to spend the money needed to ensure that their voice is well above “the level of notice” necessary for “effective advocacy.” Shrink, 528 U.S. at 397, 120 S.Ct. 897.

Although candidates for statewide office utilize more expensive media and techniques, they are permitted to spend larger amounts and can thus also engage in “effective advocacy.” In part, this reflects the particular qualities of Vermont, especially the relatively inexpensive cost of television advertising in the State. 118 F.Supp.2d at 472. In reaching this conclusion, the District Court rejected testimony of plaintiffs’ witnesses that much larger amounts of money — amounts so large that no Vermont candidate has ever spent them — are required to wage an effective campaign for governor or other statewide offices. See id. (“The Court rejects [the witness’s] testimony that it is necessary to spend between $800,000 and $1 million to run an effective campaign for Governor of Vermont ... Nor does the Court accept that candidates must spend approximately $500,000 in order to run effective campaigns for Lieutenant Governor and the other lower statewide offices of Secretary of State, Treasurer, Auditor, and Attorney General.”).

And though conflicting evidence was presented at trial, there was ample evidence aside from the specific statistics and campaigns cited in the District Court opinion to support the District Court’s effective campaign findings. For example, plaintiffs’ evidence emphasized what they described as the problems under the limits for candidates running for Seriate in Chit-tenden County, a county that includes the city of. Burlington, as well as very rural areas. But defendants presented evidence that the average spending in this district was consistently far less than the $16,500 allowed under Act 64. More specifically, defendants presented evidence that in 1994, all six of the victorious candidates in Chittenden County spent at or near the Act 64 spending limits, including two successful challengers. In 1996, all six winners spent at or below the limits, including three successful challengers. And in 1998, three of the six candidates spent below the limits. As to statewide races, one candidate for State Auditor in 2000, Elizabeth Ready, indicated that under the $45,000 limit for her race, she had been able to purchase newspaper and radio ads, and considered using television advertising later in the race.

Plaintiffs may not simply rely on the highest-spending races in order to declare the entire statute unconstitutional on its face. In Shrink, the Supreme Court went as far as to assume the truth of plaintiffs claim that the contribution limits affected his ability “to wage a competitive campaign,” and concluded that nonetheless, “a showing of one affected individual does not point up a system of suppressed political advocacy that would be unconstitutional under Buckley.” 528 U.S. at 396, 120 S.Ct. 897. Indeed, certain plaintiffs have a particularly difficult time arguing that the spending limits will burden their First Amendment rights. Plaintiff Donald Bru-nelle’s maximum expenditure in any of his past House campaigns was $1,007, and plaintiff George Kuusela has never spent more than $1,550. The new limits allow each of them, as challengers, to spend $3,000 in their House campaigns, significantly more than they have spent in the past.

In Shrink, the Supreme Court relied on and quoted the. District Court’s conclusions, made on cross-motions for summary judgment, that “candidates for state elected office [have been] quite able to raise funds sufficient to run effective campaigns,” and that “candidates for political office in the State are still able to amass impressive campaign war chests.” 528 U.S. at 396, 120 S.Ct. 897 (citations omitted). Like the Shrink Court, we affirm the District Court’s conclusion, here made after a 10-day bench trial, that candidates can meet the threshold level of “effective advocacy” when running for office in the State of Vermont. 118 F.Supp.2d at 471-72. After independent review, we agree with the District Court that these expenditure limits are not “so radical in effect” as to “drive the sound of a candidate’s voice below the- level of notice,” Shrink, 528 U.S. at 397, 120 S.Ct. 897, and therefore the limits do not prevent candidates from “amassing thé resources necessary for effective advocacy.” Buckley, 424 U.S. at 21, 96 S.Ct. 612. However, as we address in the next section, the inquiry into how much the spending limits impinge First Amendment rights is broader than it was in the Shrink contribution limits context, and therefore must go beyond merely “effective advocacy.”

.3. Are mandatory ■ expenditure limits the least restrictive means of advancing the State’s interests?

The greater level of scrutiny accorded spending limits, as compared to contribution limits, requires that Vermont must also prove that Act 64’s mandatory expenditure limit system is the least restrictive alternative for achieving the State’s compelling time-protection ■ and anti-corruption interests. This inquiry is essentially twofold. First, the State must prove that the type of regulation chosen was the least restrictive — that is, that no other type of regulation could have advanced the interests asserted while impinging less on First Amendment rights. Second, the least restrictive alternative inquiry requires scrutiny of the basis for the particular spending limits chosen — an inquiry not undertaken with respect to contribution limits. In evaluating the constitutionality of contribution limits, the Supreme Court has indicated that “[i]f it is satisfied that some limit ... is necessary, a court has no scalpel to probe, whether, say, a $2,000 ceiling might not serve as well as $1,000,” noting that “[s]uch distinctions in degree become significant only when they can be said to amount to differences in kind.” Buckley, 424 U.S. at 30, 96 S.Ct. 612. In the context of expenditure limits, and in light of the legislative history of Act 64, there may well be “differences in kind” as to the choice of specific spending limits that demand scrutiny.

a. Type of Regulation

Vermont argues that it “had already explored less restrictive alternatives and found them to be ineffective.” Specifically, in 1993, the State instituted a system of voluntary expenditure limits. Former Vt. Stat. Ann. tit. 17, §§ 2841-12 (1991) (repealed 1997). Under this system, candidates had the option of signing an affidavit indicating that they would comply with the spending limits and, in exchange, received any favorable or unfavorable publicity from that decision. At trial, Vermont presented evidence that the voluntary spending limits in place from 1993 to 1997 were not working. Participation in the voluntary system fell dramatically each year, with 90 percent of candidates participating in the first year but less than 20 percent in the second year. Witnesses testified that Vermont’s voluntary limits did not work because candidates attempted to gain an advantage in the fundraising “arms race” by ignoring the limits. Most tellingly, by 1998, not a single candidate for statewide office chose to abide by the voluntary expenditure limits.

Vermont’s voluntary limits were quite different, however, than many of the voluntary spending limits in other states.²¹ In Vermont, the only “carrot” exchanged for the voluntary agreement to the limits was the ability to publicize one’s compliance. Most other states (and New York City) offer additional incentives such as public matching funds, a higher contribution limit, or other inducements. See, e.g., Writing Reform: A Guide to Drafting State and Local Campaign FINANCE Laws, V-8-19 (Deborah Goldberg ed., Brennan Center 2001) (describing variety of mechanisms used by states to encourage compliance with voluntary limits). And many of these types of provisions have been upheld by the federal courts of appeals in the face of a First Amendment challenge. See e.g. Rosenstiel v. Rodriguez, 101 F.3d 1544, 1552-53 (8th Cir.1996) (upholding Minnesota’s voluntary public financing scheme), cert. denied, 520 U.S. 1229, 117 S.Ct. 1820, 137 L.Ed.2d 1028 (1997); see also Vote Choice, Inc. v. DiStefano, 4 F.3d 26, 39 (1st Cir.1993) (holding that Rhode Island’s voluntary public financing statute survives exacting scrutiny).

Indeed, this type of spending limit— voluntary, but with public funding as an inducement to comply — was in Act 64 as originally introduced in the House, but the provision appears to have been changed to mandatory in one of the House committees. (Ex. A, at E-0001, 0031.) It is unclear from the current record why this change occurred, and the District Court made no findings on this point. Similarly, it is possible that the Vermont legislature could have employed a public financing option for all offices, in addition to the option that was provided for Governor and Lieutenant Governor candidates. See, e.g., GeneRal Accounting Offioe, Campaign Finance Reform: Early Experienoes of Two States That Offer Full Public Funding for Political Candidates (May 9, 2003) (GAO-03-453) (assessing initial results of public funding of legislative candidates in Maine and Arizona). Notably, early versions of the bill appear to have contained such funding for a broader range of offices. (testimony of Karen Kitzmiller; Ex. 70, at E-2828.) On the other hand, after the voluntary affidavit system broke down, the legislature may have concluded that only mandatory expenditure limitations, together with contribution limitations, would adequately advance the State’s interests. Indeed, there may have been still other reasons — not yet made part of the record' — -for the legislature’s decision not to adopt, for all offices, voluntary expenditure limits with public funding incentives.

If Vermont could have utilized some of the same voluntary mechanisms employed by other states, offering either financial or other incentives for compliance with expenditure limits, then Vermont may not be able to prove that it employed the least restrictive alternative. Cf. Denver Area Educational Telecommunications Consortium v. FCC, 518 U.S. 727, 758, 116 S.Ct. 2374, 135 L.Ed.2d 888 (1996) (“[W]e can take Congress’ different, and significantly less restrictive, treatment of a highly similar problem at least as some indication that more restrictive means are not ‘essential’ (or will not prove very helpful).”) (emphasis in original). On remand, the District Court should make findings as to whether there were less restrictive alternatives available that could have been as effective in advancing the asserted interests. See Wygant v. Jackson Bd. of Ed., 476 U.S. 267, 280 n. 6, 106 S.Ct. 1842, 90 L.Ed.2d 260 (1986) (alternatives should serve the interest “ ‘about as well’ ”). Should there be proven another type of regulation that would have similarly advanced the interests asserted, while impinging less on First Amendment rights, the District Court will have a basis to find that the provision is not narrowly tailored.

b. Basis for Spending Cap Limits

If the District Court finds, however, that mandatory spending limits were the only type of regulation that could sufficiently advance Vermont’s time-protection and anti-corruption interests, then the court must also inquire into the basis for the particular amount of the spending limits chosen. There was evidence presented at trial that Act 64’s spending limits were intended to map onto existing spending levels from past campaigns. In defendants’ proposed findings of fact after trial, defendants indicated that the legislature “considered factors such as population, historical information on past spending levels and campaigns, Vermont’s previous voluntary spending limits, and the testimony of numerous witnesses concerning the appropriate level for the limits, with the legislature balancing different viewpoints as is necessary with most pieces of legislation.” This is consistent with the District Court’s findings that the average spending on past races was consistently under the limits imposed by Act 64. 118 F.Supp.2d at 471-472. See also Richard Briffault, Nixon v. Shrink Missouri Government PAC: The Beginning of the End of the Buckley Era?, 85 Minn. L. Rev. 1729, 1769 (2001) (suggesting that median spending levels of candidates in recent races could be an appropriate basis for spending limits).

Nonetheless, the specific basis for the final limits is not entirely clear from the existing record. Defendants noted that for the Governor’s race, the Senate bill would have set a $250,000 limit, while the House bill originally set a $400,000 limit. In the final bill as adopted, the limit was set at $300,000. Similarly, the limits for one-seat Senate races varied in the various committee bills, either $4,000, $5,000 or $6,000, with an additional $2-3,000 for each additional seat in the district. The final limits were those that emerged out of the Senate Finance Committee (the lowest in any of the committee bills) — $4,000 plus an additional $2,500 for each additional seat. Such decisions, of course, could be the product of typical legislative compromise— a process into which we will not intrude. However, plaintiffs also presented evidence that the ultimate decision as to the amount of certain statewide spending limits was motivated by a desire to preserve the public fisc, because of Act 64’s public financing option for gubernatorial and Lieutenant Governor candidates.²² Particularly in the context of expenditure limits, these choices demand greater scrutiny.

Thus, in undertaking its “least restrictive means” analysis on remand, the District Court should make additional findings on the impact of the legislative choices as to the appropriate spending limits on candidates’ and voters’ First Amendment rights. Even if Act 64’s spending limits are sufficiently high to permit “effective advocacy” as defined in Shrink, as the District Court found and as we have affirmed, the precise level of the limits chosen can have a significant impact on how “restrictive” the provision is on First Amendment rights. Of course, a $2 million limit for State Senate races is quite unlikely to restrict First Amendment rights, but equally unlikely to impact the “arms race” mentality and thereby advance the interests asserted. Moreover, there will always be “distinctions in degree” that could be seen as less restrictive — $305,000 is “less restrictive” than $300,000; $310,000 “less restrictive” than $305,000, etc. And such an inquiry has no logical endpoint.

However, the specific choices made by the Vermont legislature among the different spending limits contained in various bills may be “differences in kind” with respect to the impact on the First Amendment rights of candidates and voters. Inquiries into the First Amendment impact of challenged regulations have been undertaken in analogous contexts by other courts, see, e.g., National Black Police Ass’n v. D.C. Board of Elections and Ethics, 924 F.Supp. 270, 277 (D.D.C.1996) (considering a variety of factors in assessing whether the reduction in campaign funds was “so substantial that it affected the candidates’ ability to reach voters”), vacated as moot sub nom. National Black Police Ass’n v. District of Columbia, 108 F.3d 346 (D.C.Cir.1997), and, although necessarily speculative here, should be undertaken on remand. If the choice of the lower spending limit — for example, $4,000 for a Senate race as opposed to $6,000 — is significantly “more restrictive,” while no more effective in advancing the interest asserted, then the lower spending limit is not consistent with the First Amendment. Assessing the legislative alternatives, then, requires evaluating both (1) the extent to which the higher spending limit is “less restrictive” of the First Amendment rights of candidates and voters²³; and (2) the extent to which the higher spending limit would be as effective in advancing the anti-corruption and time-protection interests. The answers to these questions will determine whether the spending limits chosen not only allow for “effective campaigns” but also are the “least restrictive alternative,” and therefore narrowly tailored.

E. Conclusion: Remand for Further Findings

Quite simply, the District Court found that effective campaigns can be waged under Act 64’s stated limits, and we agree. But the District Court did not examine, because it found spending limits per se unconstitutional, whether the legislature might have chosen either another type of regulation besides mandatory spending limits, or higher limits, that would still achieve the goals we sanction and yet impinge less on the First Amendment rights of candidates and voters.²⁴ Cf. Turner Broadcasting System, Inc. v. FCC, 512 U.S. 622, 668, 114 S.Ct. 2445, 129 L.Ed.2d 497 (1994) (remanding, after grant of summary judgment, “to permit the parties to develop a more thorough factual record, and to allow the District Court to resolve any factual disputes remaining, before passing upon the constitutional validity of the challenged provisions” in part because “the record fails to provide any judicial findings concerning the availability and efficacy of ‘constitutionally acceptable less restrictive means’ of achieving the Government’s asserted interests”) (quoting Sable Communications of Cal., Inc., v. FCC, 492 U.S. 115, 129, 109 S.Ct. 2829, 106 L.Ed.2d 93 (1989)). As to the amount of the limits, there obviously comes a point where limits would be set so high as to have no impact on the interests sought to be protected. We need, however, to hear from the District Court on this fact-intensive question of whether that point is as set in Act 64 or appreciably higher.

On remand, the District Court ought consider, along with any other issues relating to narrow tailoring that it and the parties deem relevant: (1) what alternatives were considered by the legislature, including both alternative types of regulations and alternative amounts for the limits; (2) why these alternatives were rejected; (3) whether and how these alternatives would impinge less on First Amendment rights; and (4) whether the alternatives would be as effective as the mandatory spending limits in advancing the time-protection and anti-corruption interests.²⁵ Based on its fact-finding on these issues, the District Court will be able to draw a legal conclusion as to whether Vermont’s legislature chose the “least restrictive alternative” for advancing its interests, and therefore whether the expenditure limits are narrowly tailored.²⁶

With respect to treating related expenditures as candidate expenditures, see § 2809(b), the District Court did not analyze this question after deciding that candidate expenditure limits were unconstitutional per se. On remand, independent of the constitutionality of expenditure limits, the District Court should evaluate this issue.

Because a content-based regulation of speech such as this is presumptively invalid, and the District Court held it unconstitutional, we leave in place the injunction against enforcement of these provisions pending further proceedings consistent with this opinion.²⁷

II. Act 64’s Contribution Limitations

Act 64 also imposes four basic types of contribution limitations. First, contributions by individuals to candidates are limited to $200 for state representative and other local offices, $300 for state senator and other county offices, and $400 for statewide office. See Vt. Stat. Ann. tit. 17, § 2805(a). Second, PACs and political parties may not accept contributions from a single source in excess of $2000 and are subject to the individual contribution limits when contributing to candidates. See id. Third, individuals, PACs or political parties that make “related expenditures” with candidates must count those expenditures toward the relevant expenditure and contribution limits. See id. § 2809(a)-(c). Finally, candidates, PACs, and political parties may not accept more than 25 percent of their total resources from out-of-state sources. See id. § 2805(c).

A. Limitations on Contributions by Individuals to Candidates

The contribution limits of $200 (state representative), $300 (state senator), and $400 (statewide office) are subject to a lesser degree of scrutiny than expenditure limits, as explained most recently by the Supreme Court in Shrink, 528 U.S. at 386, 120 S.Ct. 897. Contribution limits can survive “if the Government demonstrated that contribution regulation was closely drawn to match a sufficiently important interest, though the dollar amount of the limit need not be fíne tuned.” Id. at 387-88, 120 S.Ct. 897 (internal quotation marks omitted).

The governmental interest in eliminating actual and apparent corruption is sufficient to support Vermont’s limits on contributions to candidates. The Buckley Court upheld limitations of $1000 on contributions to candidates for federal office on the strength of this interest alone. “It 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.... ” 424 U.S. at 26, 96 S.Ct. 612; see also McConnell, 540 U.S. at-, 124 S.Ct. at 647 (explaining that in Buckley, the Court “determined that limiting contributions served an interest in protecting ‘the integrity of our system of representative democracy’ ”) (quoting Buckley, 424 U.S. at 26-27, 96 S.Ct. 612); Jacobus v. Alaska, 338 F.3d 1095, 1107 (9th Cir.2003) (“[A] failure to regulate the arena of campaign finance allows the influence of wealthy individuals and corporations to drown out the voices of individual citizens, producing a political system unresponsive to the needs and desires of the public, and causing the public to become disillusioned with and mistrustful of the political system.”). In this case, the District Court relied on trial testimony, citizen polls, comments by public officials and media coverage to demonstrate the real and perceived threat of corruption in Vermont. 118 F.Supp.2d at 469, 478. As the District Court concluded, “[t]he threat of corruption in Vermont is far from illusory.” Id. at 478.

In addition, we conclude that the Vermont limits are “closely drawn” to this anti-corruption interest. The District Court’s findings in this respect are reasonable and based on the evidence adduced at trial. Id. at 470, 478-80. The District Court relied in part on expert testimony indicating that over the last three election cycles, less than 10 percent of contributions exceeded the limits set by the Vermont legislature. Id. at 478. Based on testimony by both plaintiffs’ and defendants’ witnesses, the District Court also concluded that the limitations approximated amounts “considered suspiciously large by the Vermont public.” Id. at 479-80. And finally, the court compared the Vermont law to similar limits upheld in Maine and Missouri. In Maine, a limit of $250 for House and Senate candidates was upheld. See Daggett v. Comm’n on Governmental Ethics & Election Practices, 205 F.3d 445, 459 (1st Cir.2000). In Missouri, limits of $1075, $525, and $275, depending on the size of the electoral district have been upheld. See Shrink, 528 U.S. at 396-98, 120 S.Ct. 897, remanded to Shrink Mo. Gov’t PAC v. Adams, 204 F.3d 838, 840-42 (8th Cir.2000).²⁸

The contribution ceilings are also sufficiently high to permit effective campaigning. Overly restrictive contribution limits might “have a severe impact on political dialogue if the limitations prevented candidates and political committees from amassing the resources necessary for effective advocacy.” Buckley, 424 U.S. at 21, 96 S.Ct. 612. Contribution limits, however, need not be perfectly set: the failure of the legislators to “engage in such fine tuning does not invalidate the legislation.” Id. at SO, 96 S.Ct. 612. As we have indicated, “distinctions in degree become significant only when they can be said to amount to differences in kind.” Id. We agree with the District Court’s conclusion that the contribution limits imposed by Act 64 do not “amount to differences in kind.”

As the District Court found, the limits imposed by Vermont hardly overwhelm the ability of candidates to engage in active and effective campaigning. The District Court marshaled evidence to support its findings, and conducted a fact-intensive analysis of what constitutes effective campaigning. 118 F.Supp.2d at 478-79. Moreover, Vermont has actually conducted an election since the imposition of these contribution limits (for Mayor of Burlington), and that election involved effective campaigns despite the contribution limitations. Id. at 471, 479. Subject to the applicable limits imposed by the statute, the mayoral candidates raised funds comparable to the amounts spent in State Senate races in the past. Id. The District Court further concluded that the limits may actually improve the ability of candidates to campaign, by freeing candidates from the time-consuming task of “wooing big donors.” Id. at 480.

B. Limitations on Contributions to and by PACs and Political Parties

Act 64 also regulates the ability of PACs and political parties to give and receive contributions. The Act prohibits such organizations from accepting contributions of more than $2000 from a single source during any two-year general election cycle. See Vt. Stat. Ann. tit 17, § 2805(a). The Act further prohibits those organizations from making contributions to political candidates in excess of the general contribution limits — $200 for state representatives or local office, $300 for state senator or county office, and $400 for statewide office. See id. §§ 2805(a)-(b).

The District Court upheld these limitations, except as applied to contributions by political parties to their own candidates. 118 F.Supp.2d at 486-87. Upon review, we hold that all of these limitations are constitutional. We thus affirm the judgment of the District Court as to the constitutionality of most of the limitations, but reject the District Court’s conclusion that political parties cannot be prohibited from contributing to candidates in excess of generally applicable limitations. We discuss three narrow issues that require further attention and, as to two of those issues, further proceedings before the District Court.

We first consider the issue of the $2000 limitation on contributions to political committees or political action committees and political parties. Act 64 defines “political committees” or “political action committees” as “any formal or informal committee of two or more individuals, not including a political party, which receives contributions or makes expenditures of more than $500.00 in any one calendar year for the purpose of supporting or opposing one or more candidates, influencing an election or advocating a position on a public question, in any election or affecting the outcome of an election.” Vt. Stat. Ann. tit 17, § 2801(4). A political party is defined as including both the central party apparatus and “any committee established, financed, maintained or controlled by the party, including any subsidiary, branch or local unit thereof and including national or regional affiliates of the party.” Id. § 2801(5).

Perhaps the most typical application of these rules would involve contributions to a political committee or political party that participates in the political process either by making contributions to or coordinated expenditures with candidates for office. As applied to these organizations, the $2000 limitation is unquestionably constitutional. Political action committees “derive rights from their members” and are accordingly due First Amendment protection. Colorado Republican II, 533 U.S. 431, 448 n. 10, 121 S.Ct. 2351, 150 L.Ed.2d 461. It is well established, however, that the state interest in fighting corruption, real and apparent, justifies limitations on contributions by individuals to particular candidate committees. Such a state interest is equally capable of justifying limits on contributions made to political parties or committees.

If the First Amendment rights of a contributor are not infringed by limitations on the amount he may contribute to a campaign organization which advocates the views and candidacy of a particular candidate, the rights of a contributor are similarly not impaired by limits on the amount he may give to a multicandidate political committee ... which advocates the views and candidacies of a number of candidates.

Cal. Med. Ass’n v. Federal Election Comm’n, 453 U.S. 182, 197, 101 S.Ct. 2712, 69 L.Ed.2d 567 (1981) (“CMA”).

The plaintiffs do not dispute that, in principle, such limitations may be constitutional. Instead, they argue that Vermont’s chosen limitations are overbroad, both because the statute applies to too many organizations and because it sets the contribution ceiling too low.

Regarding the first point, the plaintiffs assert that the restriction is an “over-broad, blunderbuss approach that punishes” even those organizations that are unlikely to corrupt the political process. They imply that certain types of PACs, “particularly legislative leadership PACs or ideological PACs,” pose a weaker danger of corruption and should therefore be permitted greater latitude in determining how to allocate their contributions. The plaintiffs argue that the limits are unconstitutional because Vermont has not shown any independent evidence that political parties and PACs have a negative or deleterious effect on Vermont’s politics. Further, the plaintiffs contend, these organizations are even less likely to corrupt in light of Act 64’s other limitations on campaign financing. Private individuals cannot, for example, effectively funnel large gifts through political parties because parties can themselves only make contributions to candidates of between $200 and $400.

An argument identical to the plaintiffs’ overbreadth argument was addressed and rejected in Buckley. There, the appellants argued that many large contributors have no interest in corrupting the political process, and the law was overbroad for restricting the rights of these unthreatening contributors. The Supreme Court upheld the constitutional validity of generally applicable contribution limits of $1000, even though “most large contributors do not seek improper influence over a candidate’s position or an officeholder’s action.” Buckley, 424 U.S. at 29, 96 S.Ct. 612. The Court reasoned that the very corruption rationale which provides a foundation for the constitutional validity of contribution limitations supports their general applicability: “Not only is it difficult to isolate suspect contributions, but, more importantly, Congress was justified in concluding that the interest in safeguarding against the appearance of impropriety requires that the opportunity for abuse inherent in the process of raising large monetary contributions be eliminated.” Id. at 30, 96 S.Ct. 612.

These arguments were also rejected by the Supreme Court in CMA, 453 U.S. at 197, 101 S.Ct. 2712. In that case, a California political action committee challenged a $5000 federal limit on annual contributions by individuals and associations to multicandidate political committees. See id. at 186, 101 S.Ct. 2712. Like the plaintiffs here, the parties in CMA asserted that such limitations do not serve the government’s strong interest in preventing actual or apparent corruption in the political process. See id. at 197, 101 S.Ct. 2712. The Supreme Court concluded that “Buckley precludes any argument” that the government may not limit the size of contributions made to multicandidate committees, and rejected the assertion that such limitations do not further the government’s interest in battling political corruption. Id. Without such limitations, individuals could evade the contribution limitation “by channeling funds through a multicandidate political committee.” Id. at 198, 101 S.Ct. 2712.

In light of these prior holdings, we are unpersuaded by the plaintiffs’ contention that Vermont had an obligation to divine which PACs and political parties pose the most serious risk of corruption, and develop a record that donations to each type of organization, narrowly defined, pose a strong threat of corruption. It is clear that, in principle, such limitations are an “appropriate means ... to protect the integrity of the contribution restrictions upheld ... in Buckley.” CMA, 453 U.S. at 199, 101 S.Ct. 2712. Thus, the Vermont provision is constitutional so long as the danger of corruption of the political system exists. Just as individuals may be limited from directly contributing to campaign organizations, individuals may be limited from doing so indirectly—that is, contributing large sums to PACs or political parties that funnel money to candidates. See Buckley, 424 U.S. at 38, 96 S.Ct. 612; cf. Federal Election Comm’n v. Beaumont, 539 U.S. 146, 155, 123 S.Ct. 2200, 156 L.Ed.2d 179 (2003) (explaining that “restricting contributions by various organizations hedges against, their use as conduits for ‘circumvention of [valid] contribution limits’”) (citation omitted; alteration in original). Vermont does not have the burden to show on a contributor-by-contributor basis that contributions have led to corruption.

The plaintiffs’ second argument is that the $200, $300, and $400 limits on contributions to candidates for office are unnecessarily low, and that political parties and PACs should be exempt. The plaintiffs in Buckley also raised this argument, contending that the $1000 limitation regulated more contributions than necessary to accomplish its anti-corruption goals. Specifically, the appellants argued that even contributions of a larger amount did not carry a risk of corruption because no politician would throw away a career and reputation for a $1000 donation. As with the earlier overbreadth argument, the Supreme Court also has rejected this contention. See Buckley, 424 U.S. at 30, 96 S.Ct. 612. “[I]f it is satisfied that some limit on contributions is necessary, a court has no scalpel to probe, whether, say, a $2,000 ceiling might not serve as well as $1,000.” Id. (quotations and citations omitted). The Court reaffirmed the validity of this approach in Shrink, stating that a contribution limit survives scrutiny only if the regulation is “closely drawn to match a sufficiently important interest, though the dollar amount of the limit need not be fine tunfed].” 528 U.S. at 387-88, 120 S.Ct. 897 (citations and internal quotation marks omitted; alteration in original); see also Montana Right to Life Assoc. v. Eddleman, 343 F.3d 1085, 1095 (9th Cir.2003) (same).

In order to succeed, then, plaintiffs must establish that when the limitations are applied to political parties and political action committees, they impose such a severe burden that it results in a “difference!;] in kind” from alternative limits. Buckley, 424 U.S. at 30, 96 S.Ct. 612. In other words, a party seeking a special exemption from such laws carries a large burden. Illustrative of the political parties’ and political action committees’ burden in this regard is Federal Election Comm’n v. Massachusetts Citizens for Life, Inc., 479 U.S. 238, 107 S.Ct. 616, 93 L.Ed.2d 539 (1986) (“MCFL”). In that case, the Supreme Court considered the constitutionality of a federal law which bans corporations from making any political expenditures from general corporate funds. The statute’s purpose was to regulate “the corrosive influence of concentrated corporate wealth.” Id. at 257, 107 S.Ct. 616. The Federal Election Commission had sought enforcement of the provision against an incorporated, non-profit pro-life advocacy organization that had “features more akin to voluntary political associations than business firms.” Id. at 263, 107 S.Ct. 616. The Court held that, as applied, the provision was unconstitutional because the stated interest does not apply to an incorporated association like MCFL. Id. at 263-64, 107 S.Ct. 616. The Court set forth specific and demanding criteria for determining when other corporations fall into this constitutionally mandated exclusion, which the advocacy organization was able to meet. Id.

We should expect that the plaintiffs here bear a similar burden of establishing their exceptionalism, even if the particular facts of MCFL do not apply. Unlike the situation in MCFL, the PACs here have offered no evidence that PACs and political parties have overriding features exempting them from the general findings about actual and apparent corruption in Vermont. Nor have they provided evidence that the limitations, when applied to these organizations, impose such a severe burden on speech as to constitute a difference in kind. As mentioned, the District Court concluded, after considering a large body of evidence, that the contribution limits are high enough so that they do not constitute a severe infringement — a difference in kind — of the ability to associate politically. 118 F.Supp.2d at 476-81. We thus agree with the judgment of the District Court and find that Act 64’s contributions limits on political action committees and parties are constitutional.

The District Court did find support for one exception to the candidate contribution limits: those made by political parties. In this regard, we reject the District Court’s conclusion that on account of their “unique role in the mechanics of our democracy,” political parties must have greater freedom to provide their candidates with financial support. Id. at 486. Relying on the central place of political parties in elections, the District Court held that the generally applicable limits were too severe when applied to parties. This was despite the fact that the District Court had already concluded that candidates can receive sufficient funds to effectively exercise their First Amendment rights even when restricted by Act 64’s contribution limits. Nevertheless the District Court held that “[s]uch limits would reduce the voice of political parties to an undesirable, and constitutionally impermissible, whisper.” Id. at 487.

We see no other way to understand the District Court’s position than as being founded on the belief that political parties operate as specially protected institutions under our Constitution and thus merit special treatment. Whatever the validity of this principle in other legal contexts, the Supreme Court has recently left no doubt that parties do not deserve a special exemption from generally applicable contribution limits. See Colorado Republican II, 533 U.S. at 480-82, 121 S.Ct. 2351. In that case, the Colorado Republican Party challenged the constitutionality of restrictions on expenditures it made in coordination with candidates for office, arguing that “coordinated spending is essential to parties because a party and its candidate are joined at the hip, owing to the very conception of the party as an organization formed to elect candidates.” Id. at 477, 121 S.Ct. 2351 (citations and quotation marks omitted). The Court held that such limitations are a constitutional mechanism for ensuring that contributors do not circumvent the federal contribution limit and rejected the claim that political parties occupy some special place in our constitutional system. Above all, the argument fails because, just as with other political organizations, political parties “are necessarily the instruments of some contributors whose object is not to support the party’s message or to elect party candidates across the board, but rather to support a specific candidate for the sake of a position on one, narrow issue, or even to support any candidate who will be obliged to the contributors.” Id. at 479, 121 S.Ct. 2351. Thus, as it does with any other contributor to political campaigns, the government has an interest in restricting the flow of money from parties to candidates in order to reduce actual and apparent corruption. “The Party’s arguments for being treated differently from other political actors subject to limitation on political spending under the Act do not pan out.” Id. at 481, 121 S.Ct. 2351. Since we agree with the District Court’s conclusion that Vermont’s limits are “vital to deter avoid-anee of the individual contribution limits,” 118 F.Supp.2d at 487, we hold that their application to political parties is supported by this strong governmental interest.

Having concluded that the restriction of contributions from political parties is supported by a constitutionally sufficient governmental interest, we turn to the question of whether the statute is sufficiently tailored to this interest. As discussed above, the District Court reviewed the limits based upon data reflecting the costs of elections and the views of citizens regarding what constitutes suspiciously large gifts. Based on this body of evidence, the District Court concluded that gifts in excess of the limits create the appearance of, and increase the likelihood of, corruption. Moreover, contributions in the amounts permitted by the Act provide citizens an adequate tool for “speaking their mind” by giving a donation in order to affiliate with a candidate. Id. at 478-80.

However, there are three narrower issues that require more individual attention. The first concerns the Act’s definition of local and state party affiliates as a single entity. For the purposes of determining whether a political party has exceeded its various contribution limitations, Act 64 defines a political party as “any committee established, financed, maintained or controlled by the party, including any subsidiary, branch or local unit thereof and including national or regional affiliates of the party.” Vt. Stat. Ann. tit. 17, § 2801(5). Vermont’s Secretary of State has interpreted this provision, in conjunction with Vt. Stat. Ann. tit. 17, §§ 2301-2320, to require that state and local branches of political parties be considered a single unit for the purposes of applying contribution limits, and determining whether those limits have been reached or violated.

Plaintiff Vermont Republican State Committee argues that this definition requires the party to treat itself as a single monolithic unit, and requires the party to abandon its current, decentralized structure. However, the plaintiff has not cited any actual changes that will need to be made, except that local and state affiliates will now have to record and coordinate their contributions. In other words, the provision does not impose any organizational burden on the party outside of the campaign finance realm, and requires no broader organizational reform. Moreover, the District Court indicated doubt as to whether the Republican Party actually demonstrated that it operates in the decentralized form that it claims. For example, the state committee brought suit on behalf of all of the town and county committees without ever consulting them or asking them to approve the lawsuit. 118 F.Supp.2d at 487-88. The District Court also noted that federal election law treats state, county, and town committees as a single unit for the purposes of campaign finance. Id. We agree with the District Court that, insofar as Vermont’s campaign finance law treats state and local affiliates as a single entity, it suffers from no constitutional defect.

Second, the plaintiffs have argued that Act 64 applies to even those political action committees that make wholly independent expenditures. Plaintiff Vermont Right to Life Committee-Fund for Independent Political Expenditures (“VRLC-FIPE”), which is affiliated with the Vermont Right to Life Committee (‘VRLC”), is a political committee that, by its charter, cannot make contributions to candidates. It has asserted that it makes only independent expenditures, that is, it never coordinates its expenditures with candidates for office. Thus it argues that when applied to itself, the $2000 cap operates as a limitation on independent expenditures.

The statute does appear to lend itself to such an interpretation. On the one hand, the Act explicitly states that it does not apply to independent expenditures. The law explicitly states that “[t]he limitations on contributions ... shall not apply to contributions made for the purpose of advocating a position on a public question, including a constitutional amendment.” Vt. Stat. Ann. tit. 17, § 2805(g). Nonetheless, it appears that VRLC may be correct that even political organizations that make solely independent expenditures, but nonetheless advocate the election of particular candidates, would be covered. See id. § 2801(4).

Thus, we remand for findings on the following points: (1) whether plaintiff VRLC makes solely independent expenditures and thus has standing to challenge this provision; (2) whether the Vermont law actually restricts independent expenditures by such organizations; and (3) whether Vermont has a sufficiently strong governmental interest in regulating PACs that do not coordinate their expenditures with candidates for office.

Finally, we remand for additional proceedings on the issue of how Act 64 implicates the ability of a state and local party affiliates to receive funds from national affiliates. Act 64 apparently limits the transfer of money from national to state and local parties, and that limit might impose a significant burden on political parties. Vt. Stat. Ann. Tit. 17, §§ 2801(5), 2805(a)-(b). How a party allocates money between its national, state and local affiliates constitutes an important component of party organization. It determines who in the party exercises decision-making authority, who speaks for the party, and how the party arranges its internal finances. At the same time, the failure to limit such transfers might create a loophole that would easily enable contributors to circumvent the $2000 limit on gifts to state parties. See McConnell, 124 S.Ct. at 671 (upholding federal legislation “sharply curbing” the ability of donors to circumvent contribution limits by channeling nonfederal or “soft” money through state political committees to finance “Federal election activity”). The District Court never made specific findings of fact regarding this issue, including how national and local affiliates of the political parties interact and how limitations on transfers of money might affect parties. Since we are reluctant to rule on this issue without the benefit of findings of fact on how such a provision might be expected to operate, we remand for further proceedings.

C. The Related Expenditure Provision is Constitutional as to Contributions

We also affirm the District Court’s holding that the “related expenditure” provisions of Act 64 are constitutional because they serve to reinforce the anti-corruption goals of the contribution limitations. Pursuant to Act 64, “related expenditures” on behalf of a candidate by a third party count toward the third party’s contribution limit as well as the candidate’s expenditure limit. The Act defines related expenditures as those “intentionally facilitated by, solicited by or approved by the candidate or the candidate’s political committee.” Vt. Stat. Ann. tit. 17, § 2809(c). The plaintiffs challenge the provision on three grounds: (1) the phrase “facilitated by” is vague; (2) political parties and PACs should have greater abilities to engage in coordinated expenditures with candidates; and (3) the Act’s rebuttable presumption that an expenditure benefiting six or fewer candidates is a related expenditure is a content-based speech restriction discouraging advertisements about a small number of candidates. We reject each claim. '

Plaintiffs argue that the “facilitated by” standard is vague because it leaves open the possibility that any communication about a candidate’s views with a third party that then undertakes independent expenditures will qualify as a contribution. The First Amendment permits the treatment of “coordinated expenditures” as contributions to a candidate. Buckley, 424 U.S. at 46-47, 96 S.Ct. 612. Independent expenditures may not be limited because “the absence of prearrangement and coordination undermines the value of the expenditure to the candidate, and thereby alleviates the danger that expenditures will be given as a quid pro quo.” Federal Election Comm’n v. Nat’l Conservative Political Action Comm., 470 U.S. 480, 498, 105 S.Ct. 1459, 84 L.Ed.2d 455 (1985). The plaintiffs’ objection to Act 64 is really one which assumes that the word “facilitated” has its broadest meaning, akin to giving any aid in support of the third-party expenditure. If that were what the statute meant, then we would agree that the provision might raise constitutional problems.

We think that, in light of the terms “solicited by or approved by” that accompany it, the term facilitated should be given a narrower reading. Such a reading would also resolve the ambiguity of the statutory language so as to guarantee the constitutionality of the statute. See William Eskridge, Legislation Statutes and the Creation of Public Polioy 873-89 (3d ed. 2001) (discussing canon of constitutional avoidance). Accordingly, we construe the phrase “facilitated by” as requiring some “prearrangement” or “coordination” with the candidate. Nat’l Conservative Political Action Comm., 470 U.S. at 498, 105 S.Ct. 1459. Under such a construction, sharing routine information about a candidate is not sufficient to meet the “facilitated by” requirement. Thus, the provision is not constitutionally invalid.

Nor is there any constitutional barrier to applying this provision to related expenditures by PACs and political parties. The plaintiffs’ argument on this point substantially restates their claim discussed above — that different contribution limits ought to apply to PACs and political parties. We reject it for the same reasons.

Finally, the provision’s rebuttable presumption, which presumes that expenditures by political parties or PACs that benefit six or fewer candidates are contributions to those candidates, does not violate the Constitution by chilling protected speech. The plaintiffs argue that the presumption is unconstitutional because (1) the law may never presume that an expenditure is coordinated and (2) the presumption could never- be rebutted and, as a result, chills independent advocacy of particular candidates. We find neither claim persuasive.

The Constitution does not bar' the use of rebuttable presumptions in this context. The plaintiffs base their argument on Colorado Republican Federal Campaign Comm. v. Federal Election Comm’n, 518 U.S. 604, 116 S.Ct. 2309, 135 L.Ed.2d 795 (1996) (Colorado Republican I). There, the Supreme Court struck down a federal provision that automatically treated all party expenditures, including those made independently, as contributions to candidates. The Court rejected the Court of Appeals’ analysis that the government was entitled to a conclusive presumption that party expenditures are coordinated. Id. at 619, 116 S.Ct. 2309. The fact that the presumption was conclusive, however, played the critical role in that decision: it eliminated the need for a finding that the expenditures were in fact coordinated and foreclosed the possibility of a defense. Id. at 625, 116 S.Ct. 2309. Act 64 does nothing of .the sort, since its presumption is rebuttable.

The plaintiffs’ argument that the presumption is functionally conclusive because one “cannot prove a negative” is, at least in the legal arena, inaccurate. There are ample strategies that an accused party can employ to demonstrate that an expenditure was truly independent from the candidate it supported. The party can, for example, testify that no discussion took place with the candidate about advertising strategies, including the sharing of information about advertising plans. Candidates can testify that they never gave feedback on an independent advertising scheme or that the third parties never solicited such feedback. Adjudicative bodies can take such evidence, or other similar testimony, as proof and infer a lack of coordination. For these reasons, we uphold Act 64’s rebutta-ble presumption concerning related expenditures.

D. The 25 Percent Limit bn Out-of-State Donations is Unconstitutional

We can find no sufficiently important governmental interest to support the provision of Act 64 that limits out-of-state contributions to 25 percent of all candidate contributions. Unlike all of the other Act 64 provisions at issue in this appeal, the out-of-state contribution limit isolates one group of people (non-residents) and denies them the equivalent First Amendment rights enjoyed by others (Vermont residents). The District Court’s decision in this regard should be upheld.

The District Court concluded that Vermont’s interest in eliminating excessive out-of-state contributions was confined to unusually large contributions. 118 F.Supp.2d at 484. The District Court also noted that many non-residents have legitimate and strong interests in Vermont and have a right to participate, at least through speech, in those elections. Id. We find no support in the record for the alternative claim that Vermont has an important interest in singling out one class of contributors for limitations. See 1997 Vt. Laws P.A. 64 (H. 28) (1997) (finding No. 5) (“Increasing campaign expenditures require candidates to seek and rely on a smaller number of larger contributors, often outside the state, rather than a large number of small contributors.”). There are only vague references to the danger of out-of-state contributions, and all refer to the danger of excessively large (not cumulatively great) contributions.

In the two reported decisions on the issue, courts have split on whether limitations of non-resident contributions may be upheld on corruption grounds. The Ninth Circuit has rejected, almost in bright-line form, limitations on non-resident restrictions. In VanNatta v. Keisling, the court struck down an Oregon initiative that effectively limited the use of non-resident contributions to 10 percent of total campaign expenditures. See VanNatta, 151 F.3d 1215, 1217-18 (9th Cir.1998), cert. denied sub nom., Miller v. VanNatta, 525 U.S. 1104, 119 S.Ct. 870, 142 L.Ed.2d 771 (1999); but see Montana Right to Life Assoc. v. Eddleman, 343 F.3d 1085, 1091 n. 2 (9th Cir.2003) (suggesting that VanNatta was superseded by Shrink). Addressing the asserted anti-corruption justification, the court held that the provision suffered from both over and underbreadth. Its overbreadth stemmed from the fact that it prevented all non-resident contributions once the 10 percent threshold had been reached, even those too small to have any corruptive influence. See VanNatta, 151 F.3d at 1221. The provision was underb-road because it did nothing to prevent corruptive (ie., large) resident contributions; nor did it prevent corruptive nonresident contributions until the 10 percent limit had been reached. See id. In other words, the non-resident cap was “not closely drawn to advance the goal of preventing corruption.” Id.

Because Act 64 contains contribution limits, it does not share all of the flaws of the Oregon statute considered in VanNat-ta. Act 64 does, for example, limit large resident and non-resident contributions. Nonetheless, the provision is overbroad in that it prohibits small contributions from out-of-state sources once the 25 percent threshold has been reached, even though such contributions are no more likely to corrupt than in-state contributions. Under this analysis, sustaining the provision would require an additional explanation for why exactly Vermont has an interest in eliminating such small donations only from non-residents.

The Alaska Supreme Court has attempted to craft such an explanation in State v. Alaska Civil Liberties Union, 978 P.2d 597 (Alaska 1999), cert. denied, 528 U.S. 1153, 120 S.Ct. 1156, 145 L.Ed.2d 1069 (2000). The Alaska law at issue capped out-of-state contributions but at lower percentages than Vermont’s law. The court upheld the limitations on the grounds that out-of-state contributions have the ability to distort the Alaskan political system: “These nonresident contributions may be individually modest, but can cumulatively overwhelm Alaskans’ political contributions. Without restraints, Alaska’s elected officials can be subjected to purchased or coerced influence which is grossly disproportionate to the support nonresidents’ views have among the Alaska electorate, Alaska’s contributors, and those most intimately affected by the elections, Alaska residents. These restraints therefore limit the ‘potential for distortion.’ ” Id. at 617. Put another way, Alaska’s “[m]ore than 100 years of experience ... have inculcated deep suspicions of the motives and wisdom of those who, from outside its borders, wish to remold Alaska and its internal policies.” Id. The out-of-state limitation, according to this view, restrains their distorting influence. Id.

The analysis in the Alaska case is a sharp departure from the corruption analysis adopted by the Supreme Court in Buckley and Shrink. Even under the more expansive Shrink analysis, the fear was that candidates would become too compliant with the wishes of large contributors because they must rely on private interest groups for funding. The Alaska analysis permits limitations not to ensure candidate independence generally, but to limit the influence of one set of people— untrustworthy outsiders. Even assuming that the Alaska Supreme Court is correct that outsiders have bad motives and little to contribute to its political discourse, the government does not have a permissible interest in disproportionately curtailing the voices of some, while giving others free rein, because it questions the value of what they have to say.

The Alaska Court’s concern could be understood another way: that when candidates are beholden to fundraisers, and not voters, then large contributions from nonresidents distort the system. Again, this problem would endure even if officials were beholden to in-state contributors. Moreover, Vermont’s expenditure limitations eliminate the major force behind candidates’ excessive reliance on campaign contributors — their need to maximize their ability to raise funds by remaining pliant to the wishes of those who contribute to the political campaign system.

Based on our review of these cases and the governmental interests asserted by the defendants, we are unpersuaded that the First Amendment permits state governments to preserve their systems from the influence, exercised only through speech-related activities, of non-residents. Vermont has asserted no valid interest sufficiently strong to justify the provision, and we therefore hold it unconstitutional. Pursuant to Act 64’s severability provision, the unconstitutional provisions should be severed.

CONCLUSION

In summary, we conclude that Vermont has a sufficiently important governmental interest in support of Act 64’s contribution limits — fighting the real and apparent corruption that accompanies unlimited campaign gifts — and that those contribution limits are closely drawn to achieve this goal. Accordingly, except as noted below, we uphold Act 64’s contribution limits.

Vermont also has established two interests in favor of Act 64’s expenditure limitations that, taken together, are constitutionally compelling: namely, protecting the time of candidates and elected officials, and preventing the reality and appearance of corruption. Although we find that these limits permit candidates for public office to engage in effective campaigns, additional fact-finding is required to complete this narrow tailoring inquiry. On remand, the District Court must determine whether the legislature might have chosen either another type of regulation or higher limits that would still achieve the goals we sanction and yet impinge less on the First Amendment rights at stake.

For the reasons set forth, we affirm the District Court’s holdings that the following provisions of Act 64 are constitutional: (1) the limit on contributions that candidates may accept from individuals or political action committees (§ 2805); (2) the limit on contributions that political action committees and political parties may accept from any source (§ 2805(a)); (3) the definition of political parties as including state, county and town entities (§ 2801(5)); and (4) the classification of related expenditures as contributions (§ 2809(a)). We also affirm the District Court’s finding that the limits on contributions from non-Vermont residents and organizations (§ 2805(c)) are unconstitutional and we uphold its injunction against enforcement of that provision.

We vacate the District Court’s injunction against enforcement of the limitation on contributions by political parties to candidates (§ 2805(a)). We also vacate the judgment and remand for further proceedings with respect to the constitutionality of (1) limiting candidate expenditures (§ 2805a); (2) treating the “related expenditures” of third parties as candidate expenditures (§ 2809(b)); (3) restricting the ability of PACs to make wholly independent expenditures, to the extent the Act’s provisions are read to impose such restrictions (§§ 2801(4), 2805(g)); and (4) limiting transfers of funds from national to state political party affiliates (§§ 2801(5), 2805(a)-(b)). Finally, we affirm the District Court’s injunction against enforcement of Act 64’s expenditure limitations, pending further proceedings.²⁹

In vacating aspects of the District Court’s injunction, we are mindful that Act 64’s limitations are premised on a two-year election cycle. Given that further proceedings must be held, we remand to the District Court the issue of when the various limitations revived by this opinion should be given effect. We thus authorize the District Court to designate an appropriate effective date for these limitations that causes the least disruption to the current election cycle.

Each party shall bear its own costs on this appeal.

1 . We draw some support for our interpretation of Buckley from the Supreme Court's recent decision in McConnell v. Federal Election Comm’n, 540 U.S. 93, 124 S.Ct. 619, 157 L.Ed.2d 491 (2003) (upholding legislation that, inter alia, limited "soft-money” campaign contributions and regulated electioneering communications). We recognize, of course, that McConnell addresses contributions rather than expenditures. Nevertheless, the McConnell Court, by focusing on political and societal developments since Buckley, see id. at -, 124 S.Ct. at 648-54, clearly rejected a static approach to campaign finance reform. Just as the McConnell Court deferred to Congress' "predictive judgments” about the need for federal regulation of soft-money contributions, id. at -, 124 S.Ct. at 673 (observing that Congress had "been taught the hard lesson of circumvention by the entire history of campaign finance regulation”), we respect the Vermont Legislature's similar reliance — in enacting regulations on both campaign contributions and expenditures — on its substantial historical experience with campaign finance reform and its informed predictions about Vermont candidate and donor behavior.

2 . Our consideration of these issues could not help but take particular notice of the McConnell majority's observation, in a discussion replete with descriptions of the distinct evil of ever larger sums of money in American politics, that many of the corporate soft-money contributions at issue in that case were motivated by the "desire for access.” McConnell, 540 U.S. at -, 124 S.Ct. at 649. The Court held that "Congress' legitimate interest” in regulating such contributions "extends beyond preventing simple cash-for-votes corruption to curbing undue influence on an officeholder's judgment, and the appearance of such influence.” Id. at -, 124 S.Ct. at 664 (citation and internal quotation marks omitted). The same issues were identified by the Vermont legislature in this case.

3 . In 1916, Vermont took early steps to ensure the accountability of its elected officials by passing direct primary elections and mandating the post-primary disclosure of candidate expenditures. 1915 Vt. Laws 4, § 22; 1916 Vt. Laws (Sp. Sess.) 4, § 1. In 1961, the legislature adopted mandatory expenditure limits in primary elections, 1961 Vt. Laws 178, and applied those limits to general elections in 1971, 1971 Vt. Laws 259. In 1976, after Buckley, Vermont repealed its expenditure limits but continued to limit the maximum contribution that candidates might accept. 1975 Vt. Laws (Adj. Sess.) 188. Over several decades, Vermont witnessed a period of growing disillusionment with its electoral system, and in 1993 instituted a system of voluntary expenditure limits. Former Vt. Stat. Ann. tit. 17, §§ 2841-42 (1991) (repealed 1997).

4 . A previous lawsuit challenged other provisions of Act 64 which required (1) disclosure of who pays for "political advertisements” and the candidate, party or political committee "on whose behalf” the advertisement is published or broadcast; and (2) reporting of expenditures of "mass media activities ... which included the name or likeness of a candidate for office” occurring within 30 days of a primary or general election. Vt. Stat. Ann. tit. 17, §§ 2881 — 2283. See Vermont Right to Life Committee v. Sorrell, 221 F.3d 376 (2d Cir.2000).

5 . Plaintiff Neil Randall is an incumbent representative in the Vermont legislature. Plaintiff George Kuusela is chairman of the Wind-ham County Republican Party and has run for state legislative office. Plaintiff John Patch is chair of the Chittenden County Democratic Party and has plans to run for State Senate. Plaintiff Steven Howard was previously a candidate for State Auditor and a former state representative. Plaintiff Libertarian Party is a political party in Vermont and ran 44 candidates for office in 1998. Plaintiff Jeffrey Nelson is a longtime resident of Vermont and a financial supporter of the Republican Party. The plaintiffs also include the Vermont Right to Life Committee, the Vermont Republican State Committee, and various additional individuals.

6 . Unease with Buckley is not limited to those who argue that campaign finance regulations have a proper place in our constitutional system. In Shrink, Justices Thomas and Scalia both advocated overruling Buckley not to give legislatures greater leeway to pass needed campaign finance reform, but to heighten the constitutional review given to contribution limits. 528 U.S. at 410-12, 120 S.Ct. 897 (Thomas, J., joined by Scalia, J., dissenting); see also McConnell, 540 U.S. at -, 124 S.Ct. at 737 (Thomas, J., concurring in part and dissenting in part). Justice Kennedy has expressed sympathy for their view. Shrink, 528 U.S. at 409-10, 120 S.Ct. 897 (Kennedy, J., dissenting); cf. McConnell, 540 U.S. at -, 124 S.Ct. at 755-56 (Kennedy, J., concurring in part and dissenting in part) (characterizing Buckley’s "unworkable” distinction between limits on contributions and limits on expenditures as having created an "awkward and imprecise test”).

More recently, in their respective McConnell dissents, Justices Scalia, Kennedy and Thomas observed with dismay that the McConnell majority went even further than the Court had gone in Buckley. See McConnell, 540 U.S. at -, 124 S.Ct. at 729 (Sca-lia, J., concurring in part and dissenting in part) (characterizing the majority as "having abandoned most of the First Amendment weaponry that Buckley left intact”); id. at -, 124 S.Ct. at 742 (Kennedy, J., concurring in part and dissenting in part) ("To reach today's decision, the Court surpasses Buckley’s limits and expands Congress’ regulatory power.”); id. at -, 124 S.Ct. at 730 (Thomas, J., concurring in part and dissenting in part) (accusing the majority of building upon the errors of Buckley by "expanding the anti-circumvention rationale beyond reason”). Hence, it would be unrealistic for us to fail to notice the Court's expanding views.

7 . The arguments to the contrary that are raised by the dissent largely reprise those presented by Senator Buckley and like-minded commentators in the early 1970s. While some of those arguments were then successful, they failed to result in having the Supreme Court hold that expenditure limits are per se unconstitutional—a fact that our dissenting colleague disputes.

8 . Some scholars think that such regulations are more properly classified as "content-neutral.” See, e.g., Elena Kagan, Private Speech, Public Puipose: The Role of Governmental Motive in First Amendment Doctrine, 63 U. Chi. L. Rev. 413, 517 n.151 (1996) ("I treat the expenditure ceilings in Buckley as content neutral, as is the norm, because they cover spending for expression supporting all political candidates. It could be argued, however, that the ceilings imposed a subject-matter limitation because they applied only to spending in political campaigns.”) (citing Laurence H. Tribe, American Constitutional Law § 13-27, at 1132-36 (2d ed.1988)).

9 . See, e.g., City of Richmond v. J.A. Croson Co., 488 U.S. 469, 500, 109 S.Ct 706, 102 L.Ed.2d 854 (1989) (plurality opinion) ("The factfinding process of legislative bodies is generally entitled to a presumption of regularity and deferential review by the judiciary.”) (citing Williamson v. Lee Optical of Oklahoma, 348 U.S. 483, 488-89, 75 S.Ct. 461, 99 L.Ed. 563 (1955)).

10 . Amicus Brennan Center for Justice, arguing in favor of the spending limits, relies on the interest of "preventing] the demonstrably corrosive effects of uncontrolled campaign spending on the democratic process.” Ami-cus further characterizes "uncontrolled campaign spending” as a problem that "harms the quality of democratic representation,” that is "undermining faith in the democratic process,” and that "has a harmful impact on voter turnout.” Amici States rely on the first three interests identified by Vermont, and also relied upon by intervenors, described as (1) "eliminating] corruption and the appearance of corruption in the political process”; (2) "ensuring] that officeholders can spend less time fund-raising and more time performing their duties” and (3) "bolstering] equal access to political office and restoring] the public’s confidence in the electoral system.”

11 . Reiterating points it had already made in Buckley and Shrink, the Supreme Court in McConnell clearly held that large campaign donations — even the so-called "soft-money” contributions at issue in McConnell — "can corrupt or, at the very least, create the appearance of corruption of federal candidates and officeholders.” McConnell, 540 U.S. at -, 124 S.Ct. at 661. The McConnell majority observed that during the 1996 and 2000 campaigns, "more than half of the top 50 soft-money donors gave substantial sums to both major national parties, leaving room for no other conclusion but that these donors were seeking influence, or avoiding retaliation, rather than promoting any particular ideology.” Id. at-, 124 S.Ct. at 663 (emphasis in original); see also id. at-, 124 S.Ct. at 649 (citing evidence that "many corporate contributions were motivated by a desire for access to candidates and a fear of being placed at a disadvantage in the legislative process relative to other contributors, rather than by ideological support for the candidates and parties”).

12 . This phenomenon was also chronicled by the McConnell Court, which quoted one former Senator as admitting that members of Congress "have limited amounts of time” but do make themselves available to meet with those who give "large sums.” McConnell, 540 U.S. at-, 124 S.Ct. at 664 (citations and internal quotation marks omitted). The same former Senator went on to explain that those meetings are "not idle chit-chats about the philosophy of democracy .... Senators are pressed by their benefactors to introduce legislation, to amend legislation, to block legislation, and to vote on legislation in a certain way.” Id. at-, 124 S.Ct. at 664-65 (citations and internal quotation marks omitted).

Our dissenting colleague criticizes Vermont’s reliance on this type of anecdotal evidence but the Supreme Court expressly credited similar testimony in McConnell where, as here, the record was "replete” with anecdotal examples of the type of access-peddling that concerned Congress. McConnell, 540 U.S. at -, 124 S.Ct. at 664. The McConnell Court specifically observed the type of "particularized evidence of improper influence” required by our dissenting colleague, post at 190, would be particularly hard to come by. "Even if it occurs only occasionally, the potential for such undue influence is manifest. And unlike straight cash-for-votes transactions, such corruption is neither easily detected nor practical to criminalize.” McConnell, 540 U.S. at-, 124 S.Ct. at 666.

13 . Although Judge Winter criticizes Act 64's proponents' reliance on the phenomenon of "bundling” or "pooling” contributions as a justification for the legislation, see post at 185-87, we note that the McConnell Court lauded Congress’ similar efforts to anticipate, document, and respond to candidates’, donors’ and parties' evolving strategies for circumventing existing campaign finance reforms. See id. at -, 124 S.Ct. at 661 ("[T]he First Amendment does not require Congress to ignore the fact that candidates, donors, and parties test the limits of the current law.") (citation and internal quotation marks omitted). Indeed, for example, no sooner had the Buckley Court construed federal law to cover "express advocacy,” did the political spending system create "issue advertising.” See Id. at-, 124 S.Ct. at 650-52. Similarly, elaborate schemes involving the "solicitation, transfer, and use of soft money” grew out of FECA’s imposition of limitations on the source and amount of contributions of "federal” or "hard” money. Id. at-, 124 S.Ct. at 648-50. The McConnell Court recognized that such new developments — and Congress' general awareness of the realities of campaign fundraising — quite correctly had moved Congress to legislate. See id. at-, 124 S.Ct. at 665-66. Clearly, bundling practices allow donors to achieve indirectly what contribution limits prevent them from doing directly. We do not diminish the legislature's reliance on this concern because, like the Supreme Court, we refuse to take a "crabbed view of corruption, and particularly of the appearance of corruption.” Id. at-, 124 S.Ct. at 665.

14 . Amicus Brennan Center also points to evidence nationally that "raising necessary campaign funds has turned into a full-time job for many candidates and officeholders.” And there is consensus among the five amici states — including Connecticut and New York — that "[ajlthough the States have a compelling interest in ensuring that the demands of fundraising not drain elected officials of the time and attention necessary to carry out their official duties, they are not succeeding." We find it significant that all three states that make up our Circuit have expressed concern over the all-consuming nature of unrestrained fundraising.

15 . Moreover, Buckley could not have anticipated the whole range of individual concerns faced by specific states, such as Vermont, given those states’ unique experiences with state-level campaign finance reform in the three decades since Buckley was decided.

16 . The General Assembly minced no words when describing this phenomenon: "Some candidates and elected officials, particularly when time is limited, respond and give access to contributors who make large contributions in preference to those who make small or no contributions.” 1997 Vt. Laws P.A. 64 (H. 28) (finding No. 2).

17 . Our dissenting colleague challenges our deference to the Vermont legislature with respect to its predictions about the effect that Act 64’s contribution limits, standing alone, would have on candidate behavior. See post at 189-94. Such "predictive judgments” by a legislature, however, should be accorded “substantial deference ..., particularly when, as here, those predictions are so firmly rooted in relevant history and common sense.” McConnell, 540 U.S. at-, 124 S.Ct. at 673 (citation and internal quotation marks omitted); see also supra at 120 n. 14 (noting the consensus among the amici states that this time-protection interest is not unique to Vermont).

18 . It further appears that the two interests analyzed here — anti-corruption and time protection — overlap and might be better described as one interest — "to protect the integrity of the electoral process- — the means through which a free society democratically translates political speech into concrete governmental action.” Shrink, 528 U.S. at 401, 120 S.Ct. 897 (Breyer, J., concurring), or, as amicus Brennan Center puts it, to "protect the quality of democratic representation.” Since defendants did not defend Vermont’s law on the basis of such an interest, however, and the Supreme Court has not relied on such an interest in its recent analyses of the constitutionality of campaign finance laws, we do not explicitly assess that interest here. We do note that this interest appears to underlie, in part, both of the interests asserted by the State — anti-corruption and time protection for candidates and elected officials — that we do analyze.

19 . One commentator has described the Supreme Court’s narrowly tailoring requirement this way: “The narrow-tailoring prong, then, involves essentially factual questions about whether the law is indeed narrowly drawn: Does the law further the interest; is the law limited to speech that implicates the interest; does the law cover all such speech; are there less restrictive alternatives that will serve the interest equally well?” Eugene Volokh, Freedom of Speech, Permissible Tailoring and Transcending Strict Scrutiny, 144 U. Pa. L. Rev. 2417, 2424 (1996) (footnotes omitted).

20 . We note that examination of other forms of expenditure limits has often involved analysis of effective advocacy and campaigning. For example, the ability of corporations to communicate an effective political message, despite a state ban on the use of general treasury funds for candidate-oriented independent expenditures, was noted in Austin v. Michigan Chamber of Commerce, 494 U.S. 652, 660, 110 S.Ct. 1391, 108 L.Ed.2d 652 (1990) ("We find that the Act is precisely targeted to eliminate the distortion caused by corporate spending while also allowing corporations to express their political views.") (emphasis added).

21 . Voluntary campaign expenditure limitations of one sort or another have been promulgated by governments in Hawaii, Kentucky, Maine, Maryland, Michigan, Minnesota, New Hampshire, Rhode Island, and Wisconsin. See, e.g., Haw.Rev.Stat. § 11-219 (2003); Ky.Rev.Stat. Ann. § 121A.030 (2003); Me.Rev.Stat. Ann. tit. 21-A § 1125 (2003); Md.Code Ann. Elec. Law § 15-105 (2003); Mich. Comp. Laws Ann. § 169.267 (2003); Minn.Stat. § 10A.25 (2003); N.H.Rev.Stat. Ann. § 664:5-b (2003); R.I. Gen. Laws § 17-25-20 (2003); Wis. Stat. Ann. § 11.31 (West 2003) (enforcement enjoined by Wisconsin Realtors Ass'n v. Ponto, 233 F.Supp.2d 1078 (W.D.Wis.2002)).

22 . We draw no conclusions at this stage about whether such a reason would be constitutionally acceptable.

23 . In addition, our dissenting colleague asserts that Act 64’s definition of "expenditures” and "related expenditures” may mean that the spending totals for past campaigns understate the amount of expenditures as defined by Act 64. He also points out that the legal and record-keeping costs of compliance with the new law will also inflate future candidates' expenditures. See post at 172-73. Findings in this regard will also help inform the inquiry on the degree to which this provision impinges on First Amendment rights.

24 . The dissent asserts that this least restrictive means inquiry is a "mixed issue of legislative fact and law” that does not require remand. See post at 203. The dissent’s extensive discussion of the significance of the distinction between adjudicative and legislative facts seems to us tangential and unhelpful. The fact that this Court may ultimately undertake de novo review of any legislative facts found by the District Court on remand or that appellate courts take judicial notice of legislative facts under appropriate circumstances, see id., does not mean that we must resolve disputed legislative facts — particularly facts that are dispositive of the case before us — on an insufficiently developed record. Nor should we, under these circumstances, take judicial notice of facts relating to Vermont's political and electoral system, an arena in which this Court lacks substantial experience or expertise. By contrast, the types of "legislative facts” that have been addressed most recently in our caselaw deal with much more straightforward questions, e.g., geography and jurisdiction or the fact that cocaine is derived from coca leaves. See, e.g., United States v. Hernandez-Fundora, 58 F.3d 802, 812 (2d Cir.) (citations omitted), cert. denied, 515 U.S. 1127, 115 S.Ct. 2288, 132 L.Ed.2d 290 (1995).

Judge Winter’s fundamental disagreement with our opinion seems not to be one of appropriate characterization of facts. Rather, his dispute with our position is that he believes that a remand is unnecessary because there is nothing more to be learned. Obviously, we disagree. There are gaps in the record — likely the result of the District Court's abridged consideration of the issue of narrow tailoring — that ought be filled before any court undertakes to resolve the ultimate issue in this case: whether there are less restrictive alternatives that could have advanced Vermont's compelling interests.

25 . Our dissenting colleague criticizes these guidelines for further inquiry on several grounds, none of which we find persuasive. At the outset, he asserts that the first two queries are simply straightforward questions of legislative history. See post at 204. Unfortunately, as noted supra at 132-34, the answers to these questions are not discernible from the existing record. To the extent that the dissent is suggesting that these questions are irrelevant, see post at 206-07, we disagree.

We reject the suggestion that we are overly focused on what the legislature did. Of course, the ultimate issue for the District Court on remand is whether there exists a less restrictive type or degree of regulation that would serve Vermont's compelling anti-corruption and time-protection interests; it is not merely whether the legislature considered such an alternative. See supra at 132-35. Because our dissenting colleague concludes, without citation, that "a combination of public and private financing with low contribution limits is infinitely less restrictive ... and accomplishes all of the ostensible purposes of Act 64’s expenditure limits,” he deems a remand unnecessary. See post at 207. We, however, do not believe that proposition to be self-evident, nor is it supported by the current record.

We are loath to presume, on an incomplete record, the reasons for the legislature's decision not to pursue such alternatives and we believe the District Court's — and perhaps eventually this Court's — evaluation of whether less restrictive alternatives exist will undoubtedly be aided by the legislature’s own analysis of those alternatives, to the extent such analysis occurred. Indeed, even a modicum of deference to the legislature and consideration for principles of federalism would seem to require consideration of its reasons for rejecting a potentially-less-restrictive alternative scheme. But see post at 207 (concluding that the only possible reason for rejecting such a measure is the drafters’ ulterior, incumbent-protective motive).

So as not to restrict its proceedings upon remand in any way — and to avoid the strictures of the mandate rule, see, e.g., United States v. Ben Zvi, 242 F.3d 89, 95 (2d Cir.2001) (explaining that the mandate rule "compels compliance on remand with the dictates of the superior court and forecloses relitigation of issues expressly or impliedly decided by the appellate court”) (citation, emphasis and internal quotation marks omitted) — we have elected to speak broadly in framing the inquiry to be undertaken on remand. To do as our dissenting colleague suggests, on the limited record before us, would unnecessarily and unwisely encroach on the essential role of the district judge in these proceedings.

26 . Before leaving the issue of the constitutionality of the spending limits, we briefly address a concern raised by our dissenting colleague: that there is language in Act 64 relating to the expenditure limits that renders the provision impermissibly vague. See post at 161-68. As a threshold matter, it must be pointed out that plaintiffs did not raise this claim either in the District Court or on appeal — and it is not because they were unfamiliar with such a ■ claim. Indeed, plaintiffs made this kind of argument with regard to other provisions in Act 64 also at issue in this case, see infra at 87; and one plaintiff, represented by the same counsel as here, challenged other provisions of Act 64 in a previous case on these grounds. See Vermont Right to Life Comm., Inc. v. Sorrell, 221 F.3d 376, 387 (2d Cir.2000) (“VRLC identifies an additional constitutional problem posed by [the disclosure provisions]: They are imper-missibly vague.”). We will not strike down a state statute, in a pre-enforcement facial challenge, on grounds upon which it has not been challenged.

The dissent takes issue specifically with several aspects of the statutory language that pertain to the expenditure limits — that the spending is made “for the purpose of influencing an election,” that it applies to "anything of value,” and the spending is attributable to a "candidate.” See post at 161-65. Of these, the first two provisions have been a part of federal and state campaign finance laws for decades, and they have been upheld by the Supreme Court. See Buckley, 424 U.S. at 145-47, 96 S.Ct. 612; 2 U.S.C. § 431(8)(A)(i). As to the last provision regarding “spending attributable to a candidate,” the notion that candidates do not know when they are candidates is belied by the specificity of the provision itself. As the dissent appears to acknowledge, Act 64 defines "candidate” as someone who has done one of three specific things. See post at 165. Though the dissent raises the possibility that a candidate could be deemed a candidate by some other "affirmative action” not defined by statute, see id., no plaintiff has raised this concern in light of the statute’s clarity in this regard.

Finally, we note that, of course, as with all campaign finance regulations, minor ambiguities, omissions or statutoiy quirks will be resolved in the normal course of administrative interpretation, legislative amendment and litigation. In any event, the ambiguities raised by the dissent do not, we believe, render the statute unconstitutional. Indeed, though raising the specter of a multitude of pitfalls latent in the statutory text, even our dissenting colleague would uphold the constitutionality of the majority of the Act — despite that many of those same ambiguities speak to portions of the Act unanimously upheld in Part II of this opinion.

27 . Any amendment by the Vermont legislature to mandatory spending caps in the interim has the potential, of course, to moot litigation over its constitutionality.

28 . The District Court noted that by way of comparison, the Vermont law has a limit-constituency ratio (i.e., the maximum contribution allowed divided by the number of constituents) of .00068, compared to .00040 under the Missouri statute. 118 F.Supp.2d at 479.

29 . Because our original opinion was withdrawn, the petition for rehearing as to that opinion became moot. Upon the filing of this substituted opinion, however, the parties may, if they so desire, file petitions for rehearing as though this were our original decision.