dissenting.
I concur in part in the result reached by my colleagues’ opinion. I respectfully dissent from their holdings as to Act 64’s limits on expenditures by candidates, including “related expenditures” by individual supporters and political parties, and as to the Act’s forced centralization of local political parties. In view of the length of this separate opinion, I begin with a table of contents.
CONTENTS
I. INTRODUCTION.150
II. APPLICABLE CONSTITUTIONAL PRINCIPLES .152
a) Overview.152
b) Money and Protected Political Speech.153
c) Freedom to Organize Political Parties.154
d) Sufficient Governmental Interests.155
e) Requisite Precision of Regulation.156
f) Appropriate Level of Scrutiny.!.157
III. THE PROVISIONS OF ACT 64 . Gj LO
a) Overview. Os lO tH
b) Two-Year Cycle. '■ — > ÍO t — 1
c) Limits on Expenditures by Candidates t — l
d) Limits on Contributions to Candidates LO CD t — 1
*150e) Limits on “Related Expenditures”.
f) Costs of Compliance.
g) Treatment of the Press.
h) Administration and Enforcement.
IV. THE BURDEN ON PROTECTED SPEECH . ti — I
a) The Burden on Grassroots Political Activity. ti — I
1) Burden on Volunteer Activity. t> i — l
2) Burden on Local Party-Funded Activity . . t-rH
b) The Burden on Candidates’ Speech. rH
1) Past Experience. t-rH
of From Past Elections. tH
B) Inadequacy of Average Spending in Past Elections as a Constitutional Standard. 05 Cr-i — I
C) Evidence of Contested Elections in Vermont. UD ti — I
2) Effect of Act 64’s Expenditure Limits on Candidates. CO ti-H
e) The Burden on Challengers. CO t-rH
d) The Burden on the Press. H CO i — i
e) The Burden on Party Affiliates. N 00 rH
V. THE CONSTITUTIONAL RESOLUTION. CO 00
a) Restrictions on Political Activity for Which No Governmental Interests are Asserted. 00 ^
b) Buckley Forecloses the Asserted Justifications for Expenditure Limits 00 CR
c) The Insufficiency of the Governmental Interests. 00 CO
1) Anti-Corruption. 00 CO
2) Time Protection. CO CO
3) Public Confidence in Government. CO ^
d) Stopping the “Arms Race,” “Effective Advocacy,” and Incumbent Protection. t — i CO 05
1) The “Arms Race”. I — l CO Q
2) “Effective Advocacy”. ( — i CO -3
3) Incumbent Protection. i — l CO CO
e) The Excessive Discretion Accorded Administrators. 1 — a CO CO
VI. THE REMAND ON NARROW TAILORING . CO o tO
a) Legislative Facts, Adjudicative Facts, and Mixed Issues of Fact and Law tO o tO
1) The Distinction Between Legislative and Adjudicative Facts CO o CO
2) Remand for “Findings” of Legislative Fact and of Law .... CO o
b) Failure to Define What is Restrictive About Act 64 . CO o Ol
e) The Existence of a Less Restrictive Alternative. CO o CTi
d) Remanding to the Wrong Forum. CO o -3
VII. CONCLUSION. to o co
Appendix >0 -i o
I. INTRODUCTION
On August 7, 2002, my colleagues filed an extensive majority opinion in this matter. I filed a partial dissent that was almost as long as the majority opinion. Although our opinions, subsequently vacated by my colleagues, were a clear and present danger to the forests of the nation, they failed to join issue in many ways.
My colleagues’ opinion in the main adopted the stated purposes of the Vermont legislature and the opinions of its proponents as sufficient constitutional justifications for Act 64. My dissent was largely a detailed statutory analysis of Act 64, concluding that it limits or prohibits a vast range of ordinary political activities. The dissent also discussed the Act’s perva*151sive ambiguities to be resolved through the often ad hoc discretionary decisions by those who must administer it and was critical of my colleagues for viewing the Act through the prism of what its proponents said about it instead of what the Act itself says.
Although my dissent has been substantially revised to address issues raised by my colleagues’ decision to remand rather than reverse with regard to the expenditure limits issue, to elaborate in more detail the evidence that Act 64 is intended to protect incumbent legislators, and to provide a more particularized discussion of the evidence in the record, not much has changed since my colleagues vacated our earlier opinions. Their revised opinion does not discuss or even mention, to take only a few examples, the following concrete effects of Act 64:
(i) Act 64’s limits on expenditures are so low that they are below the amount spent in the past by third-party candidates, see infra Part IV(b)(l)(C);
(ii) Act 64 limits a candidate who must run in both a primary and general election to the same expenditures as an opponent who runs only in the general election, see infra Parts 111(b), IV(c);
(iii) a favorable press editorial after an interview with a candidate must be valued and treated as a contribution to, and expenditure by, the candidate’s campaign subject to Act 64’s limits, see infra Parts 111(g), PV(d);
(iv) according to Vermont’s Secretary of State, if a candidate provides a photo or written materials to a person who uses the photo or materials in a publication, the candidate must treat the cost of the publication as an expenditure by the candidate, see infra Parts 111(g), IV(d); Appendix A;
(v) the draconian limits on the activities of autonomous local units of political parties resulting from Act 64’s requirement that funding for all local activities, such as a town committee picnic, must be from a single statewide party bank account with the permission of the person who controls that account, see infra note 1, Part IV(a)(2), (e);
(vi) the need for campaign volunteers to keep records of every mile they drive to meetings, to campaign events, or on other campaign business over a two-year period, see infra Parts 111(e), IV(a)(l);
(vii) the need of a campaign to treat the miles driven by volunteers as a campaign expenditure subject to Act 64’s limits (and to keep records and report that mileage), and to prohibit further driving by volunteers whenever limited (by Act 64) campaign resources are needed for other activities or the expenditure limits have been reached, see infra Parts 111(e), rV(a)(l); or
(viii) the restrictions on ordinary homeowners wanting to hold meet the candidates events (and again the need for a candidate’s campaign to record and limit such events as campaign expenditures), see infra Part IV(a)(l).
My colleagues now remand Act 64’s expenditure limits principally for an inquiry into whether the Vermont legislature considered somewhat higher limits as a less restrictive alternative. Even putting aside the Supreme Court’s holding that expenditure limits are per se unconstitutional, see Buckley v. Valeo, 424 U.S. 1, 45, 54, 57, 96 S.Ct. 612, 46 L.Ed.2d 659 (1976) {per cu-riam), that language used in Act 64 is unconstitutionally vague, 424 U.S. at 44, 80, 96 S.Ct. 612, see infra note 6, and the fact that, even under the standard applied by my colleagues, the levels of Act 64’s limits are unconstitutionally low and clearly protective of incumbents, the remand is something of an oddity. First, it involves largely legal issues described by my col*152leagues as factual. Second, the issue should be whether less restrictive alternatives exist, not whether the Vermont legislature considered them. Third, the degree of an alternative’s restrictiveness cannot be evaluated without knowing what is restrictive, and to what degree, about the law in question. However, the only discussion of Act 64’s specific restrictions on political activity is found in this dissent. Fourth, a speech-supportive and constitutionally sound alternative — a combination of private and public financing with low contribution limits — is obviously available. In fact, Act 64 itself limits contributions to such small amounts — $400 for statewide candidates, $300 for Senate candidates, and $200 for House candidates — that there is no evidence showing that the possibility of improper influence on officeholders is at present anything but negligible.
I therefore respectfully continue to dissent as to the constitutionality of two aspects of Act 64. See 1997 Vermont Campaign Finance Reform Act (codified as Vt. Stat. Ann. tit. 17, §§ 2801-2888) (“Act 64”). Those aspects are Act 64’s limits on expenditures by candidates, including related expenditures by individual supporters and political parties, and its restrictions on fundraising and spending on party-building activities by state, county, and local committees of a political party. Id. §§ 2801(5), 2805a. Otherwise, based on Supreme Court precedent, I concur in the result reached by my colleagues.
II. APPLICABLE CONSTITUTIONAL PRINCIPLES
a) Overview
Neither Act 64’s limits on expenditures nor its restrictions on independent fund-raising and expenditures by state or local party committees involve new issues of constitutional law.
Buckley v. Valeo, 424 U.S. 1, 96 S.Ct. 612, 46 L.Ed.2d 659 (1976) (per curiam), held, without qualification, that government may not limit campaign expenditures by candidates for electoral office. Id. at 45, 54, 57, 96 S.Ct. 612. Act 64 limits such expenditures notwithstanding Buckley. Indeed, the proponents of Act 64 never doubted its unconstitutionality under Buckley and enacted it for the explicit purpose of creating a vehicle for litigation to overturn Buckley. See infra note 2 and accompanying text. Act 64’s limits on expenditures violate the First Amendment because they limit a broad spectrum of political speech and activity, including ordinary grassroots activities and editorializing and reporting by the press, for no permissible purpose. Further, they entrust those who enforce the law with unfettered, and unconstitutional, discretion to determine, often on an ad hoc basis, what acts of political advocacy are permitted and what are prohibited. Even if expenditure limits were not per se unconstitutional, the low level at which the limits are set by Act 64 so heavily favors incumbents that it can be upheld only by application of a legal test similarly skewed toward incumbents. See infra Part V(d).
Moreover, Act 64 treats a contribution to a local political party affiliate as a contribution to all affiliates and requires that all such contributions be initially deposited in the state party bank account. See infra note 1. This means that all funding for a local affiliate’s activities — even for a six-pack of diet soda for a town committee picnic — must be approved and paid by the person who controls that statewide account. At a stroke, and without any proffered reason, much less a statement of a compelling governmental interest, this revolutionary provision destroys the autonomy of local affiliates of political parties from each other and from the state party organization, and thereby violates both *153freedom of speech and freedom of association.
Act 64 suppresses ordinary political activity at every level of the electoral process. It reflects the philosophy of one witness for the defense who testified that government ought to regulate political speech the way it regulates public utilities. Trial Tr. vol. V at 167 (Helen David-Friedman). Act 64 may be a popular law— although this dissent will note several instances of great disquiet and even shock among proponents upon learning what it actually says — but only because its proponents systematically divert attention from the law’s actual provisions to the nobility of their goal — here the transfer of political power from “special interests” to “ordinary citizens.” Maj. Op. at 116, Appellant’s Br. at 24-29; Trial Tr. vol. IX at 57-62 (Gordon Bristol); id. at 137 (Elizabeth Ready); Trial Tr. vol. VII at 88 (Cheryl Rivers); Trial Tr. vol. VIII at 63-64 (Peter Smith). However, even this attractive rhetoric cloaks sinister purposes. Foiling “special interests” while empowering “ordinary citizens” is a rhetorical staple of electoral politicians of every viewpoint because the terms are used as synonyms for one’s opponents and supporters respectively. In this light, the pursuit of this goal through the regulation of political speech is the road to the suppression of opponents.
As Justice Brandéis once noted, “The greatest dangers to liberty lurk in insidious encroachment by men of zeal, well-meaning but without understanding.” Olmstead v. United States, 277 U.S. 438, 479, 48 S.Ct. 564, 72 L.Ed. 944 (1928) (Brandéis, J., dissenting). And Justice Black has reminded us that “[h]istory indicates that urges to do good have led to the burning of books and even to the burning of ‘witches.’ ” Beauharnais v. Illinois, 343 U.S. 250, 274, 72 S.Ct. 725, 96 L.Ed. 919 (1952) (Black, J., dissenting). Act 64, which has its greatest impact in silencing those ordinary citizens whose active participation in politics takes place through organized, groups, provides us with a modern reminder of the wisdom of those two statements.
b) Money and Protected Political Speech
The activities limited by Act 64 are the ordinary stuff of democracy that constitutes the core of the conduct protected by the First Amendment. There is “practically universal agreement that a major purpose of [the First] Amendment was to protect” political speech. Mills v. Alabama, 384 U,S. 214, 218, 86 S.Ct. 1434, 16 L.Ed.2d 484 (1966), quoted in McIntyre v. Ohio Elections Comm’n, 514 U.S. 334, 346, 115 S.Ct. 1511, 131 L.Ed.2d 426 (1995). Indeed, the First Amendment “has its fullest and most urgent application precisely to the conduct of campaigns for political office.” Buckley, 424 U.S. at 15, 96 S.Ct. 612 (quoting Monitor Patriot Co. v. Roy, 401 U.S. 265, 272, 91 S.Ct. 621, 28 L.Ed.2d 35 (1971)).
As Buckley held, because money is needed for access to resources of communication, any limit on the use of money for political speech is a limit on that speech. 424 U.S. at 19. Political speech without an audience is not worth the effort. Political speakers must therefore go to where voters are or speak through a medium that voters watch or hear. Without resources of communication, no speech is effective. Without money, resources are not obtainable. Cars use gas. Gas costs money. A candidate who has reached Act 64’s limits on expenditures and may not even drive the family car to a town green to make a speech is as effectively barred from speaking as he or she would be if the law flatly *154prohibited the speech itself. As the Supreme Court has stated:
A restriction on the amount of money a person or group can spend on political communication during a campaign necessarily reduces the quantity of expression by restricting the number of issues discussed, the depth of their exploration, and the size of the audience reached. This is because virtually every means of communicating ideas in today’s mass society requires the expenditure of money. The distribution of the humblest handbill or leaflet entails printing, paper, and circulation costs. Speeches and rallies generally necessitate hiring a hall and publicizing the event.
Buckley, 424 U.S. at 19, 96 S.Ct. 612.
Proponents of Act 64 rarely acknowledge this fact in stressing their preference for limiting political speech to the “old-fashioned handshake campaign,” Trial Tr. vol. IX at 47 (Gordon Bristol), including “meet and greet” events, Trial Tr. vol. X at 187 (Karen Kitzmiller), such as “spaghetti suppers,” Trial Tr. vol. IX at 221 (Anthony Pollina), “little parties” for “150 people” to which “a couple hundred” people are invited by mail, id. at 141-42 (Elizabeth Ready), Rotary Club and Jaycees meetings, Trial Tr. vol. V at 43 (Donald Hooper), booths at county fairs, Trial Tr. vol. IX at 129 (Elizabeth Ready), barbecues, op-ed articles published in the press, id. at 135 (Elizabeth Ready), women’s groups meetings, “various boards” meetings, Trial Tr. vol. VII at 17 (Toby Young), and so forth. However, all such activities consume resources for which someone makes, or has made, expenditures of money, e.g., use of a vehicle, gas, food, soft drinks, meeting rooms, postage, salaries for editors and deliverymen, a printing facility, and so forth. Act 64’s proponents do not recognize these hard facts, but the Act does, and its limits on campaign expenditures directly affect — either by limiting or requiring a largely discretionary exemption for — each of the items described above.
c) Freedom to Organize Political Parties
The First Amendment requires that citizens be allowed freely to form political organizations at various levels of government. This protection extends to allowing organizations to be related to each other as affiliates of the same political party while still retaining much local autonomy. See Timmons v. Twin Cities Area New Party, 520 U.S. 351, 358, 117 S.Ct. 1364, 137 L.Ed.2d 589 (1997) (stating that “political parties’ government, structure, and activities enjoy constitutional protection” and noting that a political party has “ ‘discretion in how to organize itself, conduct its affairs, and select its leaders,’ ”) (citing Eu v. San Francisco County Democratic Cent. Comm., 489 U.S. 214, 230, 109 S.Ct. 1013, 103 L.Ed.2d 271 (1989)); see also Buckley, 424 U.S. at 15, 96 S.Ct. 612 (recognizing the right “‘to associate with the political party of one’s choice’ ” and noting that “ ‘[effective advocacy of both public and private points of view, particularly controversial ones, is undeniably enhanced by group association’ ”) (internal citations omitted). Act 64 treats state, county, and local affiliates of a political party as a single aggregated unit for purposes of fundraising and contribution limits, see Vt. Stat. Ann. tit. 17, § 2801(5), and requires that all contributions to a political party so defined be initially deposited in a single, statewide checking account.1 As a result, *155local affiliates can raise and spend money only through access to, and with the permission of, whoever controls.that bank account.
d) Sufficient Governmental Interests
The Supreme Court has held that only the prevention of “corruption or the appearance of corruption” constitutes a sufficiently compelling interest to limit contributions to candidates. See Buckley, 424 U.S. at 25-28, 96 S.Ct. 612 (holding that limiting the actuality and appearance of corruption is a “constitutionally sufficient justification” for a contribution limitation, but dismissing other proffered justifications for the limitation). It has also held, however, that neither the anti-corruption rationale, the interest in equalizing the financial resources of candidates, nor the increase in money spent on political campaigns justifies the limiting of amounts that candidates for office may spend to promote their candidacy. Id. at 45, 54, 57, 96 S.Ct. 612. Indeed, the Court has stated that
[t]he First Amendment denies government the power to determine that spending to promote one’s political views is wasteful, excessive, or unwise. In the free society ordained by our Constitution it is not the government but the people — individually as citizens and candidates and collectively as associations and political committees — who must retain control over the quantity and range of debate on public issues in a political campaign.
Buckley, 424 U.S. at 57, 96 S.Ct. 612.
Since Buckley, the Court has adhered to the distinction between the regulation of contributions and the regulation of expenditures. See Federal Election Comm’n v. Colorado Republican Federal Campaign Comm., 533 U.S. 431, 440-41, 121 S.Ct. 2351, 150 L.Ed.2d 461 (2001) (“Colorado II”). In Colorado II, the Court made the statement, reaffirmed even more recently in McConnell v. Federal Election Comm’n, 540 U.S. 93, 124 S.Ct. 619, 655, 157 L.Ed.2d 491 (2003), that “ever since we first reviewed the 1971 Act, we have un*156derstood that limits on political expenditures deserve closer scrutiny than restrictions on political contributions,” because “[rjestraints on expenditures generally curb more expressive and associational activity,” and “limits on contributions are more clearly justified by a link to political corruption.” Id. The Court went on to state that “[gjiven these differences, we have routinely struck dorm limitations on independent expenditures by candidates, other individuals, and groups, while repeatedly upholding contribution limits.” Id. at 441-42, 121 S.Ct. 2351 (citations and footnotes omitted) (emphasis in original).
One would think that the unqualified statements of the Supreme Court regarding the unconstitutionality of expenditure limits might be the end of the matter at this level of the court system, particularly since the sponsors of Act 64 have made no secret of their intention to enact it in order to provoke a test case to overrule Buckley with regard to expenditure limits. See Memorandum from Secretary of State Deborah L. Markowitz re: Review of Practical Policy and Legal Issues of Vermont’s Campaign Finance Law (Jan. 9, 2001), available at http://vermont-elec-tions.org/electionsl/2001GAMemoCF.html (“2001 Memorandum”)-, see also Hearing on H. 28 Before the Vt. House Comm. on Local Gov’t, 64th Biennial Sess. (1997) (statement of Anthony Pollina); Hearing on H. 28 Before the Vt. Senate Comm, on Gov’t Operations, 64th Biennial Sess. (1997) (statement of Sen. William Doyle); Vt. House Comm, of Conf., Report on Campaign Finance, H. 28, 64th Biennial Sess. (1997).2 However, the views of my colleagues require that I describe in some detail why Act 64 is unconstitutional in the particular respects noted above, even under the constitutional test that they create.
e) Requisite Precision of Regulation
There is another body of First Amendment jurisprudence that is of relevance here: Any regulation of protected speech must embody valid criteria sufficiently precise to ensure that officials apply those criteria. See Thomas v. Chicago Park Dist., 534 U.S. 316, 323, 122 S.Ct. 775, 151 L.Ed.2d 783 (2002) (stating that the Supreme Court has “required that a time, place, and manner regulation contain adequate standards to guide the official’s decision and render it subject to effective judicial review”); Forsyth County v. The Nationalist Movement, 505 U.S. 123, 131, 112 S.Ct. 2395, 120 L.Ed.2d 101 (1992) (“ ‘[A] law subjecting the exercise of First Amendment freedoms to the prior restraint of a license’ must contain ‘narrow, objective, and definite standards to guide the licensing authority.’ ”) (quoting Shuttlesworth v. City of Birmingham, 394 U.S. 147, 150-51, 89 S.Ct. 935, 22 L.Ed.2d 162 (1969)). Otherwise, the officials who administer the law will have the discretion to fashion and apply their own criteria without restraint. See Thomas, 534 U.S. at 323, 122 S.Ct. 775 (“Where the licensing official enjoys unduly broad discretion in determining whether to grant or deny a permit, there is a risk that he will favor or disfavor speech based on its content.”); Forsyth, 505 U.S. at 131, 112 S.Ct. 2395 (“If the permit scheme involves appraisal of facts, the exercise of judgment, and the *157formation of an opinion by the licensing authority, the danger of censorship and of abridgment of our precious First Amendment freedoms is too great to be permitted.”) (internal citations omitted).
Far from precise, much of Act 64 is more a theory than a body of legal rules. What it actually means in practice has been, in a literal multitude of critical respects, simply left to future executive or judicial rulings. Act 64 bristles with interpretive issues — the meaning of “anything of value,” “candidate,” “for the purpose of influencing an election,” “primarily benefits six or fewer candidates,” “single source,” “affirmative action to become a candidate,” “services by individuals volunteering their time,” and so on — and with valuation questions — of mileage, use of a room, office, computer, phone, professional services, etc. — -and leaves resolution of all of these issues to those who must administer and enforce the statute. See Vt. Stat. Ann. tit. 17, § 2801; 2001 Guide, supra; see also discussion infra Part 111(h).
f) Appropriate Level of Scrutiny
It is standard First Amendment jurisprudence that governmental restraints on protected speech must be subjected to exacting scrutiny to survive a constitutional challenge. See Buckley, 424 U.S. at 16, 44-45, 96 S.Ct. 612 (referring to the “exacting scrutiny required by the First Amendment,” and applying exacting scrutiny to “limitations on core First Amendment rights of political expression”); Smith v. California, 361 U.S. 147, 151, 80 S.Ct. 215, 4 L.Ed.2d 205 (1959) (applying “stricter standards” to a statute that has “a potentially inhibiting effect on speech,” and noting that “a man may the less be required to act at his peril here, because the free dissemination of ideas may be the loser”) (citing Winters v. New York, 333 U.S. 507, 509-10, 517-18, 68 S.Ct. 665, 92 L.Ed. 840 (1948)); see also McIntyre, 514 U.S. at 347, 115 S.Ct. 1511 (applying exacting scrutiny to invalidate an Ohio law that prohibited the distribution of anonymous campaign literature); Meyer v. Grant, 486 U.S. 414, 420, 108 S.Ct. 1886, 100 L.Ed.2d 425 (1988) (holding that a statute prohibiting use of paid petition circulators burdens core political speech and is therefore subject to exacting scrutiny); Lerman v. Bd. of Elections, 232 F.3d 135, 146 (2d Cir.2000) (applying “exacting scrutiny” to restriction of “core political speech” in overturning local residency requirement for petition witnesses).
The most exacting scrutiny must be given to legislation that expressly seeks to reallocate political power — in the view of Act 64’s 'proponents, from “special interests” to “ordinary citizens” — by limiting the political activity of candidates for office and their supporters. See Buckley, 424 U.S. at 14-15, 96 S.Ct. 612 (calling political campaigns the “fullest and most urgent application” of the First Amendment guarantee, and invoking the “ ‘profound national commitment to the principle that debate on public issues should be uninhibited, robust, and wide-open’ ”) (quoting New York Times v. Sullivan, 376 U.S. 254, 270, 84 S.Ct. 710, 11 L.Ed.2d. 686 (1964)).
As Justices. Brandéis and Black have reminded us, the high-mindedness of a law’s proponents is no guarantee that it does not flagrantly violate principles of freedom of expression. This is particularly true with regard to legislation that was, as I detail in Part VI(d) of this dissent, examined in the legislative process more for the nobility of its stated purposes than for'what it actually says. Since Act 64’s passage, surprise at its actual provisions and actual effects has been expressed by *158many of the law’s proponents.3 Notably, one vigorous supporter who has been described as its author, Anthony Pollina, see Vermont Reformer Says Law He Authored is Unconstitutional, Political Finance, The Newsletter, March, 2002, has since sought to run for office and brought a lawsuit claiming that Act 64 violates the First Amendment. See Ross Sneyd, Progressives Sue to Ensure Public Financing for Pollina, Associated Press, Mar. 12, 2002.
Moreover, high-mindedness is, for some, a mode of self-deception obscuring self-serving motives or, for others, a facade useful in disadvantaging political opponents, routinely referred to as “special interests.” When campaign finance legislation is considered by those in power, there is both motive and opportunity to craft rules that will restrain the political activity of opponents. My colleagues caution that the self-interest of incumbents should not cause us to presume that such legislation is unconstitutional. However, most of the major factual premises underlying Act 64 posit incumbents who value reelection over their duties to constituents and personal honor. These premises should not hold center stage when examining the ostensible justifications for Act 64 only to disappear when scrutinizing what its actual effects will be.
I also note that some of Act 64’s proponents relied upon and quoted by my colleagues have themselves demonstrated the importance of self-interest among its supporters. For example, one (then) incumbent state senator testified that Act 64 was needed to stop the “arms race” in which some of her opponents buy “ads” and “yard signs” that catch voters’ attention and cause voters to wonder whether she is running for reelection. Trial Tr. vol. IX at 148 (Elizabeth Ready). Another, as noted, *159brought a constitutional challenge to Act 64 when it impeded his political career.
Moreover, our experience in a similar area suggests that great caution is in order where incumbent legislators pass laws affecting their electoral fate. Legislatures can directly affect the outcome of elections through two kinds of legislation: reapportionment and campaign finance regulation. Our experience with reapportionment is that, over time, the self-interest of incumbents has become the sole guiding star. See infra Part IV(c).
Indeed, whenever Congress takes up legislation involving campaign finance, the press now openly discusses how various proposals will affect the prospects of particular political parties and candidates. See, e.g., Ruth Marcus & Dan Balz, Democrats Have Fresh Doubts on “Soft Money" Ban; Some Fear GOP Would Gain Edge in Campaign Finances, Washington Post, Mar. 5, 2001, at Al; John Mintz, McCain’s “Soft Money" Pledge Alarms GOP; Republican Leaders Say Curbs Would Hurt Party’s Election Chances, Give Fund-Raising Edge to Democrats, Labor Unions, Washington Post, Feb. 22, 2000, at A6. The assumption that these possible effects never enter the minds of the candidates for reelection who enact such legislation might be questioned by even the least cynical observer. Truly searching scrutiny of campaign finance legislation is therefore essential.
I respectfully submit that my colleagues have not given this legislation careful, much less exacting, scrutiny. Their opinion describes the provisions of Act 64 in only cursory fashion. In a show of deference exceeding even that accorded decisions of an administrative body, it accepts the theory and factual assumptions proffered by the law’s supporters at face value even when their actions belie their words. See infra Part VI(d) (failure to comply with reporting requirements); infra Parts IV(b)(l)(C), IV(c) (spending more than Act 64’s limits); supra Part 11(f) (bringing a lawsuit to challenge the constitutionality of the Act); infra V(e) (same). And it ignores the holding of Buckley that expenditure limits are per se unconstitutional.
Even without the direct precedent of Buckley, First Amendment jurisprudence does not allow laws that burden and prohibit political advocacy to be justified by the proffer of a theory based on spoken and unspoken factual assumptions without the most exacting judicial scrutiny of that theory, those factual assumptions, and the actual provisions of the law as enacted. Such scrutiny requires an examination of the details of the law passed, the degree of burden it imposes on protected speech, and the interests asserted as its justification. Accordingly, I turn to what Buckley directs as the first step of constitutional analysis, the details of the law challenged. 424 U.S. at 12, 96 S.Ct. 612.
III. THE PROVISIONS OF ACT 64 ■
a) Overview
Beginning with an overview, Act 64 limits the amount of resources — money and things of value — that may be used by candidates in campaigns and that may be provided by individual supporters or political parties.4 See Vt. Stat. Ann. tit. 17, §§ 2805, 2805a. It therefore contains lim*160its on direct expenditures of money or use of things of value by candidates for electoral purposes and on direct contributions of money or things of value to campaigns. See id. Such limits would not necessarily reach activities that consume resources purchased and used by individuals and political parties to support a candidate’s campaign. However, Act 64 styles these activities “related expenditures” and treats them both as candidate expenditures and as contributions to a candidate subject to the statutory limits on those expenditures and contributions. See id. §§ 2809(a), (b). Act 64 also requires that, for purposes of applying the limits, contributions to, and campaign expenditures by, state, county, and local affiliates of political parties are determined by aggregating, that is, by treating all affiliates as a single unit. See id. §§ 2801(5), 2805(a), 2809(d). To that end, Act 64 requires that all money raised by state, county, and local affiliates be put in a single bank account. See supra note 1, and accompanying text.
Act 64 provides a public financing option for candidates for Governor and Lieutenant Governor. See id. §§ 2851-2856. Eligibility for public financing turns in part on Act 64’s definitions of “contributions” and “expenditures,” and, therefore, of “related expenditures.” Were a candidate to raise or expend more than $500 before February 15 of the election year — or have supporters, including a political party, make related expenditures in excess of that amount — the candidate would not be eligible for public financing. See id. § 2853(a).
An effort like Act 64 of course must provide some definition of the conduct regulated and the substance of what is prohibited and what is permitted. Where limits on campaign expenditures and contributions are imposed by dollar value, a time frame must be selected. The statutory scheme must also include an enforcement scheme, a delicate matter when electoral speech by candidates and their supporters is regulated by governmental officials— often their opponents — and a multitude of statutory ambiguities and problems of interpretation and valuation abound. Scrutiny of the details of such regulation is necessary to inform the constitutional inquiry regarding the degree of impact on protected speech and conduct, the requisite nexus between the regulation and constitutionally permissible goals, and the accuracy, reliability, and likely adherence to those goals of the designated enforcement mechanisms.
b) Two-Year Cycle
As noted, establishing a basic legal framework for regulating political campaigns first requires selection of a time frame(s) for the provision of public financing and for totaling candidate expenditures, contributions, and related expenditures by individuals and political parties in order to enforce limits on their size. Act 64 is schizophrenic in that regard. For purposes of public financing, it establishes separate time periods and separate funding for primary and general elections in recognition of the obvious fact that some candidates must fund both a primary and general election campaign while others need fund only a general election. See id. § 2855(a).
For purposes of limiting contributions and expenditures, however, Act 64 imposes a so-called “two-year cycle” approach. See id. §§ 2805(a), 2805a(a); see also 2001 Guide, supra. Under that approach, expenditures by candidates, contributions, and related expenditures by individuals and political parties supporting candidates are totaled over a two-year period for purposes of enforcing the statutory limitations. The effect of the two-year cycle is *161not inconsequential. Vermont, like most American states, provides both for primaries and for subsequent general elections. See Vt. Stat. Ann. tit. 17, §§ 2103(15),(25), 2351. Because the two-year cycle lumps these elections together, contribution and expenditure limits, including related expenditures by individuals and political parties, are imposed on the total raised and spent by individual candidates in both electoral periods. In other words, Act 64 limits a candidate who must wage a serious primary fight to the same amount of total financing as his or her general election opponent who did not face a primary contest.
The two-year cycle introduces another complexity — and creates much room for anti-democratic manipulation — because party primaries in Vermont are not restricted to voters registered in the particular party but are open to all voters, including those registered in other parties. See id. § 2363; see also Ian Urbina, Leveling Politics in the Green Mountain State, The American Prospect, Sept. 25, 2000, at 41 (discussing Vermont’s cross-over voting in primaries); Vermont’s Senate Race, The Common Man, The Economist, Sept. 5, 1998, at 25. The amount that a candidate must spend in a primary, therefore, may be substantially affected by voters who are seeking to disadvantage the candidate in the general election.
c) Limits on Expenditures by Candidates
Act 64 defines candidate “expenditures” to include “payments, distributions, and disbursements of money or anything of value for the purpose of influencing an election.” Vt. Stat. Ann. tit. 17, § 2801(3).5 The breadth of this language is indisputable. Given its ordinary meaning, the language includes the value of the use of phones, computers, offices, rooms in residences or elsewhere, paper, pencils, autos, etc. See 2001 Guide, supra; 1999 Memorandum, supra. For example, according to Vermont’s Secretary of State, a candidate’s use of an auto is an expenditure. See 2001 Memorandum, supra. Candidates therefore may not drive their personal vehicles for campaign purposes without recording every mile driven and treating the costs of that driving as a campaign expenditure. See id. Vermont’s Secretary of State has suggested that 31<t per mile is an accurate measure of expense for this purpose. See id. (She has not changed this figure notwithstanding the facts that the price of gasoline has risen and official Vermont travel is now compensated at 37/£<c per mile. See id; Vermont Dept, of Personnel, Collective Bargaining Agreements, at http://www.Vermontpersonnel.org/employ-ee/labor_dcba.cfm (effective July 1, 2003 to June 30, 2005) (setting mileage reimbursement for Vermont employees at level established by the U.S. General Services Administration, currently 37<t)). These expenditure limits also apply to candidates who exclusively use personal funds to fund their campaigns. See Vt. Stat. Ann. tit. 17, § 2805a(a).
Two terms are critical to determining what activities are “expenditures” subject to the limits: “for the purpose of influencing an election,” see id. § 2801(3), and “candidate,” see id. § 2801(1). Notwithstanding the assertion of a footnote in my *162colleagues’ opinion,6 the breadth of the phrase “for the purpose of influencing an election” is such as to be in substantial part hopelessly ambiguous and was said by the Supreme Court in Buckley7 to be unconstitutionally vague. 424 U.S. at 44, 80, 96 S.Ct. 612; see supra note 6. At one end of an interpretive spectrum, that phrase *163would probably not include a candidate’s cost of driving to a town hall to register to vote and, later, of driving to vote, although even that driving fits within Act 64’s literal definition of expenditure. At the other end of the spectrum, candidate Jones’s purchase of an ad stating “Vote for Jones Next Tuesday” would certainly be an expenditure. Between those extremes are a multitude of activities that may influence an upcoming election but lack an accompanying statement of express purpose. As to these, the statute offers no guidance.
Potentially the most significant area of ambiguity involves activities of incumbent officials. Members of the executive and legislative branches engage in relatively continuous communication with the public that involves the use of resources in a way that will help a reelection effort and would therefore fit within the definition of “expenditure,” if done by a “candidate” “for the purpose of influencing an election.” For example, Vermont’s Secretary of State has a publicly funded website that does not avoid capitalizing on the political opportunity offered. See Vermont Secretary of State Website, at http:// www.sec.state.vt.us/. The home page features a photo of her with a backdrop of mountains and pine trees. Other pages of the site also find it necessary bo include a photo of the incumbent. Such a website not only offers favorable exposure but also involves the preparation of materials easily put to political use as flyers and ads.8 The *164Vermont Democratic Party website underlines the political usefulness of the official Secretary of State website by offering visitors to the party’s website a link to the official site. See Vermont Democratic Party Website, at http://www.vtdemoc-rats.org. See also Burlington GOP Website, at http://www.burlingtongop.com (linking to republican Jim Douglas’ official Vermont Governor Website).
The statute offers no guidance on the many questions of how the relevant language is to be applied in practice to incumbents’ activities, even though the answers may have a decisive impact on particular candidates. If most of the resource-consuming activities of officeholders are not “expenditures” because they occur in the course of the officeholders’ public duties, incumbents will have an enormous advantage over challengers under expenditure limits. If most of these activities are “expenditures,” an incumbent officeholder might well use the bulk of permitted expenditures in the first year of the two-year cycle. There are also hundreds of intermediary positions, all of which are arbitrary to one degree or another.
Some interpretive guidance, but not much, may be gleaned from the definition of “candidate.” A “candidate” is someone who “has taken affirmative action to become a candidate.”9 Contrary to the assertion in a footnote in my colleagues’ opinion,10 the elastic phrase “affirmative action” and the self-evident circularity of using a word in its own definition leave ample room for disputes over the definition’s meaning. Persons who fully intend to run for office, but have not announced, engage in all sorts of conduct to bring themselves into the public eye, to appear interested and informed on public issues, and to commend themselves as potential candidates to the media and political lead*165ers. They attend meetings of school boards, selectmen, and various public forums. See Trial Tr. vol. IX at 135 (Elizabeth Ready). Even these efforts require the use of money or things of value, are intended to influence the outcome of an election, and therefore meet the definition of expenditure if done by a “candidate.” That issue thus turns on whether such conduct constitutes an “affirmative action.”
A degree of clarity is added by the next sentence of the definition, which states that affirmative action shall include three kinds of acts. However, most of the basic ambiguity is left in place because the use of language of inclusion does not suggest that what follows is an exclusive list of “affirmative act[s].” The first set of included acts involves accepting “contributions” or making “expenditures” in excess of a total of $500. See id. § 2801(1)(A). This brings into the definition of candidate all of the ambiguities of the term “expenditure” — and “related expenditure” — including the pre-campaign conduct noted above that is fully intended to influence the outcome of an election. In addition, as the Secretary of State has noted, an individual not fully decided upon, but considering, a run for statewide office will trigger the definition of candidacy by driving four round trips between Swanton and Brattle-boro at (the now obsolete) 31<t per mile. See 2001 Memorandum, supra. A person’s official candidacy can also be triggered by acts of the person’s, political party deemed to be “related expenditures” valued in excess of $500. See Vt. Stat. Ann. tit. 17, §§ 2809, 2853(a); see also Ross Sneyd, Progressives’ Poll Raises Question About Public Financing, Associated Press, Feb. 21, 2002. The two other acts included are filing a petition for nomination or announcing a candidacy. See Vt. Stat. Ann. tit. 17, §§ 2801(1)(B), (1)(C). However, these provisions clarify things that were not ambiguous.
The limits on expenditures by candidates over the two-year cycle vary with the office sought, as follows:
Governor — $300,000
Lieutenant governor — $100,000
Other statewide offices — $45,000
State senator — $4,000 plus $2,500 for each additional seat in the district
County office — $4,000
State representative, single member district — $2,000, two member district— $3,000.
See id. § 2805a(a).
Incumbents may spend 85% — except for legislators, who may spend 90% — of the expenditure limits. See id. § 2805a(c).
d) Limits on Contributions to Candidates
“Contributions” are similarly broadly defined as any “payment, distribution, advance, deposit, loan or gift of money or anything of value paid or promised to be paid to a person for the purpose of influencing an election .... ” Id. § 2801(2).11 The limits apply to “single source” donors, defined as “an individual, partnership, corporation, association, labor organization or any other organization or group of persons *166which is not a political committee or political party.” See id. §§ 2801(6), 2805(a). Exempted from the definition of contribution are “services provided without compensation by individuals volunteering their time on behalf of a candidate.” Id. § 2801(2).
Ambiguities lurk in the words “paid to a candidate” with regard to a resource used in a campaign by the resource’s owner, for example, a campaign worker’s use of a personal vehicle. Some of these ambiguities are cured in part by the definition of “related expenditures,” discussed below.
Uncured are the ambiguities in the term “services provided without compensation” by volunteers. These uncertainties are particularly great — and very important— with regard to professional services, particularly legal services, which are of great value to a candidate who runs for office under Act 64. A few of the many questions that will arise are: If an employee or partner engages in political activity during working hours and the firm does not dock the appropriate amount of compensation, is that a contribution by the firm? Can professionals who are not solo practitioners provide free professional services to candidates? If a professional is not generally free under a firm’s employment arrangements to moonlight professional services to others, is the provision of such services to a candidate in non-working hours a contribution by the firm to the candidate valued according to the firm’s usual billing rate? And so on.
The definition of “single source” also contains ambiguities. For example, rendering a non-obvious interpretation, the Secretary of State has stated that partnerships may make contributions as separate entities from the partners themselves, who are free to make identical contributions as individuals. See 2001 Guide, supra. Questions also arise about corporations with only one shareholder, e.g., are professional corporations operated by solo practitioners firms separate from their owners for purposes of the contribution limits?
As noted, the contribution limits also apply to money, goods, or services provided to political parties, and the various affiliates of a party are treated as one unit for the purpose of these limits. That is, a contribution to a Democratic town committee is limited as noted immediately infra, see Vt. Stat. Ann. tit. 17, §§ 2801(5), 2805(a), and is viewed as a contribution to all Democratic town, county, and state committees. Further, that contribution must be paid into a single statewide bank account, see supra note 1 and accompanying text. This provision therefore necessitates statewide reporting and control of spending by affiliates, which can receive funds only from the statewide bank account.
The limits on contributions also vary by office sought and political committee as follows:
Political party/political committee— $2,000
Statewide office — $400
State senate/county office — $800
State representative/local office — $200.
See id. § 2805(a).
e) Limits on “Related Expenditures”
Turning now to “related expenditures,” they are defined as “expenditures” (including, therefore, things of value and importing the ambiguities described above) “intentionally facilitated by, solicited by or approved by the candidate.” Id. § 2809(c).12 They include “expenditures” *167by individual supporters of candidates and by the political parties that sponsor the candidates. “Related expenditures” therefore include the value of mileage driven by campaign volunteers, the use by a volunteer of a residence, house phone, or computer, or other expenditures by volunteers for items such as paper, pens, etc. They also include the cost of polling, the printed information on candidates, and offices and phones, etc., provided by political parties.
The law regulates “related expenditures” in two ways. First, it treats them as contributions subject to the limits on contributions described above. Every use of an in-kind resource — car, phone, computer, etc. — must thus be valued and totaled with direct cash contributions on an ongoing basis. See id. § 2809(a). Use of the in-kind resource must cease when the contribution limit is reached.
Second, Act 64 also treats related expenditures as candidate expenditures. When an individual’s or party’s related expenditures exceed $50, the candidate on whose behalf they were made must treat them as campaign expenditures limited by the statute. See id. § 2809(b).13 This means that, over a two-year period, every supporter of a candidate who drives the family car to campaign meetings or provides paper, pens, phones, refreshments, or rooms for meetings, must keep a running total and, when the total exceeds $50 — driving an average of seven miles per month at 31$ per mile triggers this — the candidate must fit the amount under the statutory limit on candidate expenditures.
As noted, related expenditures include activities of political parties, such as polls, mailings, dinners, and other events.14 If such party activities fall within the definition, they must be treated as contributions to, and expenditures by, the candidate. Such activities can, therefore, trigger an official candidacy, destroy eligibility for public financing, or exhaust the total that a candidate may spend in the two-year cycle. See 2001 Guide, supra; see also Vt. Stat. *168Ann. tit. 17, §§ 2805a, 2853. It will be recalled that the district court struck down the provisions of Act 64 subjecting related expenditures by parties to the Act’s contribution limits, e.g., no more than $400 in cash or related expenditures to a candidate for statewide office. Given the holding of Colorado II, 533 U.S. at 465, these provisions are now revived.
To illustrate the effect of these provisions, I have added as Appendix A a letter from the Secretary of State responding to an inquiry as to whether certain party activities should be deemed related expenditures attributable to a particular candidate. The letter makes it clear that parties and their candidates can avoid the risk of an unexpected attribution of a large sum to a campaign only by eschewing normal and necessary political activities. For example, according to the Secretary of State, there is danger in sharing party-funded poll results with candidates or potential candidates; candidates or potential candidates must avoid any knowledge of party mailings; candidates must avoid participation in planning or even approving a party event- (a party event at which a candidate is introduced apparently must be a surprise party); and parties must avoid mailings that have a “primary thrust” of supporting candidates. See Appendix A, infra. . The Secretary and Attorney General wisely advise, “Each party and potential candidate should review proposed activities with their own counsel,” id., although this will be difficult for the candidate where he or she must remain ignorant of the event.
f) Costs of Compliance
The costs of complying with the law’s various provisions are not exempted from the limits on expenditures. See 2001 Guide, supra. Raising contributions itself costs money and is an expenditure. See id. Indeed, the limits on the size of contributions increase these fundraising costs. Moreover, for a candidate to comply with the expenditure limits, he or she must, over a two-year period, either restrict the activities of supporters and the party organization, including the driving of personal vehicles, that constitute related expenditures, or keep in constant contact with supporters and the organization to monitor the size of such expenditures. A failure either to restrict or to monitor related expenditures will create the very real risk that, at a critical stage of the' campaign, several supporters or party officials will report that they have exceeded the $50 limit and have, therefore, made expenditures that must be counted as candidate expenditures and may exhaust the campaign limit. In a statewide campaign, the monitoring and limiting of related expenditures by individuals or party organizations might well require a full-time staff member.
Moreover, a candidate who does not have legal counsel and other professional services runs great risks. The ambiguities detailed above and problems of valuation will confront candidates and supporters— or at least those who seek to comply with the law as written — with an ongoing need for professional advice. In fact, the Secretary of State and Attorney General advise that parties and candidates retain their own separate attorneys. See Appendix A. As noted, the cost of these attorneys, or the value of their services if obtained as unpaid-for related expenditures by individuals or a political party, are éxpenditures. See Vt. Stat. Ann. tit. 17, § 2801(3); infra Part V(e).
g) Treatment of the Press
Although this legislation'was fostered by groups experienced in these matters, it does not contain the usual exemption for *169editorials, op-ed pieces, or even letters to the editor that endorse a particular candidate. See, e.g., N.Y. Elec. Law § 14-124 (exempting “any person, association or corporation engaged in the publication or distribution of any newspaper or other publication issued at regular intervals in respect to the ordinary conduct of such business”); Conn. Gen. Stat. § 9-333w(c) (exempting “any editorial, news story, or commentary published in any newspaper, magazine or journal on its own behalf and upon its own responsibility and for which it does not charge or receive any compensation whatsoever”).
Vermont’s Secretary of State (the one with the photo-heavy website) has warned that if any individual or organization “requests a photograph, written presentation, or other assistance or information and informs the candidate that the requested information will be used in a publication ... [providing such] will trigger a related expenditure.” 2001 Guide, supra. Therefore when: (i) a Vermont candidate meets with an editorial board, commentator, or columnist hoping for an endorsement; (ii) a supporter of the candidate uses campaign materials to author an op-ed article for a paper; (iii) a campaign official sends a letter to the editor; or (iv) a campaign official conveys information to a reporter hoping for a news story; the value of any such publication is, under Act 64, a contribution and a related expenditure. See Vt. Stat. Ann. tit. 17, § 2809(c). See also infra Part IV(d).
h) Administration and Enforcement
I turn now to the processes governing administration and enforcement of this law. Power is delegated to the Secretary of State to “adopt rules necessary to administer the provisions” regarding related expenditures. Vt. Stat. Ann. tit. 17, § 2809(f). Additionally, the Secretary of State has a general administrative role under Act 64, see, e.g., id. §§ 2803, 2810a, and she has actively offered her interpretations of the scope and application of the various provisions of Act 64, see, e.g., 2001 Guide, supra; 1999 Memorandum, supra; 2001 Memorandum, supra. As the discussion above indicates, interpretive and valuation questions abound and, as Appendix A indicates, answers that are not prolix or ambiguous are often not available. Moreover, there is at present every indication that the power to “adopt rules necessary to administer” the statute will be viewed by the Secretary of State as a very broad delegation of power. See infra Part V(e). For example, interpreting a provision requiring that all contributions in excess of $50 be made by check, the Secretary has said that Act 64 allows so-called “pass-the-hat” fundraisers at which persons may anonymously contribute up to $50 in cash. See 2001 Guide, supra. Because the givers are anonymous, a candidate is not expected to monitor how many times an individual may have put $50 into ever-moving “hats” at several “pass-the-hat” fundraisers. This ruling thus promotes fundrais-ing practices that do not really control the size of cash contributions, unless, of course, an anonymous donor foolishly drops a $100 bill into a hat.
Finally, candidates who want to seek a determination that an expenditure is a related expenditure made on behalf of their opponents may bring an expedited action in the Vermont Superior Court. See Vt. Stat. Ann. tit. 17, § 2809(e). Candidates wanting clarification of their own expenditures, or persons wishing to make expenditures on behalf of candidates, may request advisory opinions from the Secretary of State. See 2001 Guide, supra. The costs of bringing or defending such actions, or making such inquiries, are not exempted from the definition of expenditure.
*170IV. THE BURDEN ON PROTECTED SPEECH
a) The Burden on Grassroots Political Activity
I begin with Act 64’s burdening of grassroots political activities, not only because such activities are core-protected speech under the First Amendment — not to say indispensable to our democracy — but also because proponents of Act 64 purport to justify its ubiquitously restrictive provisions in the name of increasing and enhancing such activities. See 1997 Vt. Laws P.A. 64 (H. 28) (findings nos. 6 and 8); Trial Tr. vol. IX at 124, 131 (Elizabeth Ready). In fact, Act 64 relentlessly limits such activities and often renders them impossible. Indeed, the Act’s most intrusive impact is not on the rich and powerful, who if necessary can engage in constitutionally protected independent political activity, but on the ordinary citizen, who needs to participate in organized activity to have a political voice.
As the Supreme Court noted in Buckley, even the humblest kind of political activity requires the expenditure of resources. Buckley, 424 U.S. at 19, 96 S.Ct. 612. If the law as drafted is upheld, the quality and quantity of grassroots activities will be severely diminished, although one may question whether a free people will even attempt serious compliance with a law that, were the constitutional stakes not so great, might easily be regarded as an act of legislative silliness. See infra Part VI(d) (failure of Act 64 supporters to comply with its reporting requirements); infra note 32 (reporting mileage expenses in round numbers, despite per mile valuation of 31$); supra note 4 (lapel buttons and bumper stickers must identify who paid for them, the payor’s address, and the candidate benefited).
1) Burden on Volunteer Activity
As noted, Act 64 treats all related expenditures as contributions and, when they exceed $50, as expenditures by the candidate whose candidacy was supported. See Vt. Stat. Ann. tit. 17, § 2809(b). Related expenditures must therefore be counted in determining whether the candidate has complied with Act 64’s spending limits. Because in-kind expenditures are included within related expenditures, any supporter engaging in the most common kind of political activities must keep detailed records over a two-year period of their value— every mile driven, every stamp used, every use of a residence for campaign events, refreshments, pads, pencils, use of phones, etc. — so that the total amount of such in-kind expenditures can be determined. If a supporter’s related expenditures, alone or when added to monetary contributions to a candidate, reach the contribution limit, the supporter must stop all activities — even driving to a campaign event — nr violate the law.
For example, if a supporter holds a “meet the candidate” event in his or her house, the value of the space used, possibly the costs of refreshments,15 and the purchase of stamps and envelopes for mailing *171invitations to local citizens are all related expenditures. See 1999 Memorandum, supra. Vermont’s Secretary of State has stated that it usually takes one hundred invitations to attract twenty persons to a “meet the candidate” event. See id. Thirty-seven dollars would thus be used for postage alone for one event for twenty people. See United States Postal Service, First-Class Mail Rate Highlights, available at http://www.usps.com/rate-case/first.htm. As the Secretary of State of Vermont has noted, such “meet the candidate” events are therefore severely limited by Act 64. See 1999 Memorandum, supra.
Adding to Act 64’s intrusiveness on grassroots activities is its treatment of such related expenditures as expenditures by the candidate. Driving to meetings is among the most garden variety of grassroots political activities. But, under the law, a supporter who averages seven miles per month over the two-year cycle will have exceeded $50 in mileage expenses, and the candidate in question must treat that and all other resource-consuming activities by the individual as campaign expenditures.
One effect is to prevent a candidate’s supporters from exercising free choice as to what activities to undertake. Because the candidate’s total expenditures are limited, the activities of all supporters must be coordinated and controlled top-down by the candidate for two full years so that the candidate can budget a campaign and not have to end it prematurely because of belatedly discovered related expenditures in excess of $50 that exhaust the expenditure limits. Were that to happen, a candidate, or any supporter over the $50 limit, would not even be able to drive the family car to the local town green to make a speech.16
Act 64 therefore creates great incentives for campaigns to reduce the level of grassroots activities. The danger of unexpectedly reaching the expenditure limits will require a campaign to monitor, at a cost of time and resources, those grassroots activities it allows. Fewer such activities will be allowed because of these monitoring costs and because expenditure limits require that priority be given to activities that reach the largest number of voters, such as media advertising.
2) Burden on Local Party-Funded Activity
Many grassroots political activities are sponsored and subsidized by local political party affiliates. See Trial Tr. vol. IX at 138-39 (Elizabeth Ready) (party helps with “grass roots organizing,” “voter I.D.,” and “get-out-the-vote”). Act 64 diminishes almost to the point of elimination financial support for local party activity by treating all state, county, and local party committees as a single fundraising unit for purposes of raising, and therefore spending, money. Because all contributions must go to the state party account, all expenditures by every local committee must necessarily be funded out of it. Vt. Stat. Ann. tit. 17, §§ 2801(5), 2831; see supra note 1. The effect is, first, to reduce severely the amount of funding for such activity — under the limits on party related expenditures and contributions — and, second, to force *172political parties into some form of top-down control — under the single unit/single-bank-account rule for contributions to parties. See Secretary of State Being Criticized for Fund Raising Ruling, Associated Press, May 28, 1999 (reporting that both Republican and Democratic party leaders were shocked by the single unit rule); supra note 1. A town committee of a party may not even hold an organizational event with refreshments for its members without obtaining the modest funding needed from the statewide authority.
Moreover, limits on contributions and expenditures force political decisionmakers to give priority to activities that reach the largest number of voters. It is now known that Act 64 forces party committees, even without the newly-revived limits on party contributions and related expenditures, to concentrate more on mass media activities than grassroots activities. See 2001 Memorandum, supra; Campaigns Meant Cash for Vermont Media, Associated Press, Nov. 10, 2000 (“ ‘That was one of the unintended consequences of the campaign finance law, that we saw much more spending on the media,’ [Secretary of State Markowitz] said.”).
b) The Burden on Candidates’ Speech
The district court concluded that Act 64’s limits on campaign expenditures are based on past experience and, with limited exceptions, are substantially the same as average expenditures by candidates in the past. See Landell v. Sorrell, 118 F.Supp.2d 459, 471-72 (D.Vt.2000). Putting aside for purposes of argument that expenditure limits are per se unconstitutional under Buckley, the level of the limits set by Act 64 clearly places unconstitutional restraints on the speech of candidates for office. First, past experience is no guide. Second, average spending in past elections is a standard that strongly favors incumbents and imposes a one-size-fits-all philosophy that severely constricts debate in the most important elections. Third, Act 64’s spending levels are so low that, combined with the draconian restrictions on party spending, they will drastically reduce political debate in Vermont.
1) Past Experience
A) Inaccuracy of Reports From Past Elections
It is impossible to determine the level of relevant campaign spending by Vermont candidates in the past, that is, “expenditures” using Act 64’s definitions. It is not altogether clear what evidence the district court specifically considered in reaching its conclusions. However, on the face of the district court’s decision, it appears that the court relied heavily on testimony, some of which was conflicting, see id. at 470-72, and did not scrutinize in detail documentary evidence of past practices.
More significantly, even the candidate disclosure reports filed under Vermont law for past elections will vastly understate the level of spending when Act 64’s two-year election cycle and its new and much broader definitions of expenditures and related expenditures are used. There are some expenditure reports in the Trial Exhibits, but they are limited to particular candidates’ out-of-pocket expenditures made during a “campaign.” See, e.g., Trial Exs. vol. IV at E-1311 (Campaign Finance Report of Peter Brownell). Act 64’s limits on expenditures, however, apply, as noted, to all expenditures made over a two-year period immediately following the last general election and ending with the next general election. See Vt. Stat. Ann. tit. 17, §§ 2801(9), 2805a(a).
Furthermore, under prior law, there was no provision regarding related expenditures. There was, therefore, no reason even to collect information on, much less to *173calculate and report, related expenditures by supporters and political parties. In particular, there was no reason to calculate the value of in-kind related expenditures by supporters, such as mileage, all of which count toward the expenditure limits under Act 64. Finally, there was also no need under prior law for candidates to segregate and calculate expenditures on their behalf by party committees, likely a huge amount. See infra note 19 and accompanying text. The value of such support must be treated under Act 64 as a candidate expenditure. See Vt. Stat. Ann. tit. 17, § 2809(d).
We know only one thing for certain: what candidates deemed to be expenditures in the past — generally direct cash expenditures out of a campaign’s checking account — will be vastly less than what must be so regarded under Act 64’s definitions of expenditures and related expenditures.
There is another reason why prior spending is not a reliable guide for the needs of campaigns operating under Act 64. Act 64 imposes substantial costs of compliance with its terms that were not encountered under the prior law. As noted, Vermont’s Secretary of State has indicated that most candidates will be unable to proceed safely without legal advice, see Appendix A and candidates running for statewide office may need the services of an accountant as well. In the case of legislative candidates, legal assistance alone could literally exhaust all the expenditures — e.g. $2,000 for House candidates — allowable under Act 64. Again, such assistance from a candidate’s political party will be limited because retention of counsel by a party organization to help candidates would be a related expenditure allocable to individual campaigns.
Much time and possibly much support staff will also be consumed by the need to monitor, coordinate, and control related expenditures by supporters that must be charged to the campaign. Finally, some candidates may encounter costs in bringing and defending lawsuits concerning the myriad of interpretive questions that will arise as a result of Act 64’s ambiguous provisions.
B) Inadequacy of Average Spending in Past Elections as a Constitutional Standard
The district court deemed the average campaign expenditures in past elections to be a relevant legal guide, Landell v. Sorrell, 118 F.Supp.2d at 471-72, and my colleagues agree, see Maj. Op. at 130-31.
However, even if accurately determined using Act 64’s definitions, the use of average expenditures in past elections inevitably yields expenditure levels that strongly favor incumbents. First, incumbent legislators in Vermont and elsewhere have ample advantages over challengers under spending limits, discussed infra, and therefore prefer low limits. Second, the average of past expenditures is calculated by including legislative elections that were not seriously contested or perhaps not contested at all — elections in which little communication took place and little was spent. See Trial Exs. vol. Ill at E-0967 (appellees’ expert’s calculation of average expenditures, which includes low-spending candidates whose spending is unknown by assuming they spent $500, the maximum allowed before filing is required). It is altogether possible, therefore, that the average expenditure in past elections is less than the amount spent by every candidate who ran in a seriously contested race and perhaps even probable that it is less than the amount spent by any challenger who *174successfully challenged an incumbent.17 Average past spending therefore has little relevance unless the goal is to disadvantage challengers.
The use of average past expenditures is inappropriate for other reasons. The average will reflect only past patterns of citizen behavior in acquiring political information and prior methods of candidate communications with citizens. When citizens congregate in very large numbers for frequent community events, those events may well be effective vehicles for candidate communication with voters. When large community events become less frequent or less important in peoples’ lives and voters turn to other sources to acquire political information — some may rely heavily on a certain newspaper, some on particular radio or television stations, some on websites — other means of communication, perhaps far more expensive, must be used by candidates for effective communication.
Part of the problem is simply the one-size-fits-all philosophy of expenditure limits. Different legislative districts may require different modes of communication. Geographic size, the existence and nature of local newspapers, radio and television stations, demographic factors, the issues, and so on, all affect the costs of communication with voters and may do so differently in many districts for the same legislative house.
Moreover, because the one-size-fits'-all philosophy fails to recognize the differences between elections, it strikes at the heart of democracy. The view that there is an average election that can serve as the compulsory norm for all elections is quite dangerous, even apart from its pro-incumbent bias. Past averages have almost nothing to do with the communication needs in elections in which candidates strongly disagree over issues that divide large portions of the public and a clear-cut attempt is being made to alter government policies on those issues.
It is the non-average election that is often the historic election, one in which the outcome is heavily contested, the debate is most widespread, the public interest is at its highest, and the most money is spent. Such an election was the New Hampshire primary of 1968, in which Eugene McCarthy, later a plaintiff in Buckley, badly damaged a sitting President in a debate over the Vietnam war in one of the most heavily financed primary races in history. McCarthy spent a then-unprecedented $12 per vote received in that single primary. See George F. Will, Rules to Keep the Rascals In, Newsweek, Jan. 26, 1976, at 80.
Vermont had a similar election in 2000, in which civil unions and other divisive issues were at stake. See Ellen Goodman, Vermonters Are Caught up in a Civil War over Civil Unions, Boston Globe, Nov. 2, 2000, at A27; Tom Puleo, Governor’s Race Tests Vermont Values; “Gay Marriage” Issue Is Monopolizing a Bitter Battle, Hartford Courant, Oct. 30, 2000, at Al. More money was spent in the 2000 election than in any prior Vermont election. See Lawmakers To Revisit Campaign Finance Law, Associated Press, Nov. 14, 2000 (noting that the 2000 gubernatorial campaigns set the record for money spent); see also *175Ross Sneyd, Campaign 2000 Involved Lots of Spending, Associated Press, Dec. 18, 2000 (describing record spending levels for many elections across Vermont in 2000).
McCarthy’s New Hampshire campaign of 1968 had national significance, while the 2000 Vermont gubernatorial election had unquestioned state, and possibly national, ramifications. Both involved unprecedented citizen participation. See 2000 General Election Results for Gubernatorial Race, available at http://cgi.sec.state.vt.us/cgi-shl/nhayer.exe (“2000 Election Results”) (showing that voter turnout increased in Vermont by 34.6% in the 2000 election compared to previous election); Hugh Gregg, A Tall State Revisited, at app. (1993), available at http://www.politicalli-brary.org/TallState/1968dem.html. And both involved, not surprisingly, unprecedented campaign spending.
C) Evidence of Contested Elections in Vermont
Campaign finance reports of Vermont candidates provide ample evidence, of which we may take judicial notice, Fed. R.Evid. 201; Kramer v. Time Warner, Inc., 937 F.2d 767, 774 (2d Cir.1991) (court can take judicial notice of records filed with the SEC), that contested elections in Vermont involve spending well in excess of Act 64’s limits, even without including related individual and party expenditures. In the last two gubernatorial elections in Vermont, the major party candidates reported expenditures in amounts that were double or almost triple Act 64’s limits. See Campaign Finance Report of Howard Dean, Dec. 18, 2000;18 Campaign Finance Report of Ruth Dwyer, Dec. 18, 2000; Campaign Finance Report of Doug Racine, Dec. 14, 2002; Campaign Finance Report of Jim Douglas, Dec. 16, 2002. Moreover, even a third party candidate for governor exceeded the limits in one election, Campaign Finance Report of Anthony Pollina, Dec. 18, 2000, and another third party candidate came within a whisker of the limits in the next gubernatorial election, Campaign Finance Report of Cornelius Hogan, Dec. 16, 2002. In the last race for Lieutenant Governor, both major party candidates and a third party candidate exceeded the limits by 63%, 40%, and 38% respectively. See Campaign Finance Report of Peter Shumlin, Dec. 14, 2002; Campaign Finance Report of Brian E. Dubie, Dec. 16, 2002; Campaign Finance Report of Anthony Pollina, Dec. 16, 2002.
The factual support for the conclusion reached by my colleagues — that contested elections will not be substantially affected by Act 64’s limits — is found largely in opinion testimony offered by proponents of the Act. The already slim value of that testimony is further undermined by the fact that many of those witnesses, when they ran for office, actually exceeded Act 64’s limits in contested elections, again not counting related expenditures by individuals and parties. See Campaion Finance Report of Anthony Pollina, Dec. 18, 2000 (spent $335,412.46 in Governor’s race, with expenditure limit of $300,000); Campaign Finance Report of Cheryl Rivers, Dec. 18, 2000 (spent $19,290.39 in senate race, with *176expenditure limit of $9,000); Campaign Finance Report of Elizabeth Ready, Dec. 18, 2000 (spent $77,313.47 in auditor’s race, with expenditure limit of $45,000, and outspent opponent by 20%, although she had testified at trial that she would abide - by Act 64’s limit in that race); Trial Tr. vol. IX, at 147-51 (Elizabeth Ready) (testifying that she exceeded the current expenditure limits in four of her six senate races). In the view of some of these witnesses, of course, contested elections are “arms races” that should be prohibited. See Trial Tr. vol. IX, at 147-151 (Elizabeth Ready). See infra Part V(d)(l).
2) Effect of Act 64’s Expenditure Limits on Candidates
Act 64’s expenditure limits will, thereT fore, greatly hamper Vermont candidates in getting their message to the public. The Secretary of State has noted that the expenditure limit for State Treasurer— $45,000 — leaves, after advertising, “no money to hire a campaign manager, do direct mail, lawn signs or bumper stickers.” David Gram, Dems Needle Each Other On Spending in Treasurer’s Race, Associated Press, May 29, 2002. Expenditure limits should be expected to have precisely such' effects because they force candidates to give exclusive priority to the methods of communication that reach the greatest number of voters.
The Secretary of State has also noted that the “tight contribution limits” of Act 64 were part of the cause of an “unprecedented amount” of independent expenditures in the 2000 Vermont election. See 2001 Memorandum, supra. As a result, candidates complained that “mailings or advertisements made on their behalf attributed to them opinions they did not hold, or sent negative messages about their opponent, in violation of their stated intent to run a positive campaign.” Id. Expenditure limits will encourage even more extra-campaign spending and leave candidates without the means to set the record straight.
Even grassroots “meet the candidate” events in supporters’ homes are severely limited, see 1999 Memorandum, supra, as noted above, and, although my colleagues mention, among other things, town barbecues and dinners as cheap but effective campaign methods, see Maj. Op. at 130, the nature — when and where held and how often in the campaign season — usefulness- — -what kind of voters and in what numbérs attend — cost—who pays — and legal status under Act 64 — an “expenditure” or “related expenditure”- — of these events is not elaborated in the record or discussed in my colleagues’ opinion, notwithstanding the critical role such factors logically play in their opinion’s analysis. In fact, these methods may be neither cheap — at least by Act 64’s meager standards — nor effective.
There is, therefore, simply no data in the record suggesting that anything other than a drastic reduction of political speech will result from Act 64’s expenditure limits. Indeed, the effect will likely be much harsher than most would expect for the reasons that follow.
First, low limits exacerbate the highly discriminatory and arbitrary effect of Act 64’s selection of a two-year cycle. In a single-member Vermont House district, a candidate may spend — counting related individual and party expenditures — only $2,000 over the two-year cycle. Vt. Stat. Ann. tit. 17, § 2805a(a)(5). A candidate who seeks to comply fully with Act 64 will be severely constricted. Any legal fees will count toward that limit. Id. § 2801(3). Mileage and other expenses of supporters can quickly eat up the spending allowed. A candidate who has to run in a contested primary election may well be unable to *177communicate with the public at all — literally stuck in his or her driveway — in a general election against an opponent whose campaign is just beginning.
Second, the harshness of the limits on candidate expenditures is greatly exacerbated by the fact that a candidate’s campaign cannot expect the candidate’s party to provide the usual supplemental support of polls, offices, computers, phones, advertisements, mailings, and other events, such as a party-funded booth at a country fair, if the conduct “primarily benefits” fewer than seven candidates. See supra notes 12, 14. Parties may make contributions and related expenditures benefiting candidates that total, over a two-year period, no more than $400 for each candidate for statewide office, $300 for each candidate for the Senate, and $200 for each candidate for the House. See supra Part III(b)-(c). In Vermont, there are six statewide offices, thirty State Senators, and 150 State Representatives. Under Act 64, a political party — the state party and all affiliates combined — can, over a two-year period, make a total of only $41,400 in contributions to, or related expenditures on behalf of, all its candidates for non-federal office.
Although Colorado II allows such restrictions, 533 U.S. at 465, 121 S.Ct. 2351, their effect must be considered in gauging the impact of candidate expenditure limits. A statewide poll regarding candidates for the six statewide offices would cost over $6,000. Letter from Mark F. Michaud, Vermont Democratic Party, to Vermont Attorney General William Sorrell 1 (Feb. 28, 2002). If the poll data were shared with the six candidates, the poll would exceed Act 64’s limits ($400 by 6) by over 100%. See Appendix A. The full effect of Act 64’s limits has, of course, not been experienced yet,19 because the district court invalidated candidate expenditure limits and the contribution/related expenditure limits on political parties. With these limits now revived, limits on expenditures by candidates will dramatically lessen political debate in Vermont.
Finally, as noted, Act 64’s one-size-fits-all approach makes no provision for candidates to adjust to economic, demographic, cultural, or technological changes that increase the costs of campaigns. Although Act 64 is premised on the view that elections are “too expensive” — a view expressly rejected as a valid reason for expenditure limits in Buckley, 424 U.S. at 57, 96 S.Ct. 612 — and that candidates can make them cheaper, the “costs” of campaigning are not within the control of candidates.
The costs of resources to be used in campaigns are determined by competitive markets. Resources used in campaigns are also used, and far more extensively, for non-political communication. The prices of those resources are therefore set in markets that are independent of political campaigns and in which candidates for office must compete with non-political consumers. An inability to pay market price for communication resources will stifle political speech. Nevertheless, there is no provision for future inflation in Act 64’s limits, although even slight annual increases in the consumer price index will in a few short years substantially reduce further the ability of candidates to communicate with voters. For example, the cost of postage stamps is now higher than when Act 64 was passed. See It Now Costs 3 *178Cents More to Mail a First-Class Letter, N.Y. Times, June 30, 2002, at 18. For another example, the price of gasoline has risen considerably since then. (The reimbursement rate for mileage driven by federal employees has-been increased from 31$ at the time of Act 64’s passage to 37/¿<t in July, 2004. See U.S. General Services Administration, Privately Owned Vehicle Reimbursement Rates, ■ at http:// www.gsa.gov (effective Jan. 1, 2004)).
As noted above, the effect of rising costs has already been observed by Vermont’s Secretary of State. With regard to a campaign for State Treasurer — with an expenditure limit of $45,000 — she noted that, “The cost of paid media has changed quite a bit in the last four or five years. With prices for television ads, and even radio ads, running a campaign on $45,000 will leave you no money to hire a campaign manager, do direct mail, lawn signs or bumper stickers.” David Gram, Dems Needle Each Other on Spending in Treasurer’s Race, Associated Press, May 29, 2002. It goes without saying that thqre also would be no room under the spending cap for grassroots activities that would have to be included as related expenditures.
Act 64 also does not take into account the fact that a combination of demographic, cultural and technological changes may require resort .to ever more costly methods of communication. Campaigns may communicate with voters only by going to where voters are or using a medium watched or listened to by voters. My colleagues assume the existence of numerous public events in which large groups of voters frequently congregate and can be personally addressed. See Maj. Op. at 130. However, voters are not required to abide by such assumptions and to congregate in great numbers at scheduled times during political campaigns or even to welcome an interruption of free time at home by a candidate’s personal visit. Rather, they may prefer to get their information through technology that puts candidates at the mercy of a competitive market and technological advances.
For example, because the development of cable television broadened viewership opportunities for the public, it also required candidates who wished to communicate through television to buy ads on many, instead of a few, channels. For another example, the development of the Internet has made it possible for candidates to offer websites to provide information to potential voters, a fantasy fifteen years ago. The Executive Director of the Vermont Democratic Party has opined that a candidate website is now “a necessity.” Nancy Remsen, Election Notebook 2002, Burlington Free Press, Sept. 9, 2002, at IB.
Act 64 takes a Luddite view of political communication — one witness for the defense even suggested the bicycle as a means of travel in campaigns, Trial Tr. vol. IX, at 134 (Elizabeth Ready) — and prevents candidates from adjusting to new demographic and cultural patterns and costs, even though the Supreme Court has expressly declared that government is not to make that choice. See Meyer, 486 U.S. at 424, 108 S.Ct. 1886; Buckley, 424 U.S. at 57, 96 S.Ct. 612.
e) The Burden on Challengers
The fact that limits on candidate expenditures tend to disadvantage challengers in campaigns against incumbents is recognized both in the provisions of Act 64— which has slightly lower limits for incumbents — and in its legislative history. See, e.g., Hearing on H. 28 Before the Vt. House Comm. on Local Gov’t, 64th Biennial Sess. (1997) (statement of Rep. Terry Bouricius); Hearing on H. 28 Before the *179Senate Comm. on Gov’t Operations, 64th Biennial Sess. (1997) (statements of Sens. Seth Bongartz and Jean Ankeney). The Supreme Court has also noted that limits on candidate expenditures may “handicap a candidate who lacked substantial name recognition or exposure of his views before the start of the campaign.” Buckley, 424 U.S. at 57, 96 S.Ct. 612.
Incumbents in effect have capital — name recognition, an existing organization, tested donor lists, etc. — to draw upon without making expenditures as defined in Act 64, while virtually every significant capital-building activity by newcomers requires the use of resources that count toward the expenditure limits. Equally important is the fact that incumbents have methods of getting their name before the public that are not limited by Act 64, while challengers do not.
To take an example from Vermont, the State Treasurer, an elected official, publishes newspaper ads at state expense listing names of Vermonters who may have funds in dormant bank accounts, unclaimed insurance refunds, or unclaimed stock dividends. These ads have contained photos of the incumbent Treasurer and have run in the October of election years. In 2001, the ad read, “Jim Thompson, Vermont state treasurer, may have money for you.” Treasurer Candidates Show Signs of Restraint, Burlington Free Press, Sept. 23, 2002, at IB. To take another Vermont example, whereas a challenger to an incumbent Secretary of State can obtain a Madison Avenue-type website only by spending money counted as an expenditure, the incumbent can use state funds for such a site, and post on it materials casting a favorable light on the incumbent, omitting only the words “Vote for me.” See Vermont Secretary of State’s Website, at http://www.sec.state.vt.us; Office of the Attorney General Website, at http://www.atg.vt.us; see also supra Part 111(c). Moreover, the Secretary’s political party provides visitors to its site a link to the Secretary’s site. See supra Part 111(c).
The term “arms race” has acquired an almost talismanic quality in the course of this litigation, serving as a quip that answers every concern about Act 64’s effect on political speech. In fact, however, slogans about stopping the “arms race” are often cover for the disarming of challengers. See infra Part V(d).
For example, one of the witnesses whose testimony is relied upon heavily by my colleagues, Maj. Op. at 116-17, 118, 122, 127, 130-31 is a very successful electoral official in Vermont who testified to the “arms race” at the State Senate level, Trial Tr. vol. IX, at 147, 150 (Elizabeth Ready). As an incumbent and undoubtedly well-meaning supporter of Act 64’s limits, which she had exceeded (not including related individual and party expenditures) in most of her (always successful) campaigns, see id. at 147, supra Part IV(b)(l)(C), she testified that she had been forced to make larger expenditures because her opponents ran advertisements and posted yard signs that gained the attention of voters. She was then compelled to do the same, instead of relying on her preferred campaign method of person-to-person contact, because voters seeing her opponents’ ads and yard signs wondered whether she was running for reelection. See id. at 147-48. In her view, the “arms race” forcing her to run ads and post yard signs should be stopped.
So used, “arms race” is a pejorative term that refers to contested elections in which challengers spend resources to run serious campaigns. Incumbents do not restrain their own acquisition and use of perquisites of office that help them win reeleetion, but these perquisites are never *180mentioned as part of the “arms race.” For example, the witness described above, who offered yard signs as evidence of an “arms race,” now holds the elected post of State Auditor with a website that has her photo and various pages listing her goals and accomplishments. Office of the Vermont State Auditor Website, at http:// www.state.vt.us/sao. To boot, the Vermont Democratic Party website offers visitors a link to the Auditor Website. See supra Part 111(e). Expenditure limits therefore stop “arms races” by challengers, leaving incumbents with ample weapons.
Moreover, Act 64’s selection of the two-year cycle as the governing time period collapses primary and general elections under one expenditure limit and will in the main favor incumbents, who face serious primary challengers less frequently than those seeking a party nomination to challenge an incumbent. Indeed, there appears to be little other reason justifying the choice of the two-year cycle.
The degree of the adverse effect of Act 64 on challengers will depend in large part on discretionary, arbitrary, and often ad hoc rulings on what kinds of activities and speech by incumbents will be deemed to be official communication by officeholders to the public and what kinds will be deemed to be campaign expenditures. Of course, virtually every activity by an incumbent officeholder intending to seek reelection will have a political effect, and such officeholders will to one degree or another take that effect into account in determining their behavior. The law provides for marginally lower limits on expenditures by incumbents, but this largely inconsequential difference will be rendered irrelevant so long as the substantial communications by incumbents are not deemed campaign expenditures. Conversely, the advantage of incumbents under the Act’s limits will also depend on what activities by non-announced challengers are deemed to be by a “candidate” and “for the purpose of influencing an election.” See generally Vt. Stat. Ann. tit. 17, § 2801(1), (3).
These issues, of course, will likely be addressed in the first instance by an incumbent official, the Secretary of State. See 2001 Guide, supra; see also infra Part V(e). Should these rulings be adverse to incumbents — a not very likely scenario— the incumbents can overturn them by legislation. If the rulings favor incumbents, challengers have no such option.
Because the hands-off approach of my colleagues accords expansive deference to legislative judgments as to expenditure limits, see Maj. Op. at 114 & n.9, incumbents are given a weapon that can be manipulated as needed in the future. Should the limits of Act 64 prove inadequate and the “arms race” continue to result in ads and yard signs by challengers, they can be altered at will.20
Act 64’s major factual premise is that Vermont incumbents so crave reelection that they ignore official duties and personal honor to that end. My colleagues aban*181don this premise in reassuring us that self-interest will not influence campaign finance regulation despite the considerable evidence that self-interest contributed, albeit below the public radar, to the level of expenditure limits set by Act 64 and to adoption of the two-year cycle.
Moreover, there is powerful evidence that self-interest will prevail. Incumbent legislators can exercise a direct influence on the outcome of elections in two ways: campaign finance regulation and reapportionment. The importance of self-interest is dramatically confirmed by the effect of legislative reapportionment on election districts for the United States House of Representatives, an area in which courts have deferred to legislative judgment. See White v. Weiser, 412 U.S. 783, 794-95, 93 S.Ct. 2348, 37 L.Ed.2d 335 (1973) (“From the beginning, we have recognized that reapportionment is primarily a matter for legislative consideration and determination.”) (internal citation omitted); see also Miller v. Johnson, 515 U.S. 900, 915-16, 115 S.Ct. 2475, 132 L.Ed.2d 762 (1995). Reapportionment of federal House districts now has one and only one guiding star: incumbent protection. See John Harwood, No Contests: House Incumbents Tap Census, Softtvare to Get a Lock on Seats, Wall St. J., June 19, 2002, at A1 (“Thanks to the play-it-safe strategies of Republicans and Democrats alike, and to the sophisticated technology now used in redistricting, competition is being squeezed out of the House — with huge consequences.”); id. (describing the practice of “sweetheart” gerrymandering by incumbents of both parties); Richard Perez-Pena, With 2 Congressional Seats Lost, Albany Begins Battling Over Who Must Go, N.Y. Times, Jan. 22, 2002, at Bl. These sources demonstrate that reapportionment is now widely regarded as little but the “rigging” of elections in the name of the special interest — incumbency—that dominates legislative decisions. I know of no reason why the guiding star in reapportionment decisions will not become the guiding star in campaign finance regulation.
d) The Burden on the Press
As noted, the law does not exempt the media from the definitions of “contribution,” “expenditure,” or “related expenditure.” Media support is a “thing of value” that, if “facilitated” or “solicited” by a candidate, would be a “related expenditure.” See Yt. Stat. Ann. tit. 17, §§ 2801(3), 2809(c); see also 2001 Guide, supra. Indeed, as noted above, Vermont’s Secretary of State has warned candidates that providing a photo, written materials, or “other assistance or information” to anyone for use in a publication will trigger a related expenditure. See,id.
The extent of the burden imposed on the press by Act 64 is potentially vast, again depending largely on discretionary rulings by those who must administer Act 64. See infra Part V(e). Certainly, editorials or op-ed endorsement(s) of candidates fall directly within Act 64’s language, as would publication of letters to the editor from a candidate, a campaign official, or even a supporter. One witness relied upon in my colleagues’ opinion testified that she authored articles on issues for the press as a no-cost campaign tactic, Trial Tr. vol. IX, at 135 (Elizabeth Ready), but publication of these articles would clearly fall within the related expenditure language of Act 64. Media sponsorship of debates might also, particularly if some candidates were excluded or if some did not want to appear. Ordinary news stories about campaign events brought to a newspaper’s attention by a candidate or campaign official, or even news reports on interviews with candidates, also fall squarely within the *182ruling by the Secretary of State described above. See 2001 Guide, supra,.
None of this is inconsistent with Act 64’s underlying philosophy. The theory underlying Act 64 would easily include the media as a powerful special interest having a stake in government action just like any other profit-making business or organized economic interest — e.g., the newspaper quoted by my colleagues is part of a huge multi-national organization, undoubtedly one of the larger companies doing business in Vermont. See Gannett Co. Inc. Operations, available at http://www.gan-nett.com/map/units.pdf (listing The Burlington Free Press as a subsidiary). A candidate enjoying the editorial support of the local press, favorable coverage of his or her campaign events, and publication of his or her op-ed articles receives benefits bestowed at considerable expense, including past capital investment and current spending. Unless expenditure limits include the value of media support, an opponent who does not enjoy such support and labors under an expenditure limit exempting media support is in a real sense facing a candidate who is allowed to spend more because of a powerful economic supporter. Of course, unconstitutional restraints on the press are not validated by a need for fairness created by unconstitutional restraints on political candidates. See infra note 21.
However, many proposals to regulate campaign finance exempt the media, see, e.g., N.Y. Elec. Law § 14-124; Conn. Gen. Stat. § 9-333w(c), often including a definition of what organs of communication constitute exempted media, see e.g., Bipartisan Campaign Reform Act of 2002, § 201(f)(3)(B), Pub. L. No. 107-155, 116 Stat. 81 (codified at 2 U.S.C. § 431 et seq.).21 Nevertheless, Act 64 does not contain such an exception. Given the plain language of Act 64 and the consistency of its theory with that language, Act 64 can fairly be said to burden the press, and any candidate who seeks its support, quite as much as the Act burdens other candidates and their supporters.
e) The Burden on Party Affiliates
As noted, Act 64 treats a contribution to a state, county or local party affiliate as a contribution to all affiliates, and requires that all such monies be deposited in a single bank account. See Vt. Stat. Ann. tit. 17, §§ 2801(5), 2831; see also supra note 1.
My colleagues note that “the local and state affiliates will now have to record and coordinate their contributions,” but reassure us that “the provision does not impose any organizational burden on the party outside of the campaign finance realm, and requires no broader organizational reform.” Maj. Op. at 144. Of course, under Act 64, the “campaign finance realm” is not some incidental, out-of-the-way matter but covers virtually every organizational activity, including even the cost of char*183coal, hotdogs, hamburgers, and soft drinks for a town committee picnic. Moreover, the “recording] and coordina[tion]” is required to be done through a single bank account. See Vt. Stat. Ann. tit. 17, §§ 2801(5), 2831; supra, note 1. If a party town committee wants $50 for a picnic, it must petition the state party official authorized to sign checks from the statewide account to obtain the only money that can legally pay for those items. Local party-funded booths at country fairs were mentioned as a person-to-person method of campaigning by one witness relied upon by my colleagues, see Trial Tr. vol. IX, at 138 (Elizabeth Ready), but such funding must also come from the statewide account.
By requiring that the recording and coordination of all party financing be done through a statewide party organization that parcels out funds, Act 64 not only disrupts but revolutionizes the organization of American political parties and their critical role in a free society. The centralizing of party funding will of course make all grassroots political activities by local party affiliates — the indispensable stuff of American politics — subject to the whim of state party officials. In fact, when the Vermont Secretary of State ruled that a contribution to one party organization constituted a contribution to all affiliates, both Republican and Democratic state leaders registered shock — another example of the lack of scrutiny given the actual provisions of Act 64 — and opined that grassroots activities would be severely inhibited. See Secretary of State Being Criticized for Fund Raising Ruling, Associated Press, May 28, 1999 (noting one political operative’s stunned response as being “Someone’s totally taken leave of their senses”).
V. THE CONSTITUTIONAL RESOLUTION
Some of Act 64’s severe limitations on political advocacy are unconstitutional because no governmental interest, much less a compelling interest, has been offered to justify them. These include the two-year cycle for limits on contributions and expenditures, see Vt. Stat. Ann. tit. 17 §§ 2801(a), 2805(a), 2805a(a), the forced centralization of local party affiliates, see id. §§ 2801(5), 2831, and the imposition of expenditure limits on candidates’ self-funded campaigns, see generally id. § 2805a(a).
Even if expenditure limits may be constitutionally enacted notwithstanding Buckley’s holding to the contrary, Act 64’s limits are so low that they are unconstitutional by any reasonable test. Indeed, they survive in the present case only by my colleagues’ creation of a standard that allows legislatures to adopt low, incumbent-protecting limits.
With regard to the governmental interests asserted as justifications for Act 64,22 my colleagues rely upon two in particular: government’s interests in eliminating corruption or the appearance thereof and in affording candidates more time to spend *184with non-donor voters. See Maj. Op. at 124-25. As discussed in my colleagues’ opinion, however, these interests are broadly defined and include eliminating special access or the appearance of special access of donors to officeholders, reducing the influential or agenda-setting effect of “bundled” contributions, and increasing citizen confidence in the electoral process. See id. at 115-25; see also 1997 Vt. Laws P.A. 64 (H.28). All of these interests were rejected by the Supreme Court in Buckley, 424 U.S. at 25-26, 96 S.Ct. 612, but, even if they had not been, they have not been demonstrated in the present case as having a constitutionally sufficient nexus to the limits on candidate expenditures and related expenditures.
a) Restrictions on Political Activity for Which No Governmental Interests are Asserted
As to some of the restrictions on political activity imposed by Act 64, no governmental interest whatsoever has been proffered as a justification. Absent the assertion of a justifying governmental interest, any significant restraint on political speech should be struck down as per se unconstitutional. See Eu, 489 U.S. at 222, 109 S.Ct. 1013 (“If the challenged law burdens the rights of political parties and their members, it can survive constitutional scrutiny only if the State shows that it advances a compelling state interest.”)
First, no reason is offered for the adoption of the arbitrary and highly discriminatory two-year cycle for limiting contributions and expenditures. See Vt. Stat. Ann. tit. 17, §§ 2801(9), 2805(a), 2805a(a). A two-year cycle does not reduce the influence or access to officeholders of special interests, reduce time pressures on candidates, or increase citizen or voter confidence in government. Act 64’s public financing provisions recognize the self-evident need for greater financing on the part of those who must run two campaigns rather than one. See id. §§ 2855(a), (b). Collapsing primary and general elections under a single expenditure limit is thus a flat-out suppression of speech for no asserted reason, save perhaps for the unspoken reason of incumbent protection, as discussed supra.
Similarly, no justification or governmental interest is offered for Act 64’s treatment of a contribution to any party affiliate as a contribution to all affiliates, with the statutory requirement that all contributions to, and thus; expenditures by, all party affiliates — state and local — be from a single bank account. See id. §§ 2801(5), 2831; supra note 1. Because Act 64 limits contributions to, and related expenditures on behalf of, particular candidates by political parties on an aggregated basis, contributions to separate affiliates cannot serve as a conduit allowing individuals to evade the limits on single source contributions. See Colorado II, 533 U.S. at 464-65, 121 S.Ct. 2351 (permitting restriction of coordinated party expenditures to minimize circumvention of contribution limits). Neither the interests in eliminating corruption, saving candidates’ time, nor increasing confidence in government are therefore served by aggregating contributions to affiliates of political parties.
While serving none of Act 64’s asserted goals, the aggregated treatment of contributions to affiliates eliminates the right of local committees to be free from centralized control in raising funds for local party-building activities. One of the most vital and fertile areas of democratic political activity in America is the local party committee, which, while loosely related to larger party organizations, is the source of grassroots activities that permit citizens to participate and seek change. Local party *185activities are a critical means by which a changing public opinion is absorbed gradually into the political system, rather than going unheard until it reaches explosive force. Because these activities require funding, Act 64 directly impairs them. See Secretary of State Being Criticized for Fund Raising Ruling, Associated Press, May 28, 1999 (noting that leaders of both parties warn that “the ruling would have the effect of undermining the goal of Vermont’s new campaign finance law to encourage more grassroots political activity”). See supra Part IV(a).
Freedom of association includes the right not only to engage in group activities but also to affiliate groups with one another, on a horizontal, vertical, or hierarchical basis, loosely or with centralized control. The choice is to be made by the citizens involved, not by government. See Timmons, 520 U.S. at 358, 117 S.Ct. 1364; Meyer, 486 U.S. at 424, 108 S.Ct. 1886; Buckley, 424 U.S. at 57, 96 S.Ct. 612.
Finally, no reason is given for applying expenditure limits to candidates who desire to fund their own campaigns. See generally Vt. Stat. Ann. tit. 17, § 2805a(a). Such candidates already have “access” to themselves and need not spend excessive time fund-raising. Again, speech is suppressed for no reason. See Buckley, 424 U.S. at 44-45, 96 S.Ct. 612.
b) Buckley Forecloses the Asserted Justifications for Expenditure Limits
Turning to the reasons given by my colleagues as constitutional justifications for expenditure limits, each of them has already been considered and rejected by the Supreme Court. Buckley rejected in the most explicit terms the notion that government may, under a Constitution containing the First Amendment, limit the amount of political speech by candidates and ordinary citizens. See id. It is no surprise that many of the arguments made in favor of Act 64 rehash those considered in Buckley because Act 64 was intended by its proponents as a vehicle to overturn the Buckley ruling. See 2001 Memorandum, supra; Hearing on H. 28 Before the Vt. House Comm. on Local Gov’t, 64th Biennial Sess. (1997) (statement of Anthony Pollina); Hearing on H. 28 Before the Vt. Senate Comm. on Gov’t Operations, 64th Biennial Sess. (1997) (statement of Sen. William Doyle); Vt. House Comm, of Conf., Report on Campaign Finance, H. 28, 64th Biennial Sess. (1997).
For example, the question of special donor access or the appearance thereof was highlighted both by the Congress that enacted the expenditure limitations struck down in Buckley, see Minority Views on Report of the Comm, on House Admin, to Accompany H.R. 16090 (July 30, 1974), reprinted in Legislative History of Federal Election Campaign Act Amendments of 197k, at 749 (1977), and by the Court of Appeals for the District of Columbia, which upheld those limitations and was reversed in Buckley. Buckley v. Valeo, 519 F.2d 821, 838 (D.C.Cir.1975), aff'd in part and rev’d in part, 424 U.S. 1, 96 S.Ct. 612, 46 L.Ed.2d 659 (1976). That court expressly noted that “[ljarge contributions are intended to, and do, gain access to the elected official,” id., an observation interchangeable with language in my colleagues’ opinion. To be sure, the Supreme Court did not use the word “access” in Buckley, but it did use the stronger term, “improper influence.” Buckley, 424 U.S. at 27, 45-46, 96 S.Ct. 612. Having held that corruption itself or the appearance thereof — bribes—was not sufficiently compelling to justify limits on expenditures by candidates, id. at 55, 96 S.Ct. 612, the Court hardly had to go on to say that access or the appearance of access — returning or taking a phone call from a donor — was also not compelling. Reduc*186ing bribes is generally regarded as a far more compelling interest than reducing phone calls.
It is also suggested that the practice of bundling contributions by those with common interests justifies expenditure limits. My colleagues assume that the practice of bundling was unknown at the time of Buckley. Maj. Op. at 118.23 However, the concept of pooling contributions by per*187sons with common interests is hardly new. Indeed, at the time of Buckley, proponents of the 1974 Act relied heavily on pooled contributions by various firms in particular industries as evidence of improper influence. See Senate Floor Debates on S. 3044 (Mar. 28, 1974, Apr. 3, 1974) (statements of Sens. Griffin, Baker), reprinted in Legislative History of Federal Election Campaign Act Amendments of 1974, at 259, 365-66 (1977); House Floor Debates on H.R. 16090 (Aug. 8, 1974) (statement of Rep. Dickinson), reprinted in Legislative History of Federal Election Campaign Act Amendments of 191k, at 917 (1977). An amendment in the House that would have prohibited pooling was introduced, voted on, and rejected. See Minority Views on Report of the Comm, on House Admin, to Accompany H.R. 16090 (July 30, 1974), reprinted in Legislative History of Federal Election Campaign Act Amendments of 1974, at 752-53 (1977); House Floor Debates on H.R. 16090 (Aug. 7, 1974), reprinted in Legislative History of Federal Election Campaign Act Amendments of 1974, at 858-59 (1977).
This was all explicitly before the Supreme Court in Buckley. In fact, “bundling” was considered so significant that the findings of the Buckley district court as to large contributions cataloged them by industry as well as by individual donor. See Joint Appendix at 86-143, Buckley (Nos. 75-436 and 75-437). In particular, much attention was given at the time of Buckley to the 1972 campaign contributions by the dairy industry, in which milk producers pooled a large sum and then broke it down into contributions by small committees to avoid disclosure. See Senate Floor Debates on S. 3044 (Mar. 26, 1974, Mar. 27, 1974, Mar. 28, 1974, Apr. 3, 1974, Apr. 4, 1974) (statements of Sens. Hathaway, Griffin, Hollings, Baker, Kennedy), reprinted in Legislative History of Federal Election Campaign Act Amendments of 1974, at 205, 225, 257, 365, 376-77 (1977). This incident was a highly publicized scandal at the time. It was featured in the Court of Appeals decision that upheld expenditure limits, Buckley, 519 F.2d at 839 n. 36, and that was reversed by the Supreme Court, Buckley, 424 U.S. at 55, 96 S.Ct. 612. Finally, an example of “bundling” was mentioned during the oral argument in the Supreme Court in Buckley. 1976 Landmark Briefs and Arguments of the Supreme Court of the United States: 2 Buckley v. Valeo 729 (Phillip B. Kurland and Gerhard Casper, eds. 1977) (appellees’ argument that disclosure of small contributions necessary because of “the problem of culminating, of combining contributions, if you have a large number of people who are affiliated”). The term “bundling” may be new, therefore, but the concept is long in the tooth.
My colleagues now emphasize a governmental interest mentioned only in passing in their original opinion, the belief that expenditure limits will reduce the time spent fundraising by candidates, with the hope that the extra time will be spent conferring with ordinary citizens. ,Maj. Op. at 120. This justification overlaps with the access issue and, to that extent, was decided in Buckley as discussed above.
*188In any event, the arguments regarding the time politicians spend fundraising were, as conceded by my colleagues, known and made before the Supreme Court in Buckley. Because my colleagues quote a commentator for the proposition that in Buckley, “ ‘candidate time protection was almost wholly ignored as a justification for campaign spending limits,’ ” Maj. Op. at 120 (quoting 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, 1285-86 & n.15 (1994)), I quote the pertinent passages from the briefs and lower court decision in the margin. The time-protection argument was relied upon by the Court of Appeals in Buckley in upholding the statute,24 was the subject of an entire subsection of the brief filed in the Supreme Court on behalf of the Attorney General and Solicitor General,25 was argued as a justification in the brief filed in the Supreme Court by intervening parties defending expenditure limits,26 and was *189mentioned by the Supreme Court itself.27 The claim that this issue was not fully before the Court in Buckley is therefore incorrect.
c) The Insufficiency of the Governmental Interests
Even putting aside Buckley’s binding precedent, the various interests asserted in defense of expenditure limits fail to satisfy First Amendment requirements.
1) Anti-Corruption
Act 64 drastically reduced the limits on contributions established in prior Vermont law. There is no evidence in the record that Act 64’s new low limits on contributions alone will not suffice to eliminate any improper influence. All of the — largely sparse, anecdotal, and conclusory — evidence of improper influence dates from a time when the contribution limits were much higher.
Prior Vermont law allowed contributions of $1000 per election to any candidate. See Vt. Stat. Ann. tit. 17, § 2805(a) (1996) (amended 1997). As noted, Act 64 limits contributions to candidates for statewide office to $400, for the state senate to $300, and for the state house to $200. See Vt. Stat. Ann. tit. 17, § 2805(a). The reduction is greater than might be perceived because Act 64’s contribution limits apply over a two-year cycle and related individual and party expenditures — e.g., mileage, house parties — must be counted in determining whether a donor has reached the limit. See id. § 2801(a), 2805(a), 2805a(a), 2809(a); see also supra Part III(b)-(e).
In assessing the anti-corruption effect of expenditure limits, my colleagues rely on a portrait of Vermont politics drawn from testimony by a handful of proponents of Act 64. That portrait is essentially as follows. The money spent on campaigns has been spiraling over the forty or more years in which Vermont has toyed with expenditure limits. See 1997 Vt. Laws P.A. 64 (H. 28) (finding no. 1); Maj. Op. at 101-02, 124-25. The urge to raise campaign money has increased accordingly, see id. at 124-25, leading to an “arms race” *190caused by the fear of being “vastly outspent,” id. at 127, in which, because one candidate may, for example, buy materials for yard signs, others must do so also, see Trial Tr. vol. IX, at 148 (Elizabeth Ready). We are also told that candidates for legislative office have historically spent less in their campaigns than the limits set by Act 64. See Maj. Op. at 130. In concrete terms, this means that the average campaign cost, the “arms race,” for single-member Vermont House or Senate races has over 40 years spiraled to $2,000 and $4,000 respectively. See generally Vt. Stat. Ann. tit. 17, §§ 2805a(a)(4), (a)(5).
We are also told that raising these, sums creates such a dependence among legislators on large contributors — at the time, a $1,000 maximum — that for the entire two years after each election, the legislators engage in “perpetual fundraising,” Maj. Op. at 97, and have little time to talk with ordinary citizens. See id. at 124-25. The selections from the testimony highlighted in my colleagues’ opinion portray members of the Vermont legislature as susceptible to corruption and so obsessed with soliciting or conferring with cash donors that they have very little time to confer with ordinary citizens. See Maj. Op. at 123 (describing candidates as being “locked away” while fundraising instead of “out with the public” because “candidate time is effectively for sale”).
The record justifying Act 64’s massive regulation of political speech is not strong; in fact, it is pitifully weak. Even if Vermont legislative candidates had to raise cash amounts well in excess of $4,000, this task would hardly leave them so obsessively dependent on large contributions (now a maximum over a two-year cycle of $400, $300, $200, depending on office), see Vt. Stat. Ann. tit. 17, § 2805(a), that large contributors would thereafter be able to demand the right to most of the legislators’ free time. What little actual evidence there is to the contrary is on its face gross hyperbole. For example, the testimony of a person described by my colleagues as a “lobbyist,” who offered descriptions of Governor Dean meeting only with contributors, Trial Tr. vol. IX at 195-96 (Anthony Pollina), is heavily relied upon in their opinion, Maj. Op. at 122, 123-24. In fact, that person was a lobbyist for the passage of Act 64, whose success in that regard— with the Governor’s support — entirely belies his assertions about elected officials listening only to large contributors.
There are also, of course, the accusations of corruption with precisely the same scripted sound-bites that are used in every talk-show discussion of these issues. In fact, my colleagues’ opinion overstates this testimony by omitting important qualifications offered by the witnesses relied upon, such as their lack of knowledge of any legislative vote ever cast solely because of a campaign contribution, Trial Tr. vol. VII at 48-49 (Toby Young); Trial Tr. vol. VII at 105 (Cheryl Rivers) (“I am not talking about selling votes.”), their denial of present improper influence in contrast to their fear of future behavior under the old contribution limits, Trial Tr. vol. VII at 37 (Toby Young) (opining that Vermont elections are clean); id. at 50 (opining that an official would grant preferential access to a thousand-dollar donor); Trial Tr. vol. VII at 137-38 (Cheryl Rivers) (“I don’t think the present situation is good, and I think if we don’t do something, it’s going to get worse.”); Trial Tr. vol. IX at 167-69 (Elizabeth Ready) (accepting $1000 and $2000 contributions appeared improper though she was not influenced), and their knowledge of the ready access of ordinary citizens to lawmakers, Trial Tr. vol. VII at 27-28 (Toby Young) (stating that typical *191state public officials in Vermont will see anyone who wants to see them).
The only particularized evidence of improper influence relied upon by my colleagues consists of one anecdote. It involved “widely reported” meetings of major dairy companies with unnamed officials when such meetings were denied to smaller dairy organizations, Maj. Op. at 122. We are asked to assume in this case that unspecified contributions — in contrast to, perhaps, numbers of voters involved — were the decisive factor. Moreover, not only are the details of this episode unknown, but it also took place before Act 64’s limits on contributions. It is therefore not particularly relevant.
In any event, the First Amendment does not permit the suppression of speech based on such untested anecdotal evidence. See, e.g., United States v. Playboy Entm’t Group, Inc., 529 U.S. 803, 820-21, 120 S.Ct. 1878, 146 L.Ed.2d 865 (2000) (requiring more than “anecdotal evidence” of “signal bleed” problem in support of a regulation requiring broadcasters to fully scramble sexually-oriented programming); Stanley v. Georgia, 394 U.S. 557, 567, 89 S.Ct. 1243, 22 L.Ed.2d 542 (1969) (“Given the present state of knowledge, the State may no more prohibit mere possession of obscene matter on the ground that it may lead to antisocial conduct than it may prohibit possession of chemistry books on the ground that they may lead to the manufacture of homemade spirits.”).
Moreover, if the claims of widespread, improper influence are true, anecdotes should not be the only available evidence. Disclosure of contributions has been required in Vermont for years, offering documentary support for the claims, if accurate, of dependence on large or bundled contributions and of the influence of those contributions. See, e.g., Vt. Stat. Ann. tit. 17, §§ 2811(a)(l)-(4) (effective July 1, 1982) (amended 1997) (requiring campaign reports for candidates’ contributions and expenditures). The lack of reference to available hard evidence of who gave what to whom even under the prior higher contribution limits suggests that the portrait of corruption painted by my colleagues is vastly overdrawn.
When the Supreme Court decided Buckley, it had before it detailed records of actual large, bundled contributions and the amount of out-of-pocket expenditures by candidates.28 See, e.g., Buckley, 424 U.S. at 32-34 & nn. 35-40, 96 S.Ct. 612; Joint Appendix at 264, 483, Buckley (Nos. 75-436 and 75-437); see also supra note 22. Nevertheless, it struck down the expenditure limits as facially unconstitutional. See Buckley, 424 U.S. at 54-55, 96 S.Ct. 612. Here, the opposite result is reached *192based on anecdotal evidence, even though better evidence, if corruption exists, is available.29
2) Time Protection
I turn now to the claim that expenditure limits are necessary because “endless *193fundraising[ ] drastically reduces opportunities that candidates have to meet with non-contributing citizens.” Maj. Op. at 122. As noted, this argument was squarely before the Supreme Court in Buckley. See supra notes 24-27. It was, understandably, given only passing attention by the Court because it is not compelling in any sense. The time protection argument, baldly stated, is that law can force candidates to engage in more personal communications with voters by limiting candidate spending on other means of communication.
The evidence in the record of excessive time consumption reveals that it is not a testable proposition but can be repeatedly stated as fact and in exaggerated terms. There is no doubt that candidate time is spent fundraising and that some of it is wearisome. However, there is also no doubt that one can say that candidates spend too much time fundraising without knowing how much time is actually spent, because no one else knows either. This particular record contains almost no evidence of the specific time spent fundrais-ing. To the extent that a particular amount of time was described, it was brief, such as an afternoon. Trial Tr. vol. IX at 151 (Elizabeth Ready) (“That afternoon that I had to raise that extra money, I wasn’t in front of the Grand Union nor was I going door to door.”).
Time protection is a useful political argument for a number of additional reasons. It allows legislative proponents of expenditure limits to avoid saying that limits are needed because they themselves might be tempted to vote for or against a measure solely to get a campaign contribution. Instead, they can say, without fear of any future verification, that they need more time to spend with “ordinary citizens.”
The time protection argument is also useful because it can easily be exaggerated. There is often no dividing line between donors and “ordinary citizens” who support a candidate or between fundrais-ing and other campaign events. Meetings with supportive voters can often be described either as “fundraising” or as “person-to-person contact with voters,” depending on the point to be made. The type of campaign events at which money is raised might well be held even if no contributions were sought. After all, even a holiday dinner with family members can be described as being “locked away” with large donors. See Vt. Stat. Ann. tit. 17, § 2805(f) (contributions by candidates and their immediate families unlimited).
In a more sinister vein, time protection is a particularly useful argument for incumbents. It is on its face either a remarkable example of incumbent self-sacrifice or a remarkable example of self-interest. Should we believe that incumbent legislators, who generally can raise campaign money more easily than non-incumbents, want major-party and third-party challengers to spend less time fundraising so that the challengers can spend more time engaged in supposedly more effective personal meetings with ordinary citizens? Or, is it possible that incumbents may want to stress the value of person-to-person contact over other methods of communication with voters, knowing that those other methods of communication with voters will still be substantially available to incumbents even under expenditure limits while not — or much less- — -available to potential challengers. For incumbents, the time protection argument is less about increasing person-to-person contact with voters than it is about limiting their opponents’ overall contact with voters. Time protection and incumbent protection thus usefully coincide. Finally, person-to-person contact between candidates *194and voters will not increase under an incumbent-rigged system. Potential challengers will be deterred from running, and incumbents who will be protected by the law’s expenditure limits will have less need for person-to-person contact.
Governmental interests that are so speculative — or incumbent protective — are not sufficient to override significant First Amendment interests. E.g., Stanley, 394 U.S. at 567, 89 S.Ct. 1243.
In any event, Act 64 actually diverts candidates from meeting with voters. As my colleagues concede, limiting the size of individual contributions necessarily increases the amount of time that must be spent raising a particular amount of money. Maj. Op. at 123. Moreover, expenditure limits also impose time consuming tasks on candidates themselves, particularly because expenditure limits may preclude the hiring of staff. As noted, expenditure limits require candidates to map-out, monitor, and put values on activities by supporters or party organizations that involve related expenditures, which, if they exceed $50 with regard to a single source, must be totaled within the permissible expenditure and contribution limits. See Vt. Stat. Ann. tit. 17, § 2809(a)-(e); see also supra Part 111(e). Every campaign event involving supporters — e.g. buying yard signs and driving to where they will be placed, holding meetings or dinners — must involve close calculations by candidates as to whether the particular activities constitute expenditures or related expenditures and what value should be attributed to particular in-kind expenditures. Act 64’s provisions will actually force candidates to spend more time than ever on non-speech-related activities.
3) Public Confidence in Government
This brings me to the argument of Act 64’s proponents that the expenditure limits of Act 64 will restore public confidence in government and thereby increase citizen and voter participation in elections. Of course, every attempt to suppress speech is based on claims that the speech in question, if allowed to go on freely, will induce behavior that is undesirable. Critics of literature, theater, or television with explicitly sexual or violent themes claim that such speech may induce the behavior portrayed. See, e.g., 1 American Psychological Association, Report of the American Psychological Association Commission on Violence and Youth, at 6 (1992), available at http://www.apa.org/pi/pii/violeneeand-youth.pdf (stating that exposure to violence in mass media increases the risk of youth involvement in violence). Those who would censor political speech will always argue that such speech will reduce confidence in government. I have no doubt that supporters of the Alien and Sedition Acts made such arguments and that many incumbent officeholders view vigorous opponents as undermining confidence in government.
Governmental suppression of speech must be based on a compelling demonstration that the speech will incite conduct— here an alleged indifference to politics on the part of citizens — that government has a right to prevent. See Boos v. Barry, 485 U.S. 312, 335, 108 S.Ct. 1157, 99 L.Ed.2d 333 (1988) (Brennan, J., concurring in part and concurring in the judgment) (“Our traditional analysis rejects such a priori categorical judgments based on the content of speech, requiring governments to regulate based on actual congestion, visual clutter, or violence'rather than based on predictions that speech with a certain content will induce those effects.”) (internal citations omitted); Tinker v. Des Moines Indep. Cmty. Sch. Dist., 393 U.S. 503, 508, 89 S.Ct. 733, 21 L.Ed.2d 731 (1969) (“[I]n our system, undifferentiated fear or appre*195hension of disturbance is not enough to overcome the right to freedom of expression.”)- I do not doubt that government can and ought to take steps to enhance citizen confidence, but suppressing political activity will not encourage either more confidence or more political activity.
In fact, no little part of the public confidence argument is a quintessential self-fulfilling prophesy. The confidence of Vermont citizens in their state government is unlikely to be substantially enhanced so long as Act 64’s proponents make unsupported claims about the corrupt nature of that government.
In any event, an indifference to politics cannot be traced to excessive spending for electoral purposes. To the contrary, the New Hampshire 1968 primary and the Vermont 2000 election involved heavy citizen participation because of voter interest in the issues and the critical fact that candidates who were divided on those issues could raise and spend money debating them. The theory of Act 64, as stated in the legislative findings, is that public involvement decreases as spending increases. See 1997 Vt. Laws P.A. 64 (H. 28) (findings nos. 4 and 10). However, record amounts were spent on the Vermont 2000 gubernatorial election, see Ross Sneyd, Campaign 2000 Involved Lots of Spending, Associated Press, Dec. 18, 2000, but a full — perhaps also record breaking — 34.5% more people voted in the 2000 election than in the prior gubernatorial election. See 2000 Election Results, supra.
Nor is there experience elsewhere to the contrary. Before 1976, presidential general elections were privately funded with no limits on contributions and expenditures. Claims of a lack of citizen confidence were made. From 1976 through 1988 — before the era in which so-called soft money played a growing role — presidential general elections were fully funded by government and subject to expenditure limitations. No appreciable increase in turnout or confidence in government was noted.
As in the one-size-fits all concept of an average election, there is much in the public confidence argument to fear. Proponents of Act 64 rely upon evidence such as a poll showing that 75% of voters believe that large corporations have too much influence and on a newspaper article (published by a very large corporation) stating similar conclusions. If polls suggesting that citizens believe that some groups have too much power demonstrate a governmental interest sufficient to silence those groups, political speech, including freedom of the press, cannot be protected.
Act 64 reduces the contribution limits for statewide races to $400, Senate races to $800, and House races $200. See Vt. Stat. Ann. tit. 17, § 2805(a). There is nothing in the record of this case to suggest that these limits are not sufficiently low to dispel any possibility of corruption or. the appearance of corruption, at least as viewed by reasonable persons. See Buckley, 424 U.S. at 55, 96 S.Ct. 612 (holding that “[t]he interest in alleviating the corrupting influence of large contributions is served by the Act’s contribution limitations and disclosure provisions,” and therefore does not justify campaign expenditure limitations). The proponents of Act 64 mention only hypothesized, large, cash contributions as leading to an improper influence on government. There is no evidence whatsoever that expenditures by supporters for “meet the candidate” events, or that supporters’ use of a residence, computer or phone, purchase of stamps, ,or driving to meetings have ever caused a problem that calls for redress.
Moreover, there is nothing in the record to suggest that disclosure of amounts and sources of a candidate’s campaign funds, in conjunction with low contribution limits *196and/or a form of public financing, is not the proper democratic method of enhancing voters’ confidence in the character of the people they elect. Nor is there anything in the record to suggest that a reduction of campaign activity, in particular grassroots activity, will lead to persons of better character being elected, particularly under the terms of a law enacted by those of purported lesser character.
The theory of Act 64 is that less political advocacy is better for us as a polity because too much political activity is engaged in by powerful groups. Because these groups are theoretically able to use every means of communication as a conduit of influence, political activity at every level must be reduced. Act 64 is, therefore, designed to impose relative silence on everyone, with two exceptions. First, incumbents will ensure that they can communicate with the public. Second, the truly rich and powerful can still engage in constitutionally protected independent political activities or buy a media outlet. Expenditure limits do not limit the influence of a Richard Mellon Scaife, a George So-ros, or the politically concerned persons that publish the Washington Post, New York Times, National Review, and New Republic. Candidate expenditure limits in fact enhance the power of these wealthy individuals to set the political agenda while the ordinary citizen — who must speak, if at all, through organizational activity' — -is silenced.
d) Stopping the “Arms Race, ” “Effective Advocacy,” and Incumbent Protection
As noted, the term “arms race” has been much used in this litigation, see, e.g., Appellants’ Br. at 26-27; Maj. Op. at 117, 119, 122, 123, 127, 132, 134; Trial Tr. vol. VIII at 57 (Peter Smith); Trial Tr. vol. VII at 56 (Cheryl Rivers), but in a way that suggests that it is so obviously an appropriate analogy that an explanation of its relevance is unnecessary. When examined, however, the analogy has no logical or factual support. It is also antithetical to the First Amendment because it suggests that government may set a low maximum limit on political speech and, to boot, one that is particularly harmful to challengers.
1) The “Arms Race”
The “arms race” analogy is a useful— and therefore oft-used — political slogan for proponents of Act 64 because it suggests subliminally a catastrophic spectre of millions being killed and perhaps elimination of the species itself. However, the analogy between nuclear-tipped ICBM’s — which many people would want to eliminate completely- — and political advertisements — in Vermont, yard signs — is not one that meets the straight-face test, much less one that fits well within First Amendment jurisprudence, and this aspect of the analogy is unworthy of further discussion.
What other lessons the slogan “arms race” is deemed to further are difficult to detect because it is used as an argument-stopper rather than argument-advancer. It may suggest that much of political spending is superfluous — beyond a certain point additional spending does not change votes. This is a suggestion that posits candidates who keep spending even though it will not benefit them. If so, the empirical basis for that suggestion is not visible in this record. Nor is there a visible basis for believing that the particular point at which further persuasion stops is the low expenditure limits imposed by Act 64.
Perhaps the implication is that spending becomes superfluous at some point but candidates have no idea where that point is and continue to spend anyway. However, if the critical point cannot be determined *197by a candidate in a particular campaign, it certainly cannot be determined on a one-size-fits-all basis by a self-interested legislature or by a reviewing court.
My colleagues apparently do not use “arms race” to imply superfluous spending. Rather, they perceive the-“race” to result from the fear of being “vastly outspent” by better financed opponents. Maj. Op. at 127. They do not deny that on election night, election officials do not count the dollars spent by a candidate to determine the winner, and thus, that a fear of being outspent necessarily associates spending with voter persuasion. See id. at 123 (quoting Act 64 supporter for proposition that contribution limits without expenditure limits would lead to competitions “to see who could raise the most money and outspend their opponent and therefore win the race”) (emphasis added). “Arms race” is therefore a pejorative method of describing competition over voter persuasion, the very heart of the democratic process.
The proponents of Act 64, like my colleagues, also use the slogan “arms race” to argue that any such race should be prevented because, in their view, (supposedly) low cost, person-to-person contacts are the most effective campaign tactics. Trial Tr. vol. IX at 143, 150 (Elizabeth Ready) (personal contact more effective than paid media); Trial Tr. vol. VII at 100 (Cheryl Rivers) (challenger spending more than incumbent is not serious detriment to incumbent, who “can make it up with grass roots effort”). Therefore, campaign expenditures above a certain level (somehow determined) can be eliminated without disadvantaging any candidate.
What is not explained, however, is why, if spending above that somehow determined level is ineffective, such spending by a candidate is feared by the candidate’s opponent. The answer must be, and is, that the additional spending does persuade voters. As the incumbent Senator, whose testimony is relied upon by my colleagues, testified, she had to spend more than she wished because her opponents’ ads and yard signs caused voters to wonder whether she was running for reelection. Trial Tr. vol. IX at 148 (Elizabeth Ready) (“[Wjhen everybody has yard signs out and everybody’s on the radio and the TV, your constituents will say, Aren’t you running Elizabeth? I see that so and so has got a million yard signs out. You don’t have any yard signs.”) Her opponents, in short, had gotten the attention of voters; who appear to have been more oblivious to her person-to-person campaign than she would have liked. In fact, as used in the present record, “arms race” is a term by which incumbents describe contested elections.
2) “Effective Advocacy”
I turn now to the test adopted by my colleagues to determine whether the level of expenditure limits set by Act 64 is unconstitutional. That test asks whether the limit is so low that it prevents “effective advocacy” by “driv[ing] the sound of a candidate’s voice below the level of notice.” Maj. Op. at 128-29 (internal citation omitted). Two aspects of this test must be emphasized. First, it is a minimum speech test, expressly authorizing government to silence candidates once they reach “the level of notice” (assuming no less restrictive methods are available). Second, my colleagues, by equating (understated) average past expenditures with,the threshold “level of notice,” id., further reduce the minimum at which government can silence candidates, see supra Part IV(b)(l).
Despite the unconditional statements by the Supreme Court that contribution limits “ ‘entai[l] only a marginal restriction upon the contributor’s ability to engage in free communication,”’ McConnell, 124 S.Ct. at *198655 (quoting Buckley, 424 U.S. at 20, 96 S.Ct. 612), while “limitations on expenditures [are] direct restraints on speech,” id. at 647, my colleagues take the “effective advocacy”/“level of notice” standard from a discussion of contribution limits in Shrink, 528 U.S. at 395-96, 120 S.Ct. 897, and force it into the quite different context of expenditure limits.
Contributions pose the spectre of improper influence and may be substantially limited. Colorado II, 533 U.S. at 440-41, 121 S.Ct. 2351 (“limits on contributions are more clearly justified by a link to political corruption than limits on other kinds of political spending are”). The language taken by my colleagues from Shrink says no more than that, even given the governmental interest in reducing improper influence, contribution limits may not be set so low as to prevent an otherwise viable candidate with large numbers of potential small donors from raising enough money even to be noticed by voters. 528 U.S. at 395-96,120 S.Ct. 897. It does not say that such a candidate may, having raised a substantial amount in small contributions, be prevented from spending more than what (government believes) is needed to reach the minimum level of notice. Indeed, Shrink itself repeatedly and conspicuously emphasized the constitutional distinction between contribution limits and expenditure limits. 528 U.S. at 386, 120 S.Ct. 897 (“expenditure limits [are] direct restraints on speech, which nonetheless sufferf ] little direct effect from contribution limits”); id. at 387, 120 S.Ct. 897 (noting a “similar difference between expenditure and contribution limitations in their impacts on the association right”); id. (“restrictions on contributions require less compelling justification than restrictions on independent spending”) (internal citation omitted). This distinction was recently reaffirmed by the Court in McConnell, 124 S.Ct. at 655 (reaffirming practice of “subjecting] restrictions on campaign expenditures to closer scrutiny than limits on campaign contributions” because “contribution limits, unlike limits on expenditures, entail only a marginal restriction on the contributor’s ability to engage in free communication”).
Unlike expenditure limits, which directly restrain speech, Buckley made clear that “[t]he quantity of communication by the contributor does not increase perceptibly with the size of the contribution, since the expression rests solely on the undifferentiated, symbolic act of contributing.” 424 U.S. at 21, 96 S.Ct. 612. “A limitation on the amount of money a person may give to a candidate or campaign organization thus involves little direct restraint on his political communication ... [because] the transformation of contributions into political debate involves speech by someone other than the contributor.” Id.
It is from this premise — as long as the candidate can speak effectively, the contributor also can speak — that the effective advocacy test was born. See also McConnell, 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 prevent candidates and political committees from amassing the resources necessary for effective advocacy.”) (internal quotation marks and alteration omitted). In no way do the cases evaluating contribution limits use the effective advocacy test to restrain the direct political speech occasioned by expenditures.
Were these cases read otherwise, they would represent a complete abandonment of the First Amendment’s standard of a free, robust discussion in which citizens “retain control over the quantity and range *199of debate on public issues in a political campaign,” even when those who control the government believe the spending to be “wasteful, excessive, or unwise.” Buckley, 424 U.S. at 57, 96 S.Ct. 612. That test, which emphasizes freedom, would be replaced by a test that permits government to cap the amount of permissible speech.
3) Incumbent Protection
The rhetoric of the “effective advocacy”/“level of notice” standard based on average past expenditures conceals a legal test lethal to challengers. A “level of notice” is something that incumbents generally have and challengers generally lack. Under my colleagues’ test, therefore, an incumbent’s campaign starts at the “level of notice” at which a challenger’s campaign may be stopped by government. This anti-challenger effect is aggravated by the use of average past expenditures to determine the “level of notice.” See Maj. Op. at 129-31. As detailed above, see supra Part rV(b)(l), past averages, even if accurately calculated — a result not attainable given Act 64’s new definitions — include uncontested, or barely contested, elections. See Trial Exs. vol. Ill at E-0967 (appellees’ expert’s calculation of average expenditures, which includes low-spending candidates whose spending is unknown by assuming they spent $500, the maximum allowed before filing is required); id. at E-1019 (appellees’ expert’s report criticizing appellants’ expert for failing to include low-spending candidates, whose spending is unknown, in his averages of campaign expenditures).
e) The Excessive Discretion Accorded Administrators
In their first opinion, my colleagues, noting that “[i]t is beyond cavil that an opponent of the Act will argue its ambiguities and statutory peculiarities,” addressed the merits of that argument and stated that the discretion accorded administrators by Act 64 was not constitutionally excessive. Landell v. Sorrell, 2002 WL 1803685, slip op. at 9156 (withdrawn). They now argue that the issue need not be addressed. Maj. Op. at 136-37, n.26.1 disagree.
The discretion issue is clearly before us. The degree to which Act 64 limits political activity is the first part of the calculus that is before the court. The second part of the calculus is the sufficiency of the reasons proffered in its support. Where only acts of administrative or judicial discretion can mitigate the harshness of restrictions on protected activity so as to render the justifications for the restrictions constitutionally sufficient, the constitutionality of that discretion itself is obviously put in issue.
Mild or incidental restrictions on political activity require less compelling justifications than do harsh restrictions. FEC v. Beaumont, 539 U.S. 146, 123 S.Ct. 2200, 2210, 156 L.Ed.2d 179 (2003) (level of scrutiny applied to “political financial restrictions” is “based on the importance of the political activity at issue to effective speech or political association”) (internal citation omitted). Much of what Act 64 says harshly limits political activity; much else is left to future elaboration. Perhaps we may rely upon the wisdom of Vermont’s Secretaries of State, Attorneys General, and its courts, to mitigate the harsh effects of Act 64’s language and to resolve its pervasive ambiguities in favor of freedom, rather than suppression, of political speech. If so, the reasons offered in support of the Act might seem sufficient. However, mitigating rules — e.g. reducing the effect on the press or allowing local party affiliates self-financing — would involve wholly discretionary or arbitrary decisions, and the very existence of that dis*200cretion is itself a constitutional problem that cannot be avoided.
Limits on campaign expenditures are like all limits on speech. If the limits are triggered, further speech is forbidden. It is standard First Amendment jurisprudence that such a restriction on speech must be precisely crafted to avoid vesting those who administer the law with excessive discretion as to its interpretation. See Forsyth, 505 U.S. at 131, 112 S.Ct. 2395 (requiring “narrow, objective, and definite standards”). The requirement that a law regulating speech embody workable and known standards is necessary both to alert those who are regulated to its terms, see Gentile v. State Bar of Nevada, 501 U.S. 1030, 1048, 111 S.Ct. 2720, 115 L.Ed.2d 888 (1991) (requiring regulation of speech to give “fair notice” to those to whom it is directed), and to prevent enforcers from making decisions based on impermissible grounds, see id. at 1050-51, 111 S.Ct. 2720 (“The prohibition against vague regulations of speech is based in part on the need to eliminate the impermissible risk of discriminatory enforcement.”); Forsyth, 505 U.S. at 131, 112 S.Ct: 2395 (noting “danger of censorship” where regulation allows excessive enforcement discretion).
Act 64 simply lacks discernible criteria for the many interpretive and valuation questions that it creates. If there is to be compliance with Act 64 — instead of candidates and their supporters generally ignoring it as a silly law — there must be constant interpretation by the Secretary of State, the Attorney General, and the Vermont courts with regard to the vast number of questions that will arise election-by-election, campaign-by-campaign, and day-by-day. The answers to those questions are the equivalent of granting or denying a permit to speak. In interpreting the statute, however, the Secretary of State and the Vermont courts are afforded almost no guidance except for the manipulable proposition that the influence of “special interests” is to be reduced and that of “ordinary citizens” increased.
For example, the statute says nothing about the payment of debts or wind-down expenses of prior campaigns during the next two-year cycle. See 1999 Memorandum, supra. It is also unclear whether the (paltry) exception for expenses for “meet the candidate” events applies only to party-sponsored events or all such affairs. See supra notes 14-15; see generally Vt. Stat. Ann. tit. 17, §§ 2809(d)(l)-(3).
If Act 64 is enforced, valuation questions regarding the donation or use “of anything of value” will themselves be a constant issue. When the Act was passed, the Secretary of State set the cost of mileage at 31$. Although travel on behalf of a candidate’s campaign is a related expenditure counting toward the contribution limit, see Vt. Stat. Ann. tit. 17, § 2809(c), she has arbitrarily not changed that figure notwithstanding the substantially increased price of gas and the pervasive availability of public and private mileage guidelines. In fact, employees of the State of Vermont are compensated at 37 ’At per mile at present. Vermont Dept, of Personnel, Collective Bargaining Agreements, at http://www.vermontpersonnel.org/employ-ee/labor_cba.cfm (effective July 1, 2003 to June 30, 2005) (setting mileage reimbursement for Vermont employees at level established by the U.S. General Services Administration, currently 37^$). Given the two-year cycle and the low contribution/related expenditure limits, mileage valuations are of enormous importance, but those valuations appear essentially to be matters of caprice under Act 64.
There will also be ubiquitous questions concerning whether particular activities of officeholders, “candidates,” or would-be “candidates” have “the purpose of influenc*201ing an election.” See generally Vt. Stat. Ann. tit. 17, § 2801. Indeed, the Supreme Court stated in Buckley that the last-quoted phrase was unconstitutionally vague unless more narrowly confined than it is in Act 64. See supra note 6. There is also little guidance as to what conduct is an “affirmative action to become a candidate,” see supra notes 9-10, or what professional services are donations or related expenditures by a firm rather than volunteer services. See id. §§ 2801(1), 2809.
Furthermore, the definition of “related expenditures” can provoke thousands of questions regarding actions of individuals or political parties as to which the answer turns — after a potentially intrusive inquiry into the fine details of what candidates and political parties want to do or did, what they said, and what they thought — on what was the “primary thrust” of the activity. See Appendix A. In addressing such questions, the Secretary herself has noted that the likelihood of so many different factual circumstances arising prevents the drafting of precise rules regarding whether particular efforts by a party will be related expenditures on behalf of candidates — one of the most important questions arising under Act 64. See id.
All of these issues are serious, bristling with First Amendment implications, and their resolution will oft-times award an election to one candidate rather than another. If a party’s poll is deemed a related expenditure on behalf of a candidate for the House, most or all of that candidate’s expenditure limits for two elections may be exhausted. If a particular activity by an incumbent legislator is deemed an expenditure, rather than the performance of an official duty, that legislator may be barred from driving the family automobile to the local town green to make a speech during the campaign. If the activity is not an expenditure, the incumbent legislator may be allowed to engage freely in very helpful electoral activities that are denied to his or her opponent. If a candidate has a supporter who is a lawyer and whose professional services are not deemed related expenditures, that candidate will have a great advantage over another whose supporters are not lawyers, including the ability to bring litigation against the opponent that will exhaust the opponent’s campaign funds. A ruling allowing lawyers to contribute professional services without counting such services as contributions or related expenditures is hardly out of the question under Act 64, even though lawyers are not less apt than other citizens to seek favors from elected officials for themselves or their clients.
The Secretary of State’s opinions allowing partners to make double donations— once by the partnership, once by the individual partners — and endorsing the legality of “pass-the-hat” fundraisers in which donors are allowed to remain anonymous and on the honor system as to how much they give are only the first examples of how Act 64 will come to mean what the Secretary of State and Vermont courts say it means. See 2001 Guide, supra. Mileage is now computed at 81<t no matter what the price of gas. See 2001 Memorandum, supra. Such uncabined discretion cannot be squared with the First Amendment’s requirements that speech be regulated according to spelled out and precise criteria.
Equally important, such discretion cannot be squared with increasing confidence in government, which, in enforcing Act 64 against those running for government office, will necessarily appear inefficient and arbitrary. Only an organ of government can administer and interpret these laws. However, that interpretation consumes time when time is of the essence30 — Ap*202pendix A. is a letter dated December 3, 1999, responding to an inquiry dated October 8 — and is susceptible to colorable claims of partisan influence.
In Vermont, the purported author of Act 64 was denied public financing because his party took a poll that, if attributed to his candidacy, would be a related expenditure causing him to exceed the maximum contribution and expenditure limits for candidates eligible for public financing. See Ross Sneyd, Progressives’ Poll Raises Question About Public Financing, Associated Press, Feb. 21, 2002 (describing Anthony Pollina’s violation of the campaign finance law). The Democratic party then objected to his receipt of public financing. See Ross Sneyd, Democrats Ask that Pollina Be Disqualified from Public Financing, Associated Press, Feb. 28, 2002. The Secretary of State and the Attorney General, both Democrats, undertook an investigation into the activities of the candidate’s party. See Ross Sneyd, Progressives Sue to Ensure Public Financing for Pollina, Associated Press, Mar. 12, 2002; see also Ross Sneyd, Polli-na’s Lawyer Says He Won’t Cooperate with AG’s Probe, Associated Press, Mar. 22, 2002. The candidate was then quoted as saying, quite understandably, “You have the Democratic Party asking the Democratic Attorney General based on an opinion of a Democratic secretary of state to investigate a Progressive Party candidate.” Christopher Graff, Anthony Polli-na’s Campaign Demeans Legislators, Associated Press, Mar. 17, -2002. A system in which partisan politicians investigate and make rulings on how vigorous a campaign their opponents may wage is not a confidence-builder.
VI. THE REMAND ON NARROW TAILORING
My concerns over the remand to the district court for various “findings” are fourfold. First, my colleagues’ opinion draws no distinction between legislative facts, mixed questions of legislative fact and law, adjudicative facts, and issues of law — on this record distinctions of crucial importance to any further proceedings in the district court and in this court. Second, it is unclear what parts of Act 64 my colleagues deem to be restrictive, and to what degree, thereby hampering if not precluding any comparison with alternatives. Third, vastly less restrictive alternatives with no constitutional implications are so obvious that a remand is unnecessary. Fourth, on many issues, the forum that needs to be heard from is not the district court but the Vermont legislature.
a) Legislative Facts, . Adjudicative Facts, and Mixed Issues of Fact and Law
One of the difficulties I had with my colleagues’ earlier opinion — but did not elaborate in my earlier dissent — was that it did not make clear which facts are of a legislative nature — facts that determine the appropriateness of a rule of law — and which facts are of an adjudicative nature— facts that affect the legal relations of the particular parties to a particular lawsuit. See Fed.R.Evid. 201, Notes of Advisory Committee on Rules (explaining the “fundamental differences- between adjudicative facts and legislative facts. Adjudicative *203facts are simply the facts of the particular case. Legislative facts, on the other hand, are those which have relevance to legal reasoning and the lawmaking process, whether in the formulation of a legal principle or ruling by a judge or court or in the enactment of a legislative body.”); see also Langevin v. Chenango Court, Inc., 447 F.2d 296, 300 (2d Cir.1971) (“Adjudicative facts” are “facts about the parties and their activities, businesses, and properties, as distinguished from general facts which help the tribunal decide questions of law and policy and discretion.”) (Friendly, C.J.) (internal quotation marks and citation omitted). Nor did my colleagues’ earlier opinion separate factual matters from mixed issues of legislative fact and law. The remand for “fact finding” has now pushed these issues to the forefront.
1) The Distinction Between Legislative and Adjudicative Facts
I recognize that the distinction between legislative and adjudicative facts is no bright line and is often judicially finessed. See Kenneth Culp Davis, An Approach to Problems of Evidence in the Administrative Process, 55 Harv. L. Rev. 364, 403 (1942) (“courts have generally treated legislative facts differently from adjudicative facts, even though the distinction has not been clearly articulated”). In the present case, however, the perceived need for a remand appears to be based on the conviction that adjudicative facts need to be resolved. In my view, that is not so.
Legislative facts are factual assumptions or conclusions that cause a court to choose one rule of law rather than another or to hold that certain circumstances meet a particular legal test. See Fed.R.Evid. 201, Notes of Advisory Committee on Rules. Legislative facts thus govern all future cases impheating the particular rule of law or its application and are not subject to future challenge by a litigant save by an attempt to have the rule of law overruled. My colleagues’ repeated reliance on the existence of an “arms race,” their adoption of the “level of notice”/“effective advocacy” test defined by average past expenditures, and their conclusion that Act 64’s expenditure limits meet that test, all rest on factual assumptions about candidate spending. These assumptions are quintessential legislative facts.
Determination of legislative facts is not governed by the Federal Rules of Evidence. See Fed.R.Evid. 201, Notes of Advisory Committee on Rules. Nor are they subject to clearly erroneous review under Fed.R.Civ.P. 52. See In re Asbestos Litigation, 829 F.2d 1233, 1252 n. 11 (3d Cir.1987) (citing Lockhart v. McCree, 476 U.S. 162, 106 S.Ct. 1758, 90 L.Ed.2d 137 (1986)). Instead, they are subject to de novo review, and appellate courts not only can find legislative facts on their own but they also usually do so. See Davis, supra, at 403-07 (describing Supreme Court and other cases in which appellate courts found legislative facts). The practice is so common that what technically might be called mixed issues of legislative fact and law are often treated simply as issues of law. Whether expenditure limits set at the level of average past spending are sufficient for “effective advocacy,” for example, is just such an issue.
Adjudicative facts determine the legal relations of particular parties with regard to particular issues of controversy. Adjudicative facts do not govern the results of future litigation — save for the application of doctrines of preclusion such as res judi-cata and collateral estoppel. Determination of these facts is governed by the Federal Rules of Evidence and may be based on credibility determinations. Significantly, adjudicative facts are subject to clearly erroneous review on appeal under *204Rule 52. Fed.R.Civ.P. 52(a). Examples of adjudicative facts are whether one party-struck another by driving an auto through a red light.
2) Remand for “Findings” of Legislative Fact and of Law
In my view, my colleagues fail to observe the adjudicative/legislative distinction because they view legislative facts relating to the constitutionality of Act 64 as the sole province of the district court, while an appellate court may address only the legal issues arising from these facts. As their opinion expressly states, “[A]l-though we do not question the validity of the factual findings developed by the legislature in support of Act 64, 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.” Maj. Op. at 114 (footnote omitted). As a result, the remand in the present case seeks findings on legislative facts, mixed issues of legislative fact and law, and even pure questions of law.
My colleagues remand for findings on whether higher limits would achieve the anti-corruption and anti-time consumption goals and “impinge less on the First Amendment rights of candidates and voters.” Id. at 135-36. This inquiry is described by them as a “fact intensive question of whether that point is set in Act 64 or appreciably higher” and apparently is viewed as an adjudicative fact as to which the district court may make credibility determinations that are binding on this court unless clearly erroneous. See id. at 136; Fed.R.Civ.P. 52(a). However, the issues are clearly ones of mixed legislative fact and law that can and should be decided by this court.
It is undisputed that: (i) the view that Act 64’s limits will not substantially affect candidate spending is arrived at only by averaging in spending in essentially non-contested elections and using a definition of spending much narrower — one that excludes all related expenditures by individuals and political parties — than that used by the Act; and (ii) even under the narrow definition, the limits are far below spending in actual contested elections in Vermont, including spending by third party candidates. All that is left is the legal question of the sufficiency of the governmental interest in justifying the restrictions on speech.
My colleagues also remand for what they describe as “fact-finding on [the] issues” of
(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.
Maj. Op. at 136 (footnote omitted). None of these issues involves adjudicative facts. Issues (3) and (4) are clearly questions of mixed legislative fact and law, if not solely of law, while issues (1) and (2) involve determinations of legislative history.
For yet another example, footnote 23 of my colleagues’ opinion indicates that “findings” are to be made on remand as to a pure matter of law: whether required reports of spending by candidates for past elections included “related expenditures” as defined in Act 64, ie. including spending by individuals and political parties. Id. *205at 135 n. 23. This is an issue of Vermont law over which there is no dispute. Prior Vermont law never mentioned “related expenditures”; the term was first introduced and defined in Act 64; before that there were contribution limits on individuals, and political parties could give and spend freely; there are, therefore, no “facts” to find.
In the same footnote my colleagues also ask for findings with regard to whether “legal and record-keeping costs of compliance with [Act 64] will also inflate future candidates’ expenditures.” Id. My colleagues appear to concede that such costs are expenditures and are therefore limited by Act 64. What more need be known? There are no adjudicative facts at issue here. The Secretary of State has advised candidates to obtain legal advice, and any legal or record keeping costs will drastically affect, say, a House candidate who must wage both a primary and general election contest with a total two-year expenditure limit of $2,000. The trees in Vermont may be beautiful, but needed professional services do not grow on them.
My colleagues now concede that this court will apply de novo review to legislative facts found by the district court. Id. at 135 n. 24. However, my colleagues make no attempt to inform the parties and the district court of what findings are to be of legislative fact and what are to be of adjudicative fact. For example, my colleagues remand for “fact-finding” on the types and amounts of limits rejected by the Vermont legislature and the reasons for rejection. Id. at 136. I see no comparative advantage in the district court’s researching this question of legislative history unless it is contemplated — and I assume it is not— that the district court will take testimony on the state of mind of the then-legislators, resolve credibility issues, and find facts on these issues.
My colleagues do agree that a distinction between legislative and adjudicative facts exists and that legislative facts are reviewed de novo while adjudicative facts are reviewed under the “clearly erroneous” standard. However, they never identify which of the many “findings” to be made on remand are legislative and which are adjudicative, and, therefore, what rules to apply to each. This failure will greatly complicate further proceedings.
In my view, virtually all the issues remanded are ones of legislative fact or of law, and, therefore, there is no reason for a remand. Moreover, if my colleagues believe that further findings of legislative fact are needed, they can request briefing of the relevant issues by the parties, rather than returning questions of law to the district court only to have us later resolve them de novo.
b) Failure to Define What is Restrictive About Act 6k
Clarification is also needed with regard to the inquiry on remand as to whether there are less restrictive alternatives available. As I said at the outset of this dissent, my colleagues still fail to analyze much of what Act 64 actually says and does, and, as a result, their opinion contains few descriptions of significant political speech that they deem to be restrained by Act 64. An inquiry into less restrictive alternatives requires some identification and discussion of those restrictions for which an alternative might be substituted. That is to say, an available legal rule may be deemed less restrictive only after its restrictions are compared with the restrictions of an existing legal rule. Colorado II, 533 U.S. at 464-65, 121 S.Ct. 2351 (comparing restrictiveness of existing limitation on coordinated expenditures with available legal rule limiting contributions alone). However, my colleagues’ opinion *206neither acknowledges nor denies that Act 64: intrudes on grassroots activities of ordinary citizens, restricts the press’ editorializing or reporting on political events, disadvantages candidates who must run in two elections rather than one, and so on and so on.
The lacunae in my colleagues’ opinion are nowhere better exemplified than by the remand with regard to related expenditures. Maj. Op. at 137. The two sentences directing this remand precede a discussion upholding Act 64’s treatment of related expenditures as contributions. That discussion notes that limits on related expenditures are designed to avoid evasions of the limits on contributions. Id. at 145-46. For example, if someone can both lend office space to a candidate and make a cash contribution at the maximum limit, that person can evade the limits.
Before this discussion, their opinion states “On remand, independent of the constitutionality of expenditure limits, the district court should evaluate [the constitutionality of treating related expenditures as candidate expenditures].” Id. at 137. Of course, one reason the related expenditure provision was included in Act 64 was that cash or in-kind expenditures by individuals coordinated with a campaign are an obvious method of evading candidate expenditure limits as well as contribution limits. To return to the prior example, if a person lends office space to a candidate, the cost must be treated as an expenditure subject to the limits on campaign expenditures, or those limits can be evaded. Because Act 64 reflects the view that expenditure limits cannot be effective without treating individual and party related expenditures as expenditures by the candidates, it is enigmatic, to say the least, to ask the district court to evaluate “this issue ... independent of the constitutionality of expenditure limits.” My colleagues appear to be troubled by Act 64’s limits on related expenditures but fail to describe or discuss the reasons for their disquiet. Such a description or discussion would surely be helpful to the district court and the parties.
c) The Existence of a Less Restrictive Alternative
I turn now to the question of whether, if Act 64 contains the restrictions described in this dissent, there are less restrictive alternatives. My colleagues frame this issue as an inquiry into what alternatives were actually considered by the Vermont legislature. See Maj. Op. at 135. That is not the proper inquiry.31 A state may not impose laws suppressing political speech and then successfully defend them on the ground that it was ignorant of alternatives. Rather, the issue is the existence of less restrictive alternatives, not whether a par*207ticular legislature considered them. See Boos v. Barry, 485 U.S. 312, 329, 108 S.Ct. 1157, 99 L.Ed.2d 333 (1988); Wygant v. Jackson Bd. of Education, 476 U.S. 267, 280 n. 6, 106 S.Ct. 1842, 90 L.Ed.2d 260 (noting that the term “narrowly tailored” “require[s] consideration of whether lawful alternative and less restrictive means could have been used”).
Moreover, when properly framed, the answer to the inquiry is so self-evident in the present case that a remand is unnecessary: a combination of public and private financing with low contribution limits is infinitely less restrictive — is actually speech supportive — and accomplishes all of the ostensible purposes of Act 64’s expenditure limits. Public financing of campaigns is of unquestioned validity, see Buckley, 424 U.S. at 57, n. 65, 85-109, 96 S.Ct. 612, and, combined with low limits on private contributions, can guarantee a critical mass of funds to all candidates, thereby freeing candidates of improper influence from particular donors and relieving candidates of the need for extensive fund-raising. Such a combination of public and private financing with low contribution limits would impose only a modest burden on taxpayers, who, we are told, are anxious to eliminate undue influence by donors.
My colleagues suggest in a footnote, see Maj. Op. at 136 n.25, that the conclusion that a combination of public and private financing would satisfy the goals of Act 64 is neither self-evident nor supported by the record. To the contrary, if any conclusion is established, it is this one. As my colleagues’ opinion repeatedly states, low contribution limits further the anti-corruption goal. And, as the Vermont legislature has stated, public financing (provided in gubernatorial races) furthers both the anti-corruption and time protection goals. 1997 Vt. Laws P.A. 64 (H. 28) (findings nos. 11 and 12). Public funding money does not come from private sources, is quickly obtained, and lessens the need for private money because it is as negotiable.
The problem with this speech-supportive alternative is not its dubious merit but the fact that incumbent legislators know that a combination of public and private financing would vastly increase the number and viability of challengers. The dark secret of campaign finance reform is that its proponents avoid this alternative in a Faustian bargain with incumbent legislators who reject it out of self-interest. Incumbents prefer speech-limiting expenditure limits, a preference that will gain in intensity if federal courts will allow such limits to be based on past average spending and to be upheld so long as they do not drive challenger campaigns below “the level of notice.”
d) Remanding to the Wrong Forum
As I have argued above, even if expenditure limits were not per se illegal, the limits set by Act 64 are so ridiculously low that they fail under any reasonable standard. In fact, they are so low that they would diminish spending by third party candidates, not generally regarded as generators of “arms races,” and would limit even modest grassroots activities by supporters. Similarly, the restrictions on spending by local party affiliates fail because no reason has been offered to justify them. Holding the expenditure limits and the restrictions on the financial organization of political parties unconstitutional would send these issues back for reconsideration to the proper forum — the Vermont legislature.
My colleagues show great deference to the Vermont legislature and to the various legislative proponents of Act 64 whose views are in the record. This deference is entirely undeserved. There is not the *208slightest evidence that either the legislature or those proponents ever examined the details of the Act or how they impact those ordinary political activities that are indispensable to democratic rule. Consequently, the legislators never weighed Act 64’s costs in suppressed activity. There is no evidence of any discussion of, inter alia, the effects on grassroots activity, the two-year cycle, the effect on the press, the costs of legal services to candidates, or the radical restructuring of political parties.
In fact, their own testimony suggesting that “meet and greet events” such as “spaghetti suppers,” “little parties” for “150 people” to which “a couple hundred people” are invited by mail, booths at county fairs, barbecues, and op-ed articles in the press are campaigning methods not involving Act 64’s expenditure limits, see supra Part 11(b), itself reflects an awesome ignorance of Act 64’s provisions. Moreover, it is certainly hard to find any explanation other than ignorance of the contents of Act 64 for the fact that the executive director of the Vermont Democratic party challenged, as harmful to grassroots activities, the ruling by the Secretary of State that Act 64’s limits on contributions to parties treat all party affiliates as a single unit, see Secretary of State Being Criticized for Fund Raising Ruling, Associated Press, May 28, 1999, even though the ruling simply followed the plain language of the Act.
The lobbyist who secured passage of Act 64, who has been described as its author, and who is quoted at great length in my colleagues’ opinion, has since brought an action in the district court to have the treatment of related expenditures declared unconstitutional as infringing on the (his) right to engage in political activities. See Veivnont Reformer Says Law He Authored Is Unconstitutional, Political Finance, The Newsletter, March, 2002 (describing Anthony Pollina’s lawsuit to have Act 64 ruled unconstitutional); Ross Sneyd, Progressives Sue To Ensure Public Financing for Pollina, Associated Press, March 12, 2002 (noting that Anthony Polli-na calls his lawsuit “ironic”); see also Complaint at 1, Pollina v. Markowitz, No. 2:02-CV-63 (D.Vt. Mar. 11, 2002) (“Plaintiffs claim that certain provisions of Act 64 violate their First Amendment free speech and association rights, do not serve compelling state interests, and violate equal protection and due process of law, both facially and as applied.”).
Officeholders have filed disclosure forms that indicate a continuing lack of knowledge of the requirements of Act 64, in particular the need to record and disclose mileage and the value of office space and outside professional services. See, e.g., Campaign Finance Report of William Doyle, Dec. 16, 2002 (listing no mileage expenses for himself or any supporters, or any value derived from use of office space, computers, utilities, etc., or any outside legal or accounting services); Campaign Finance Report of William Doyle, Sept. 25, 2002 (same); Campaign Finance Report of William Doyle, Oct. 25, 2000 (same); Campaign Finance Report of Anthony Pollina, Dec. 16, 2002 (listing no mileage expenses for any supporters or any value derived from use of office space, computers, utilities, etc., or any outside legal or accounting services); Campaign Finance Report of Jeb Spaulding, Dec. 15, 2002 (same); Campaign Finance Report of Deborah Markowitz, Dec. 16, 2002 (listing no expenses for value of office space, computers, utilities, basic office supplies, etc., or any outside legal or accounting services); Amended Campaign Finance Report of Deborah Markowitz, Oct. 29, 2002 (same); Campaign Finance Report of Deborah Markowitz, Dec. 18, 2000 (same). The actual spending by other proponents of Act 64 also belies their opinions as to the reasonableness of Act 64’s limits. See *209supra, Part IV(b)(l)(C) (describing Act 64 proponents who exceeded its limits in past elections).32
Deference to a legislative judgment is due when some minimal effort has been made by legislators to weigh the relevant factors and strike a minimally informed balance. There is no evidence of any such legislative effort with regard to Act 64’s actual effect on political activity.
Even under my colleagues’ legal theory, therefore, the Act’s limits on expenditures must be struck down as well below any reasonable constitutional standard, and the limits on local party affiliates must be invalidated for the lack of any proffered justification. Such a ruling would leave the Vermont legislature in a position to deliberate on the full ramifications of its actions in considering new legislation but still free to pursue it. See Quill v. Vacco, 80 F.3d 716, 742 (2d Cir.1996) (Calabresi, J., concurring) (“no court need or ought to make ultimate and immensely difficult constitutional decisions unless it knows that the state’s elected representatives and executives — having been made to go, as it were, before the people — assert through their actions (not their inactions) that they really want and are prepared to defend laws that are constitutionally suspect”), rev’d, 521 U.S. 793, 117 S.Ct. 2293, 138 L.Ed.2d 834 (1997); United States v. Then, 56 F.3d 464, 466 n. 1 (2d Cir.1995) (Cala-bresi, J., concurring) (“when factual developments have made a law’s prior justification constitutionally invalid, it is up to the legislature to decide whether to advance another state interest in support of that law”) (citing Abele v. Markle, 342 F.Supp. 800, 810-11 n. 18 (D.Conn.1972) (Newman, J., concurring), vacated, 410 U.S. 951, 93 S.Ct. 1412, 35 L.Ed.2d 683 (1973)); Tunick v. Safir, 209 F.3d 67, 74 (2d Cir.2000) (implying that state suicide statute could have been interpreted not to ban assisted suicide because “the New York Court of Appeals had never clarified, and the legislative history cast some doubt upon, the question of whether the New York ban on assisted suicide, first enacted in 1828, was ever meant to apply to a treating physician”) (internal quotation marks omitted).
VII. CONCLUSION
In holding, with only one dissenting vote, that limits on candidate expenditures are unconstitutional, Buckley simply followed the mainstream First Amendment jurisprudence that is applied to communicative activity of far less constitutional significance than political speech. See, e.g., Stanley, 394 U.S. at 567 (disallowing speculative governmental interest in banning obscene material as justification for statute restricting nonpolitical speech); Smith v. California, 361 U.S. 147, 151, 80 S.Ct. 215, 4 L.Ed.2d 205 (1959) (applying “stricter standards” in statutory scrutiny where statute has “a potentially inhibiting effect on speech”). That jurisprudence calls for scrutiny that does not take unquestioningly and at face value the claims of a law’s proponents without actually examining the law. See Buckley, 424 U.S. at 40-41, 96 S.Ct. 612 (stating that “[bjefore examining the interests advanced” in support of legislation, “[cjlose examination of the specifici*210ty of the statutory limitation is required where, as here, the legislation imposes criminal penalties in an area permeated by First Amendment interests”); id. at 41, 96 S.Ct. 612 (“The test is whether the language of [the statute] affords the ‘[precision of regulation [that] must be the touchstone in an area so closely touching our most precious freedoms.’ ”) (quoting NAACP v. Button, 371 U.S. 415, 438, 83 S.Ct. 328, 9 L.Ed.2d 405 (1963)). That jurisprudence demands that a law restricting speech, including editorializing speech by the press, spell out what it permits and what it prohibits in intelligible detail, see Forsyth, 505 U.S. at 131, 112 S.Ct. 2395, and not leave vast areas of discretion to those who must implement it. See Thomas, 534 U.S. at 323, 122 S.Ct. 775. That jurisprudence denies to government “the power to determine that spending to promote one’s political views is wasteful, excessive, or unwise,” Buckley, 424 U.S. at 57, 96 S.Ct. 612, and “to control ... the quantity and range of debate on public issues in a political campaign,” id. Under that jurisprudence, Act 64’s limits on expenditures and party financing cannot be upheld. Under that jurisprudence, forcing a reorganization of political parties that reduces the autonomy of local party committees for no articulated reason is unconstitutional.
If one looks at what Act 64 says instead of what its proponents say about it, it is quite apparent that the Act does indeed embody the theory of the defense witness who opined that government may regulate political speech the way it regulates electric companies, see supra Part 111(a). Although Act 64 is said to be aimed at reducing the corrupt influence of “special interests” while enhancing the role of “ordinary citizens,” it substantially disables ordinary citizens from meaningful participation in the political process. It cripples party organizations, particularly at the local level, and prevents individual grassroots activities on behalf of candidates, while organized economic interests retain the ability to engage in costly independent political advocacy. Moreover, the Act distinctly benefits the “special interest” that enacted it: incumbents.
In the beginning of this dissent, I quoted Justices Brandéis and Black on the dangers of high-minded assaults on liberty. In waging its broad attack on political activity in pursuit of its goal of transferring political power from “special interests” to “ordinary citizens,” Act 64 also exemplifies the wisdom of another, albeit less August, source, Walt Kelly: “We have met the enemy, and he is us.”
APPENDIX A
State of Vermont
Office of the Secretary of State
December 3,1999
Representative Terry Bouricius
56 Booth Street
Burlington, VT 05401
Re: Your e-mail of October 8, 1999
Dear Representative Bouricius,
Please accept my apology for the delay in responding to your questions. The review of my proposed opinion took longer than anticipated. This letter is in response to the two questions which you raised in your e-mail. I will restate the questions to make sure we understand the fact patterns which I am addressing.
1. Can a political party which has no “candidates” as defined in 17 V.S.A. § 2801(1) at the time the proposed poll is conducted, pay in excess of $500 for the conducting of a professional poll to seek potential voters opinions about selected po*211tential candidates, both Progressive and otherwise, and share the poll results with potential candidates without the polling expense and associated political party activities triggering either a person named in the poll becoming a “candidate” as defined in § 2801(1), nor disqualifying a person named in the poll from seeking public financing from the Vermont Campaign Fund, if a person named in the poll later decided to become a candidate?
The conducting of a poll by a political party to “test the waters” for various potential candidates will not trigger a “candidacy” for a person named in the poll even if more than $500 per potential candidate is expended by the party for the poll so long as only the general results are used for recruiting, media releases, or other generalized activities. The conducting of the poll itself falls within the types of activities generally pursued by a political party for its overall organization, planning, and strategy. The political party can conduct a poll and can make the general results or the poll public without triggering any candidacies. The party can use the results of the poll in a general way for recruiting candidates. (For example, telling John Jones that he was the favorite in a potential race with 3 other names, would be general information which can be used for recruiting.) However, as I will discuss below, the acceptance of detailed data and information from the poll by an individual will trigger a candidacy if the cost of the poll which is attributable on a pro rata basis to the provision of specific information and data to a particular candidate exceeds $500.
The Vermont campaign finance law does not specifically address polling activities and expenses. The definition of “candidate” in 17 V.S.A. § 2801(1) states that “an affirmative action” of “(A) accepting contributions or making expenditures of over $500” will trigger a candidacy. The definition of contribution includes “a gift of money or anything of value.” It is when an individual accepts the detailed data and information gained from the professional poll, that “a gift of anything of value” is accepted, and if the specific information given to an individual cost over $500 to produce on a pro rata basis, then a candidacy will be triggered.
Because it is the acceptance of the gift of detailed data from the poll, not the polling itself, which can trigger a “candidacy,” the political party could conduct a professional poll at any time but wait until after February 15, 2000 to offer any detailed data or information from the poll to an individual in order to avoid the prohibitions of the Vermont campaign fund (public financing). The law states that if a person becomes a candidate before February 15 of the general election year, that person shall not be eligible for Vermont campaign finance grants, 17 V.S.A. § 2853(a). Therefore, if before February 15th, a person accepts detailed data from a professional poll which is a gift of “anything of value” which cost over $500 to produce on a pro rata basis, the person has become a candidate by accepting a contribution totaling $500 or more, and will not be eligible for the Vermont campaign finance grants.
In summary, the political party can conduct polls and make the general results of the poll public without triggering any candidacies. If specific data and information is accepted by an individual which cost over $500 to produce on a pro rata basis, a candidacy is triggered. If the candidacy is triggered before February 15, 2000, the candidate will not be eligible for campaign finance grants. A political party and any individual considering accepting the detailed results of a poll should consult their *212own attorney to discuss specific fact patterns and how the law would be applied to those specific facts.
2. Can a political party spend in excess of $500 to arrange and sponsor dinners, other events, or party mailings for the purpose of educating party supporters or potential contributors about Vermont campaign finance grants and the need for “qualifying contributions”? Can the party make other preparations to assist a future candidate in gathering “qualifiers” for use after February 15, 2000?
Your question leaves room for many fact patterns so we will address three possible scenarios for such preparations, including dinners, events or mailings which we do not believe will trigger a candidacy and then discuss some additional considerations that might raise issues in the mind of an opposing candidate who could raise the issue using the process in 17 V.S.A. § 2809(e).
A. If the proposed events are for the sole purpose of educating voters about the need for many small contributors in order to qualify for Vermont campaign finance grants, to explain the importance of party organization, or to discuss any other topics related to general campaigning, the expenditures clearly would not trigger a candidacy.
B. If the party conducts mailings in which individuals are named or discussed as potential candidates because the party is hoping to generate interest in candidacies, but the individuals do not have knowledge of the fact that their names are mentioned, and the primary thrust of the activities are party organization and education of voters about the requirements of the campaign finance grant law, then the mailings would not trigger a candidacy.
C. If the party conducts other activities to develop a database of persons who might be willing to contribute “qualifying contributions” or solicits conditional pledges if an individual decided to run (see Secretary of State Letter of July 6, 1999), these activities would not trigger a candidacy.
D. However, if the party conducts dinners, events or mailings or solicits pledges in which potential candidates are introduced and discussed, and the potential candidates have participated in the planning of the events or given approval to them, the potential candidates will need to evaluate when or whether they might cross the line into candidacy by either accepting contributions or making expenditures of $500 or more by way of “accountability of related expenditures” as described in 17 V.S.A. § 2809. We cannot anticipate every possible fact pattern that may develop as the party and potential candidates proceed, so we merely want to raise the prospect that at some point the activities may raise questions in another candidate’s mind and the opposing candidate may use § 2809(e) to seek findings and a determination from a court. Each party and potential candidate should review proposed activities with their own counsel to examine the particular facts to evaluate whether they may fall into the category of activities addressed in 17 V.S.A. § 2809 as “related expenditures” and possibly trigger a candidacy based upon accountability for a related expenditure of $500 or more.
This letter represents my opinion on the issues which you have raised. You may seek your own counsel to advise you in these matters. Assistant Attorney General Michael McShane has reviewed this advisory opinion on behalf of the Office of the Attorney General and that office concurs with this opinion. If you have any ques*213tions please contact me at 828-2304 or kdewolfe@sec.state.vtus.
Sincerely,
Kathleen S. DeWolfe
Director of Elections and Campaign Finance
. Act 64 contains the following pertinent provisions. It states, by way of definition:
"Political party” means a political party organized under chapter 45 of this title or any committee established, financed, main-*155tamed, 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). In particular, it is this definition that governs the application of both the limits on contributions to, and expenditures by, "political parties,” e.g. any contribution to an affiliate of a party is a contribution to all affiliates and the state party-
The Act further provides:
Candidates who have made expenditures or received contributions of $500.00 or moré and political committees shall be subject to the following requirements:
(1) All expenditures shall be paid by check from a single checking account in a single bank publicly designated by the candidate or political committee.
(2) Each candidate and each political committee shall name a treasurer, who may be the candidate or spouse, who is responsible for maintaining the checking account.
Vt. Stat. Ann. tit. 17, § 2802.
Each political committee and each political party which has accepted contributions or made expenditures of $500.00 or more shall register with the secretary of state stating its full name and address, the name of its treasurer, and the name of the bank in which it maintains its campaign checking account within ten days of reaching the $500.00 threshold.
Vt. Stat. Ann. tit. 17, § 2831(a).
I do not read these provisions to prevent local affiliates from having separate bank accounts. I do read them, however, to .require that all contributions to parties go initially to the single checking account mentioned in Section 2831 for purposes of reporting and enforcing the limits on contributions to parties. Withdrawal from that checking account must be only with the consent of those authorized to sign checks.
. The desire to challenge Buckley, even at the cost of living under a bad law, is exemplified by an astonishing statement of Vermont's Secretary of State. In transmitting to the Vermont legislature a review of the operation of Act 64, she cautioned against any amendment or repeal that would render the present litigation moot, even if the legislature thought the amendment or repeal in the public interest, because such an action would "frustrate] the express legislative goal of giving the Supreme Court an opportunity to reevaluate its decision in Buckley v. Valeo.” See 2001 Memorandum, supra.
. Surprise at the actual provisions of campaign finance laws is as old as the laws themselves. A veteran civil-liberties lawyer tells the following story. In the summer of 1972, three old-time dissenters came into the offices of the New York Civil Liberties Union in Manhattan and told an extraordinary story. In May of that year they and a few like-minded others had drafted and sponsored a two-page advertisement that appeared in The New York Times. The advertisement was sharply critical of Richard M. Nixon, the President of the United States. The ad claimed that President Nixon had authorized the secret bombing of Cambodia, in violation of international law, and should be impeached and removed from office. The ad set forth the text of an impeachment resolution that had been introduced in the House of Representatives and contained an "Honor Roll” listing eight House members who had cosponsored that resolution. The advertisement cost approximately $17,850, and the ad hoc group called itself the National Committee for Impeachment. Before the ink on the ad was barely dry, the group was sued by the United States Justice Department for running the advertisement.
When Randolph Phillips, one of the sponsors of the ad, told this story to the lawyers at the New York Civil Liberties Union, we were incredulous. How could a group of citizens be sued by the Federal Government for publishing a criticism of the President of the United States? After all, this was 1972, and First Amendment law seemed at its most vigorous in the protections of public speech, one of the shining legacies of the Warren Court. What possible justification could the government have for suing this small group of protestors? We soon discovered the answer: campaign finance reform.” Joel M. Gora, No Law ... Abridging, 24 Harv. J.L. & Pub. Pol’y 841, 842-43 (2001) (reviewing Bradley A. Smith, Unfree Speech (2001)) (footnote omitted). The law in question was the Federal Election Campaign Act of 1971, which defined a political committee as any group that spent more than $1,000 annually "for the purpose of influencing” — language used in Act 64 — a federal election and imposed various requirements on a committee's purchase of advertisements relating to a federal candidate. See United States v. Nat’l Comm. for Impeachment, 469 F.2d 1135, 1139 (2d Cir.1972).
. Because other provisions of Act 64 strike at the heart of the democratic political system, I will note only in a footnote one egregiously overreaching provision. Act 64 requires that all political advertisements identify who paid for them, with an address, and which candidate is benefited. The Secretary of State has sought an exemption from this requirement for buttons and lapel stickers. See 2001 Memorandum, supra. So far, she has been unsuccessful.
. The pertinent provision reads:
"Expenditure” means a payment, disbursement, distribution, advance, deposit, loan or gift of money or anything of value, paid or promised to be paid, for the purpose of influencing an election, advocating a position on a public question, or supporting or opposing one or more candidates.
Vt. Stat. Ann. tit. 17, § 2801(3).
. My colleagues assert that the phrase expenditures "for the purpose of influencing an election” was "upheld” against a claim of unconstitutional vagueness by the Supreme Court in Buckley. Maj. Op. at 136 n.26. In fact, what the Court said was virtually the opposite.
The relevant passage in Buckley addressed a limit "for the purpose of influencing an election” on expenditures by persons who were neither candidates nor political committees. The "for the purpose of ...” language was modified by the phrase "relative to a clearly identified candidate.” Buckley stated that the definition of expenditures was unconstitutionally vague unless the adjectival phrase was construed narrowly to apply only to "communications that in express terms advocate the election or defeat of a clearly identified candidate.” 424 U.S. at 44, 96 S.Ct. 612 (expenditure limits apply only "to communications containing express words of advocacy of election or defeat, such as 'vote for,' ‘elect,’ 'support,' 'cast your ballot for,' ‘Smith for Congress,’ 'vote against,' 'defeat,' 'reject' ”, id. at n. 52, 96 S.Ct. 612; limits struck down on other grounds). As to reporting requirements imposed on persons who were neither candidates nor political committees regarding expenditures and contributions not made to candidates or political committees, the Court also held those to be impermissibly vague unless the phrase "for the purpose of influencing an election” was construed "in the same way” as the aforementioned terms, i.e., to apply only to expenditures and contributions expressly advocating the election or defeat of a clearly identified candidate. Id. at 80, 96 S.Ct. 612.
Act 64 does not contain the language "relative to a clearly identified candidate,” and relevant Vermont authorities have not construed Act 64 in this limited manner, see 2001 Guide, supra. As a result, the holding in Buckley invalidates Act 64's expenditure limits for vagueness.
. The Supreme Court's recent decision in McConnell did not alter the Buckley analysis. McConnell upheld the phrase "electioneering communication” — where those funding such communications faced restrictions and disclosure requirements — against a challenge of unconstitutional vagueness. The term is defined as " ‘any broadcast, cable, or satellite communication' " that " 'refers to a clearly identified candidate for Federal office,' ” is made during certain time periods, and " 'in the case of a communication which refers to a candidate other than President or Vice President, is targeted to the relevant electorate.' ” McConnell, 124 S.Ct. at 686-87 (quoting 2 U.S.C. § 434(f)(3)(A)(i)). A communication is " 'targeted to the relevant electorate' " if it " 'can be received by 50,000 or more persons' in the district or State the candidate seeks to represent.” Id. In upholding the term "electioneering communication,” the Court explained that it held "for the purpose of ... influencing” a federal election unconstitutionally vague in Buckley not because issue advocacy (as opposed to express advocacy) can never be limited, but rather because the limitations on advocacy imposed by that phrase were unconstitutionally vague. Id. at 688. "Electioneering communication,” in contrast, could be regulated because it describes only a well-defined subset of issue advocacy. Id. ("In narrowly reading the FECA provisions in Buckley to avoid problems of vagueness and overbreadth, we nowhere suggested that a statute that was neither vague nor overbroad would be required to toe the same express advocacy line.”). Another provision upheld in McConnell against a vagueness challenge was similarly specifically defined. Id. at 675 n. 64; 2 U.S.C. § 431(20)(A)(iii) (limits on contributions to fund a "public communication that refers to a clearly identified candidate for Federal office (regardless of whether a candidate for State or local office is also mentioned or identified) and that promotes or supports a candidate for that office, or attacks or opposes a candidate for that office (regardless of whether the communication expressly advocates a vote for or against a candidate),” 2 U.S.C. § 431(20)(A)(iii), not unconstitutionally vague because "[t]he words 'promote,' 'oppose,' 'attack,' and 'support' clearly set forth the confines within which potential party speakers must act in order to avoid triggering the provision,” McConnell, 124 S.Ct. at 675).
. For example, the central link on the Secretary of State website is entitled “Visit the Secretary’s Desk,” a section that includes a picture of the incumbent on each page. Vermont Secretary of State’s Website, at http://www.sec.state.vt.us/secdesk/index.html. This section also houses her “Biography,” which includes the following passage:
As Secretary of State, Markowitz has enhanced the office’s services to Vermont’s businesses, banks and professionals. She has made customer service a priority and created a state of the art web site to serve the business community. Markowitz has also protected consumers of professional services by reducing the backlog of professional licensing complaints and by starting a public information campaign to inform consumers of their rights to competent professional services. Markowitz has also promoted civics education in Vermont's schools and has encouraged Vermonters to be active participants in democracy by volunteering in town government and by voting.
ld. at http://www.sec.state.vt.us/secdesk/mar-ko.html.
The Vermont Attorney General’s Office has a similar website. The first page houses a photograph of the incumbent, whose biography page reads as follows:
Welcome to the Home Page for Vermont Attorney General William H. Sorrell. The Attorney General is the chief law enforcement officer in the state. He is charged with representing the state in all matters in which the state is a party or has an interest. The office of the Attorney General is dedicated to the protection of the health and safety of all Vermonters
A native and resident of Burlington, Vermont, Attorney General William H. Sorrell graduated from the University of Notre Dame (AB,'magna cum laude, 1970) and Cornell Law School (JD, 1974). Bill served as Chittenden County Deputy State’s Attorney from 1975-1977; Chittenden County State’s Attorney, 1977-78 and 1989-1992; engaged in private law practice at McNeil, Murray & Sorrell, 1978-1989; and served as Vermont's Secretary of Administration, 1992-1997. As State’s Attorney, he personally successfully prosecuted several significant matters, including the first case allowing the admissibility of DNA evidence in a Vermont State Court and a ten-year-old homicide in which the victim’s body had never been found.
Governor Howard Dean appointed General Sorrell to fill the unexpired term of now Vermont Chief Justice Jeffrey Amestoy, commencing May 1, 1997. He has enjoyed strong voter support in standing for election in November 1998, 2000 and 2002. His current term of office will expire in January 2005.
Bill is on the board of the American Legacy Foundation;- has served on the Judicial Nominating Board; as president of United *164Cerebral Palsy of Vermont; secretary of the Vermont Coalition of the Handicapped; and on the board of the Winooski Valley Park District. Bill has recently been elected the President-Elect of the National Association of Attorneys General (NAAG) and will assume the Presidency of that organization for a one-year term beginning in June of 2004. He is chair of the NAAG Tobacco Committee and co-chair of its Consumer Protection Committee. In June of 2003, Bill was selected by his peers from around the country to receive NAAG's Kelley-Wy-man Award, given annually to the "Outstanding Attorney General” who has done the most to further the goals of the nation's attorneys general.
Office of the Vermont Attorney General, at http://www.atg.state.vt.us/dis-play.php?smod=70.
. The pertinent provision reads:
"Candidate” means an individual who has taken affirmative action to become a candidate for state, county, local or legislative office in a primary, special, general or local election. An affirmative action shall include one or more of the following:
(A) accepting contributions or making expenditures totalling $500.00 or more; or
(B) filing the requisite petition for nomination under this title or being nominated by primary or caucus; or
(C) announcing that he seeks an elected position as a state, county or local officer or a position as representative or senator in the general assembly.
Vt. Stat. Ann. tit. 17, § 2801(1).
. My colleagues assert that the word "candidate” is not vague because candidates know when they are candidates, and because "the notion that candidates do not know when they are candidates is belied by the specificity of the provision itself.” Maj. Op. at 136 n.26. The issue, however, is not one of notice but one of enforcement. The question is not whether a candidate knows when he or she is a candidate, but what acts, or in the words of Act 64, "affirmative action[s],” will be deemed by Vermont authorities to trigger expenditure limits for purposes of the statute. Resolution of this question on a case-by-case basis ultimately rests within the discretion of some body external to the statute, whether it be the courts or the Secretary of State.
. The pertinent provision reads:
"Contribution” means a payment, distribution, advance, deposit, loan or gift of money or anything of value, paid or promised to be paid to a person for the purpose of influencing an election, advocating a position on a public question, or supporting or opposing one or more candidates in tuny election, but shall not include services provided without compensation by individuals volunteering their time on behalf of a candidate, political committee or political party. For purposes of this chapter, "contribution” shall not include a personal loan from a lending institution.
Vt. Stat. Ann. tit. 17, § 2801(2).
. The pertinent provision reads:
For the purposes of this section, a “related campaign expenditure made on the candi*167date's behalf" means any expenditure intended to promote the election of a specific candidate or group of candidates, or the defeat of an opposing candidate or group of candidates, if intentionally facilitated by, solicited by or approved by the candidate or the candidate's political committee.
Vt. Stat. Ann. tit. 17, § 2809(c).
. Section 2809(b) reads in full:
A related campaign expenditure made on a candidate’s behalf shall be considered an expenditure by the candidate on whose behalf it was made. However, if the expenditure did not exceed $50.00, the expenditure shall not be considered an expenditure by the candidate on whose behalf it was made.
Vt. Stat. Ann. tit. 17, § 2809(b).
. Section 2809(d) reads in full:
An expenditure made by a political party or by a political committee that recruits or endorses candidates, that primarily benefits six or fewer candidates who are associated with the political party or political committee making the expenditure, is presumed to be a related expenditure made on behalf of those candidates. An expenditure made by a political party or by a political committee that recruits or endorses candidates, that substantially benefits more than six candidates and facilitates party or political committee functions, voter turnout, platform promotion or organizational capacity shall not be presumed to be a related expenditure made on a candidate's behalf. In addition, an expenditure shall not be considered a “related campaign expenditure made on the candidate’s behalf" if all of the following apply:
(1) The expenditures were made in connection with a campaign event whose purpose was to provide a group of voters with the opportunity to meet the candidate personally-
(2) The expenditures were made only for refreshments and related supplies that were consumed at that event.
(3) The amount of the expenditures for the event was less than $100.00.
Vt. Stat. Ann. tit. 17, § 2809(d).
. There is an exemption for certain expenses relating to "meet the candidate” events in Section 2809(d). See supra notes 12, 14. Because Section 2809(d) relates generally to activities of political parties and committees and the exemption for meet the candidate events begins with the words "In addition,” the exemption may well have been intended to apply only to party or committee-sponsored events. The Secretary of State has indicated that the Attorney General believes it to apply to all such events, however sponsored. See 1999 Memorandum, supra. The Secretary of State’s position is unclear. In any event, the Secretary of State has opined that, even with the exemption, Act 64 unduly discourages such events because their costs involve non-exempted expenses. See id.
. Another common form of grassroots activity is the contribution of time and expert services by local professionals to candidates. Here, as is so often the case, Act 64 is ambiguous, but the provision of such services may well be deemed a contribution and related expenditure of some considerable size, particularly if the professional is in a firm that normally restricts the moonlighting of professional services. Indeed, to exempt professionals would be precisely counter to the purpose of reducing the influence of special interests — e.g., clients of lawyers or the lawyers themselves — and the well-to-do.
. For example, the 1974 Federal Election Campaign Act, the subject of the challenge in Buckley, set the limits on expenditures for House of Representative candidates in general elections at $70,000. See 424 TJ.S. at 55, 96 S.Ct. 612. Adjusted for inflation, this limit was below the amount spent by every successful challenger to a House incumbent in the 1972 election. See Joint Appendix at 269-318, Buckley (Nos. 75-436 and 75-437). Nevertheless, the $70,000 limit was far above the average expenditure by House candidates in 1972, $39,884. See id. at 447.
. When Howard Dean, the Vermont governor whose rallying cry for Act 64 is quoted in the opening paragraph of my colleagues’ opinion, ran for the Democratic nomination for President, he rejected public financing so that he could spend unlimited amounts. He made that decision at a time when all of his Democratic opponents were abiding by spending limits — i.e. no "arms race.” Although exact numbers are not available, Dean's rivals estimated that he spent as much as $6 million in the neighboring state of New Hampshire. Mike Madden, Democrats Scrambling For Cash As Well As Votes, Gannett News Service, Jan. 31, 2004.
. For example, in the 2000 election, the Democratic Party made cash contributions— not including related expenditures — in the amount of $28,000 to Elizabeth Ready for her campaign for Auditor of Accounts. See Campaign Finance Report of Elizabeth Ready, Dec. 18, 2000. Under Act 64, cash contributions and related expenditures could not have exceeded $400.
. A low level of judicial scrutiny necessarily leaves legislators with discretion to alter campaign finance regulations to affect upcoming elections. We have seen an example of this in New Jersey, where a campaign finance law was deliberately changed in an (unsuccessful) attempt to bolster the candidacy of a new entrant into the race for Governor. See David M. Halbfinger, Substitute Candidate, on Short Notice, Stakes Claim in Race for New Jersey Govemor, N.Y. Times, Apr. 27, 2001, at B5; David M. Halbfinger, New Jersey Legislature Votes To Delay Primaries 3 Weeks, N.Y. Times, Apr. 24, 2001, at B5. In Vermont, Governor Dean sought to use money reserved for the public financing of campaigns — a key part of Act 64 — to pay general state expenditures. See State May Tap Campaign Finance Fund to Ease Budget Crunch, Associated Press, Dec. 5, 2001.
. Some may doubt that Act 64 was intended to apply to media editorializing because they deem such editorializing to be constitutionally protected. However, paid advertisements have the same protection as editorials, see Sullivan, 376 U.S. at 266, 84 S.Ct. 710 (holding that "statements [that] would otherwise be constitutionally protected ... do not forfeit that protection because they were published in the form of a paid advertisement”), and if government may constitutionally limit paid advertisements, as Act 64 does, government may limit unpaid endorsements. I, of course, believe that government cannot limit either.
The reason many campaign finance laws exempt the media is, therefore, not constitutional scruple, but the desire of proponents of regulation for media exposure and support. Such support might not be forthcoming if the media realized the extent to which the theory of such laws is a dagger easily aimed at freedom of the press.
. My colleagues note that one purpose of Act 64’s expenditure limits was to reduce the use of short commercials by candidates but, in light of their disposition of this matter, do not reach the issue of whether this concern is sufficiently compelling to justify the legislation. In that regard, I note two things. First, that interest is not compelling. Indeed, the use of law to force candidates to select one medium of advocacy rather than another is an unconstitutional purpose and an additional ground for striking expenditure limits down. See Meyer, 486 U.S. at 424, 108 S.Ct. 1886 (“The First Amendment protects [individuals'] right not only to advocate their cause but also to select what they believe to be the most effective means for so doing.”). Second, Act 64 has increased reliance on the media. See 2001 Memorandum, supra; David Gram, Dems Needle Each Other on Spending in Treasurer’s Race, Associated Press, May 29, 2002.
. My colleagues cite a Sixth Circuit concurrence for the proposition that Buckley was " ‘decided on a slender factual record,' ” Maj. Op. at 109 (citing Kruse v. City of Cincinnati, 142 F.3d 907, 919 (6th Cir.1998) (Cohn, J., concurring)). To this they add citations to a treatise, a law review article, and a student note for the proposition that Buckley was decided without a "factual” record. The authors of these works could not have been familiar with the actual record before the Court in Buckley, which contained over 700 pages of statistical and testimonial data and findings of fact, as described below. Those materials are a matter of public record, and the Buckley briefs and oral arguments can be found in a published, two-volume work. 1976 Landmark Briefs and Arguments of the Supreme Court of the United States: Buckley v. Valeo (Phillip B. Kurland and Gerhard Casper, eds. 1977) (hereinafter "Landmark Briefs").
Buckley involved a major piece of legislation passed after extensive congressional hearings in which critics of the private financing of elections supported their case with massive submissions of evidence. Legislative History of Federal Election Campaign Act Amendments of 1974, The Federal Election Commission (1977). The Supreme Court’s decision indicated familiarity with this body of evidence. See, e.g., Buckley, 424 U.S. at 20 nn. 20-21, 96 S.Ct. 612 (giving election-related statistics); id. at 22 n. 23, 96 S.Ct. 612 (same); id. at 26 n. 27, 96 S.Ct. 612 (same). Other materials before the Supreme Court in Buckley include, inter alia, a Joint Appendix of 762 pages, which compiled findings of fact and statistical findings agreed to by the parties, Joint Appendix at 4-698, Buckley (Nos. 75-436 and 75-437), and the district court’s findings of fact, id. at 699-753.
The agreed upon findings of fact included data from opinion polls on public perceptions of politicians, political participation, expenditure limits, and cynicism about government, id. at 160-89, 207-53, detailed catalogs of specific contributions by labor unions, PACs, and business organizations to individual candidates, id. at 55-146, data on expenditures in presidential elections from 1912 to 1968 indicating increasing spending and an increasing cost per vote, id. at 50-51, the cost of postage and newspaper advertising, id. at 29-32, advantages of incumbents over challengers, id. at 16-24, statistics indicating declining voter participation, id. at 9, evidence that candidates generally focused on wealthy donors but that candidates who limited their expenditures had been successful in the past, id. at 256-59, and evidence that elected officials give preferential access to large contributors, id. at 256-57.
The agreed upon statistical findings included data on and analysis of contributions to, expenditures by, and election results for, all congressional candidates and political committees that filed reports in the 1972 and 1974 elections, id. at 270-440, 571-72, 619-78, votes received by challengers versus incumbents in the 1974 House races, id. at 679-96, evidence of specific individuals and groups donating money to congressional candidates on committees relevant to their businesses, id. at 462-64, 467-72, statistics on independent expenditures, id. at 472-73, data on individuals who contributed large sums to 1972 congressional elections and the Committee to Reelect the President, id. at 479-564, a detailed analysis of the 1972 elections which discussed, among other things, the costs of raising money from large versus small donors and the relationship of expenditures to success in elections, id. at 571-586, and a ten volume study by Common Cause entitled 1972 Congressional Campaign Finances, id. at 698.
The Buckley district court’s findings of fact were based on the testimony and affidavits of fifteen individuals, and basically summarized that testimony. Id. at 699. These findings included opinions on the importance of seed money to challengers, id. at 703, 714, the ways in which expenditure limits favor incumbents over challengers and third party candidates, id. at 727-33, 713-19, indices of success other than winning or losing, id. at 712, and the ability to run a successful campaign with little money, id.
In Buckley, therefore, the Court had before it extensive hard data regarding contributions to, and expenditures by, candidates for feder*187al offices, as well as a multitude of reflections and opinions on the role of money in campaigns by persons familiar with American electoral politics. Moreover, the defense in Buckley included not only the government but also various groups, including Common Cause and the League of Women Voters, who were allowed to intervene as full parties and were represented by the Washington law firm, Wilmer, Cutler, and Pickering and by Archibald Cox of the Harvard Law School, a former Solicitor General of the- United States.
. See Buckley v. Valeo, 519 F.2d 821, 838 (D.C.Cir.1975). The Court of Appeals found that:
In practice ... candidates were compelled to allot to fund raising increasing and extreme amounts of time and energy. Senator Hollings testified that survival required candidates for national office to "set down a policy where they won't go see people other than those who can give money.” Joseph Cole, finance chairman for the Democratic National Committee, testified from his experience in some four or five Presidential campaigns, how dog-tired candidates must arise early in pursuit of large contributions, and continue "all day long and all night long.” "[How] much time do you think a Presidential candidate spends on fund raising? ... at least 70 percent of his time, and I think all of his waking hours. It is really demeaning, demeaning to go through it.”
Id. (footnotes omitted).
. See Landmark Briefs: 2 Buckley v. Valeo (Brief for the Attorney General as Appellee and for the United States as Amicus Curiae). The relevant section of the brief stated:
Fund raising consumes candidate time that otherwise would be devoted to campaigning.
The court of appeals found support for the statute in the fact that in order to raise large amounts of money to support a campaign, "candidates [are] compelled to allot to fund raising increasing and extreme amounts of time and energy.” A past finance chairman of the Democratic National Committee testified that a presidential candidate is required to spend 70 percent of his time in pursuit of funds.
There is, of course, another side to this problem. If the idea is that expenditure limits relieve the pressure of raising funds, thereby giving the candidate more time to engage in speaking, there may be effective alternative means of accomplishing this end. For example, public financing of election campaigns, quite independent of restrictions upon contributions and expenditures, will provide some relief. A restriction on contributions will increase this effect even without an accompanying restriction on expenditures; once candidates are precluded from drawing upon wealthy donors who command personal attention, campaign fund raising may turn more to direct mail efforts that are sparing of the candidate's time.
Id. at 434-35 (internal citations and footnotes omitted).
.See Landmark Briefs: 2 Buckley v. Valeo (Brief for Appellee Center for Public Financing of Elections, Common Cause, League of Voters, et al.). The brief stated:
Th[e] rising thirst for money has forced candidates to divert time and energy to fund-raising activities and away from other activities, such as addressing the substantive issues, that do not fill campaign coffers. Joseph Cole, who was National Finance Chairman of the Democratic National Committee, vividly recounted the effects of the pressure to raise money on candidates in a passage quoted by the court of appeals:
"... I have been close to four or five Presidential campaigns .... I have sat next to the Presidential candidate, who was tired, who was weary and concerned with the issues and not able to handle *189them, not able to prepare, not able to think about them because he has to go downstairs at 7 in the morning to shake hands with a guy from whom he may get a large contribution.”
"It goes on all day long and all night long, and I was asked at the Senate hearings how much time do you think a Presidential candidate spends on fundraising? And I said at least 70 percent of his time, and I think all of his time, and I think all of his waking hours. It is really demeaning, demeaning to go through it.”
Id. at 97, 96 S.Ct. 612 (footnotes omitted); see also Landmark Briefs: 2 Buckley v. Valeo, (Brief of Senators Hugh Scott and Edward M. Kennedy as Amici Curiae) ("The pressure upon candidates to raise money from large contributors had become so great as to leave them little time for ordinary citizens.”) (footnote omitted).
. See Buckley v. Valeo, 424 U.S. at 91, 96 S.Ct. 612 ("In this case, Congress was legislating for the 'general welfare’ ... to free candidates from the rigors of fundraisers”); id. at 96, 96 S.Ct. 612 ("In addition, the limits on contributions necessarily increase the burden of fundraising, and Congress properly regarded public financing as an appropriate means of relieving major-party Presidential candidates from the rigors of soliciting private contributions.”); S. Rep. 93-689, 5 (cited in above quotations from Buckley, and justifying public financing because "[mjodem campaigns are increasingly expensive and the necessary fundraising is a great drain on the time and energies of the candidates”); see also Buckley, 424 U.S. at 258-59, 96 S.Ct. 612 (White, J., concurring in part and dissenting from the Court’s view that the expenditure limits were unconstitutional) (“In another major innovation, aimed at insulating candidates from the time consuming and entangling task of raising huge sums of money, provision was made for public financing of political campaigns for federal office.”).
. For example, the Court of Appeals in Buckley stated:
Looming large in the perception of the public and Congressmen was the revelation concerning the extensive contributions by dairy organizations to Nixon fund raisers, in order to gain a meeting with White House officials on price supports. The industry pledged $2,000,000 to the 1972 campaign, a pledge known to various White House officials, with President Nixon informed directly by Charles Colson in September 1970, as acknowledged by the 1974 White House paper .... On March 23, 1971, after a meeting with dairy organization representatives, President Nixon decided to overrule the decision of the Secretary of Agriculture and to increase price supports. In the meetings and calls that immediately followed the internal White House discussion and preceded the public announcement two days later, culminating in a meeting held by Herbert Kalmbach at the direction of John Ehrlichman, the dairymen were informed of the likelihood of an imminent increase and of the desire that they reaffirm their $2 million pledge. 519 F.2d at 840 n. 36.
. My colleagues suggest in a footnote citing to McConnell, see Maj. Op. at 117, n.12, that speech may be suppressed based solely on anecdotal evidence, and that reliance on a couple of untested and untestable anecdotes from Act 64's supporters is sufficient. However, McConnell — which dealt with contributions, not expenditures — does not stand for any such proposition. McConnell explained specifically how ''[t]he evidence connects soft money to manipulations of the legislative calendar, leading to Congress' failure to enact, among other things, generic drug legislation, tort reform, and tobacco legislation.” 124 S.Ct. at 664 (citing among other sources the declaration of former Senator Alan Simpson that "Donations from the tobacco industry to Republicans scuttled tobacco legislation, just as contributions from the trial lawyers to Democrats stopped tort reform”). McConnell also noted that
The evidence in the record shows that candidates and donors alike have in fact exploited the soft-money loophole, the former to increase their prospects of election and the latter to create debt on the part of officeholders, with the national parties serving as willing intermediaries.... federal officeholders have commonly asked donors to make soft-money donations to national and state committees "solely in order to assist federal campaigns,” including the officeholder’s own. Parties kept tallies of the amounts of soft money raised by each officeholder, and "the amount of money a Member of Congress raise[d] for the national political committees often affect[ed] the amount the committees g[a]ve to assist the Member's campaign.” Donors often asked that their contributions be credited to particular candidates, and the parties obliged, irrespective of whether the funds were hard or soft. National party committees often teamed with individual candidates' campaign committees to create joint fundraising committees, which enabled the candidates to take advantage of the party’s higher contribution limits while still allowing donors to give to their preferred candidate. Even when not participating directly in the fundraising, federal officeholders were well aware of the identities of the donors: National party committees would distribute lists of potential or actual donors, or donors themselves would report their generosity to officeholders.
For their part, lobbyists, CEOs, and wealthy individuals alike all have candidly admitted donating substantial sums of soft money to national committees not on ideological grounds, but for the express purpose of securing influence over federal officials.
Id. at 662-63 (quoting and citing statements by politicians, CEOs and lobbyists).
To support their argument that anecdotal evidence may, by itself, serve to repress speech, my colleagues parse McConnell's statement that "The record in this case is replete with similar examples of national party committees peddling access to federal candidates and officeholders in exchange for large soft-money donations.” Id. at 664, see Maj. Op. at 117 n.12. In fact, this statement does not reference a record "replete” with anecdotal evidence, as my colleagues seem to believe, but introduces a factual discussion:
So pervasive is this practice that the six national party committees actually furnish their own menus of opportunities for access to would-be soft-money donors, with increased prices reflecting an increased level of access. For example, the DCCC offers a range of donor options, starting with the $10,000-per-year Business Forum program, and going up to the $100,000-per-year National Finance Board program. The latter entitles the donor to bimonthly conference calls with the Democratic House leadership and chair of the DCCC, complimentary invitations to all DCCC fundraising events, two private dinners with the Democratic House leadership and ranking members, and two retreats with the Democratic House leader and DCCC chair in Telluride, Colorado, and Hyannisport, Massachusetts. Similarly, "the RNC’s donor programs offer greater access to federal office holders as the donations grow larger, with the highest level and most personal access offered to the largest soft money donors.”
Id. at 665 (noting parenthetically "records indicating that DNC offered meetings with President in return for large donations”).
. In the last mayoral election in New York City, a candidate claimed to be eligible for *202public financing in a primary race, but his application was denied in a debatable ruling. See Mirta Ojito, Badillo Campaign Denied Matching Funds-, N.Y. Times, Sept. 8, 2001, at B6; Mirta Ojito, Badillo Appeals Ruling on Campaign Fund Match, N.Y. Times, July 25, 2001, at B4. A Campaign Finance Board later overturned that ruling but only after the primary election was over. See A Bit Late for Race, Badillo Gets Funds, N.Y. Times, Apr. 12, 2002, at B3.
. In a footnote responding to this dissent, my colleagues concede that "[o]f 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.” Maj. Op. at 136 n.25 (emphasis in original). Unfortunately, the text of their opinion continues to direct the district court on remand to consider only what was "considered” by the legislature. See id. at 135 ("On remand, the District Court ought consider, ... (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.”) Were I to eliminate my criticism of their text, their footnote would become superfluous and similarly removed, leaving the text of their opinion stating the incorrect standard.
. Even where proponents do report spending for mileage, the numbers are rounded-off and therefore appear to be estimates. See, e.g., Campaign Finance Report of Anthony Pollina, Sept. 25, 2002 (reporting in kind contribution of mileage by self of $2720); see also Campaign Finance Reports of Deborah Markowitz, Sept. 25, Oct. 25, and Dec. 16, 2002 (reporting in kind contributions of "postage, travel, phone”) by Paul Markowitz of $525, $460, and $120, respectively. These round numbers are not divisible by 31<t — the per mile cost of gas assigned by the Vermont Secretary of State; see 2001 Memorandum, supra.