Dissenting in Part:
Under Buckley, contribution limitations can be upheld only “if the State demonstrates a sufficiently important interest and employs means closely drawn to avoid unnecessary abridgment of associational freedoms.” Buckley v. Valeo, 424 U.S. 1, 25, 96 S.Ct. 612, 46 L.Ed.2d 659 (1976). I agree with the majority that Montana has *1099a sufficiently important interest in preventing corruption and the perception of corruption in Montana elections. I do not disagree with the majority in upholding 'the individual contribution limits placed on individuals and PACs. Such limits are closely drawn to the significant interest of preventing improper influence, and quid pro quo arrangements arising from large contributions. As intended, the individual limits target the upper 10% of contributions.
Where I depart from the majority is on the constitutionality of the aggregate PAC contribution limit. I disagree that the State has demonstrated a “genuine threat to its important governmental interests” or has “employ[ed] means closely drawn to avoid unnecessary abridgment” of protected activity. Citizens Against Rent Control/Coalition for Fair Housing v. City of Berkeley, 454 U.S. 290, 302, 102 S.Ct. 434, 70 L.Ed.2d 492 (1981) (Blackmun, & O’Connor, J.J., concurring) (internal quotation marks omitted).
The Supreme Court has previously defined corruption as “a subversion of the political process” where “[ejlected officials are influenced to act contrary to their obligations of office by the prospect of financial gain to themselves or infusions of money into their campaigns.” Federal Election Comm’n v. National Conservative Political Action Comm., 470 U.S. 480, 497, 105 S.Ct. 1459, 84 L.Ed.2d 455 (1985) (“NCPAC”). I agree that Montana has a significantly important interest in preventing corruption associated with large contributions. However, I submit that large individual contributions from persons and PACs have been addressed by Montana’s individual contribution limits as set forth in Mont.Code Ann. § 13-37-216 (2001). I find that having a limit on the amount an individual PAC may contribute to a candidate sufficiently prevents any one PAC from exerting “unfair influence” over a candidate. Nevertheless, the State has chosen to enact an aggregate PAC contribution limit to prevent a candidate from being overly influenced by special interests generally. The predicate for such a position must necessarily be that all PACs operate with a monolithic agenda. This ignores the obvious. Like individual persons, each PAC has its own interests and its own reasons for contributing. There is no evidence to support a proposition that all PACs exert unfair influence, or are collectively capable of doing so. I conclude that not only has the State failed to demonstrate a genuine threat, i.e., that all PAC contributions exert an unfair influence over candidates to justify the State’s interest in preventing perceived and actual corruption, but the State has also failed to employ means closely drawn to that interest.
I. Inadequate evidence exists to sustain the aggregate limit.
While states should be permitted to respond to potential electoral deficiencies “with foresight rather than reactively,” the response must not significantly impinge on constitutionally protected rights. Munro v. Socialist Workers Party, 479 U.S. 189, 195, 107 S.Ct. 533, 93 L.Ed.2d 499 (1986). To sustain the aggregate PAC contribution limit, we must find under the present law a serious threat of abuse exists from collective PAC special interest contributions. See Federal Election Comm’n v. Colorado Republican Fed. Campaign Comm., 533 U.S. 431, 457, 121 S.Ct. 2351, 150 L.Ed.2d 461 (2001).
Here, Montana’s asserted purpose for the aggregate PAC limit is to prevent a candidate from being overly influenced by special interests.1 The majority finds that *1100the State’s evidence of isolated incidents speaks not to particular PACs but to the nature of special interest groups in general. From this narrow evidence, the majority concludes that the regulation is necessary to limit the total amount of PAC contributions and encourage a diverse base of support in order to “eliminate the corrosive effects of large amounts of special interest money.” It appears that Montana and the majority equate all special interest contributions with corruption without any evidentiary support.
Because the aggregate limit discriminates between PACs and individuals, the “discrimination itself [must be] necessary to serve a substantial governmental interest.” Arizona Right to Life PAC v. Bayless, 320 F.3d 1002, 1010-11 (9th Cir.2003). Moreover, while some courts have recognized that the potential danger of corruption is greater with respect to PAC contributions than with individuals, I find that the State has failed to demonstrate a serious threat of influence by all PACs in Montana to justify the aggregate limit or that the discrimination between PACs and individual donors is necessary to serve a substantial governmental interest. Even if one assumes there are instances of abuse by particular PACs, an aggregate limit on PAC contributions is no more justified than an aggregate limit on all individual contributions to regulate abuses by a particular contributor. Notably, in striking down a criminal statute limiting PAC expenditures, the Supreme Court observed that even if the large pooling of financial resources by PACs poses a potential for corruption or the appearance of corruption, a PAC expenditure limit is overly broad. See NCPAC, 470 U.S. at 498, 105 S.Ct. 1459. It applies not only to “multimillion dollar war chests,” but equally to “informal discussion groups that solicit neighborhood contributions.” Id.
By contrast, in upholding a limitation on corporate contributions and independent expenditures, the Supreme Court in Austin v. Michigan Chamber of Commerce, 494 U.S. 652, 660, 110 S.Ct. 1391, 108 L.Ed.2d 652 (1990), specified that the “mere fact that corporations may accumulate large amounts of wealth is not the justification for [the expenditure restriction]; rather, the unique state-conferred corporate structure ... warrants the limit on independent expenditures.” The Court upheld the limitation on corporation expenditures based on the inherent structure of the corporation — a structure which exists in all corporations. Here, as in VanNatta v. Keisling, the State is unable to point to any evidence which demonstrates that all PAC contributions inherently lead to the sort of corruption that Montana purportedly seeks to prevent. 151 F.3d 1215, 1221 (9th Cir.1998) (holding unconstitutional a contribution limit on out-of-district residents because it was not closely drawn to advance the goal of preventing corruption).
Interestingly, in footnote 2, the majority concludes that VanNatta and another case, Service Employees Int’l Union v. Fair Political Practices Com’n, 955 F.2d 1312 (9th Cir.1992), were “superseded” by Nixon v. Shrink Missouri Gov’t PAC, 528 U.S. 377, 120 S.Ct. 897, 145 L.Ed.2d 886 (2000). The majority’s dismissal of Van-Natta and Service Employees, is inconsistent with Miller v. Gammie, 335 F.3d 889 (9th Cir.2003) (en banc). In Miller, the court addressed “when, if ever, a district court or a three-judge panel is free to reexamine the holding of a prior panel in *1101light of an inconsistent decision by a court of last resort on a closely related, but not identical issue.” Id. at 899. The court concluded that intervening Supreme Court authority “need not be identical,” but the decision “must have undercut the theory or reasoning underlying the prior circuit precedent in such a way that the cases are dearly irreconcilable.” Id. at 900 (emphasis added).
Neither VanNatta nor Service Employees is clearly irreconcilable with Shrink Missouri. As discussed in more detail below, the volume of evidence presented in Shrink Missouri to justify its individual limits was substantial. See, e.g., 528 U.S. at 393-94, 120 S.Ct. 897. That factor alone is sufficient to distinguish Shrink Missouri from VanNatta where the court found that the State had failed “to point to any evidence which demonstrates that all out-of-district contributions lead to the sort of corruption discussed in Buckley.” VanNatta, 151 F.3d at 1221.
Similarly, Service Employees is distinguishable from Shrink Missouri because the former involved, among other things, an outright ban on certain types of campaign contributions, while the latter only involved limits on contributions. In finding a ban on inter-candidate contributions unconstitutional, the court in Service Employees noted that “[t]he potential for corruption stems not from campaign contributions per se but from large campaign contributions.” 955 F.2d at 1323. The majority does not explain how that decision is inconsistent, much less clearly irreconcilable, with the Supreme Court’s decision in Shrink Missouri upholding individual contribution limits. See Shrink Missouri, 528 U.S. at 387-89, 120 S.Ct. 897.
Moreover, under both VanNatta and Shrink Missouri, actual evidence is required; mere conjecture that special interest money corrodes politics in Montana is inadequate to carry a First Amendment burden. See Shrink Missouri, 528 U.S. at 392, 120 S.Ct. 897; VanNatta, 151 F.3d at 1221.2 Nevertheless, the majority finds that the evidence here exceeds the evidence presented in Shrink Missouri. I disagree. In Shrink Missouri, the State presented an affidavit from a state senator who expressed that large contributions have “ ‘the real potential to buy votes.’ ” Id. at 393, 120 S.Ct. 897 (quoting Shrink Mo. Gov’t PAC v. Adams, 5 F.Supp.2d 734, 738 (E.D.Mo.1998)). There were newspaper accounts of large contributions supporting inferences of impropriety. One such account examined the state treasurer’s decision to engage in substantial state business with a bank which contributed $20,000 to the treasurer’s campaign. Shrink Missouri, 528 U.S. at 393, 120 S.Ct. 897. Another report disclosed a $40,000 contribution from a brewery and one for $20,000 from a bank to a candidate for state auditor. Id. A PAC linked to an investment bank contributed $420,000 to candidates in northern Missouri; three scandals ensued including one involving a state representative who was “ ‘accused of sponsoring legislation in exchange for kickbacks.’ ” Id. (quoting Carver v. Nixon, 72 F.3d 633, 642, and n. 10 (8th Cir.1995)). Another resulted in Missouri’s former attorney general pleading guilty to charges of conspiracy to misuse state property after being indicted for using a state work*1102er’s compensation fund to benefit campaign contributors. Shrink Missouri, 528 U.S. at 393-94, 120 S.Ct. 897. Finally, “ ‘an overwhelming 74 percent of the voters of Missouri determined that contribution limits are necessary to combat corruption and the appearance thereof.’ ” Id. at 394, 120 S.Ct. 897 (quoting Carver v. Nixon, 882 F.Supp. 901, 905 (W.D.Mo.1995)). These instances of actual and perceived corruption based on large contributions sufficiently justified the individual limits upheld in Shrink Missouri.
Not only do Montana’s instances of corruption pale by comparison to the facts in Shrink Missouri, in my opinion, they simply do not demonstrate a serious threat of abuse by all PACs to justify an aggregate limit on PACs while simultaneously raising the amount political parties may contribute to a candidate. Montana cites a memorandum by a Republican legislator as evidence of corruption in the Montana legislature. The memorandum states that the legislator wants to keep the particular PAC money within the Republican party and not allow that money to be shared with Democrats. Although five separate investigations were conducted, no convictions, or even indictments ensued. Thus, this incident does not justify a restriction placed on PACs generally. It is ironic that the State cites this episode as a justification for restraining aggregate PAC contributions to candidates, when it can be cited as powerful evidence of PAC influence on political parties, parties to which this very legislation permits PACs to make unlimited contributions. As discussed below, this anomaly simply dramatizes how inadequately tailored this law is to prevent a supposed corruption of Montana politics.
Montana also cites efforts by the gambling industry to prevent the passing of an automatic system of monitoring video gambling and efforts by the electric power industry to deregulate prices as examples of corruption associated with PAC contributions. Without any evidence of quid pro quo arrangements or other illegal or improper conduct, Montana asserts that these “results” are manifest examples of undue influence by PAC money. I find that this evidence is at most inconclusive.
Next, the State points to a poll of voters in support of campaign reform as evidence of perceived corruption. This poll did not specifically address PACs or special interest groups. Even if it did, I question whether a poll of the constituents is sufficient evidence, or is even probative to show the existence of perceived corruption. Issues of fundamental freedom should not be decided by majority vote, much less by a public opinion poll; thus, the poll results here should not be considered by the panel. The Tenth Circuit noted that “[w]e should not allow generic public dissatisfaction to support the restriction of political speech.” Federal Election Comm’n v. Colorado Republican Fed. Campaign Comm., 213 F.3d 1221, 1230 n. 6 (10th Cir.2000), rev’d on other grounds, 533 U.S. 431, 121 S.Ct. 2351, 150 L.Ed.2d 461 (2001) (citing NCPAC, 470 U.S. at 499-500, 105 S.Ct. 1459) (“newspaper articles and polls purportedly showing a public perception of corruption” are insufficient to justify a limitation on the independent expenditures of PACs).3 In NCPAC, the Supreme Court *1103implicitly affirmed the district court’s exclusion of the proffered evidence of newspaper articles and polls purportedly showing a public perception of corruption as irrelevant. 470 U.S. at 499, 105 S.Ct. 1459. Likewise, in an Eighth Amendment challenge to capital punishment for 16- and 17-year-old offenders, the Supreme Court “deeline[d] the invitation to rest constitutional law upon such uncertain foundations” as public opinion polls. Stanford v. Kentucky, 492 U.S. 361, 377, 109 S.Ct. 2969, 106 L.Ed.2d 306 (1989). The panel should follow the Supreme Court’s dictate to reject public opinion polls on issues of constitutional importance.4
Finally, the State relies on low voter turnout to demonstrate the negative impact on public perception created by the involvement of large amounts of PAC money in political campaigns. I agree that voter turnout may be a better measure of the constituency’s perceived corruption of elections. The Supreme Court, however, has noted that the fact that the voters passed the initiative to establish contribution limits is not dispositive; “majority votes do not defeat First Amendment protections.” Shrink Missouri, 528 U.S. at 394, 120 S.Ct. 897. Montana enjoys one of the five highest voter turnouts in the country. (ER 246). Although the State asserts that nearly half of the eligible voters in Montana did not vote in the 1998 general election because they lacked faith that their vote made a difference, the record does not reflect increased voter confidence post-reform.
Thus, the State fails to justify the need for the aggregate PAC contribution limit, in addition to the individual PAC limit, in order to prevent the type of corruption or perceived corruption that allegedly exists in Montana. Based on the inconclusive evidence of undue influence by PAC contributions in the record, the individual PAC limit more than suffices to prevent any undue influence on legislators caused by large contributions. I have found no basis for concluding that PACs are generally corrupt or perceived to be corrupt, nor any justification for further restricting PAC participation in legislative campaigns.
II. The limit is not closely drawn to avoid unnecessary abridgement of First Amendment freedoms.
The majority concludes that there is sufficient justification in the record to demonstrate that the danger of corruption (or the appearance of corruption) is greater when dealing with PAC money as opposed to other contributions. Assuming that PACs are perceived to be a source of corruption, the State’s goal of encouraging a diverse base of support is thwarted when the State places a limit on the aggregate amount of *1104PAC contributions a candidate may receive. By instituting such a limit, the State has, unjustifiably, restricted PAC participation in legislative campaigns and unnecessarily abridged their associational freedoms.
To be “closely drawn,” the degree of restriction must bear a sufficiently close relation to the reasons proffered by the State. We must consider whether, “within the full panoply of legislative choices otherwise available to the State, there exist alternative means of furthering the State’s purpose without implicating constitutional concerns.” Supreme Ct. of Va. v. Friedman, 487 U.S. 59, 67, 108 S.Ct. 2260, 101 L.Ed.2d 56 (1988) (deciding whether the degree of discrimination is closely drawn to the State’s reasons in the Privileges and Immunities clause context); see also Elrod v. Burns, 427 U.S. 347, 363, 96 S.Ct. 2673, 49 L.Ed.2d 547 (1976) (“to survive constitutional challenge, [the encroachment of First Amendment protections] must further some vital government end by a means that is least restrictive of freedom of belief and association in achieving that end, and the benefit gained must outweigh the loss of constitutionally protected rights.”) (emphasis added).
The State argues that by placing an aggregate limit on PAC contributions, candidates obtain a more diverse base of contributions and thereby prevent undue influence by special interests. However, the State fails to demonstrate that PAC contributions, when individually limited to $400/ $200/$100 depending on the office, are not sufficiently curtailed to alleviate a perception of corruption or a danger of undue influence. The State contends that large amounts of special interest money have a corrosive effect on Montana politics. It fails to show that PACs in general create a perception of corruption. I acknowledge that courts have found that the danger of corruption is greater with PAC contributions than with individuals. Such a determination may justify individual limits to combat quid pro quo arrangements from large contributors, but it does not justify an overbroad aggregate PAC limit that does little more than restrict speech and association rights.
Unlike Shrink Missouri, where individual PAC contributions are limited to a certain amount based on the specified state office or size of constituency, here aggregate PAC contributions are capped and a candidate must give back a portion of its PAC collections in order to receive contributions from another PAC. Clearly, Montana’s aggregate limit has a more restrictive effect than the contribution limits upheld in Shrink Missouri. This restrictive effect is meaningful for purposes of determining whether the State has demonstrated that the aggregate limit is closely drawn to match a sufficiently important interest. See Beaumont, 123 S.Ct. at 2211 (“It is not that the difference between a ban and a limit is to be ignored; it is just that the time to consider [the difference] is when applying scrutiny at the level selected .... ”). This demonstrates that the State has not employed the least restrictive alternative in furthering its purpose. See Service Employees, 955 F.2d at 1312 (rejecting ban on inter-candidate contributions where the ban did not distinguish on the basis of the size of the donation).
Montana has asserted that corruption (or perceived corruption) exists when PAC contributions make up a large portion of a candidate’s campaign treasury, but has not made a closely drawn determination of what portion of a candidates’ contributions is “large” enough to create the perception of corruption. Rather, Montana has made an arbitrary determination of what amount of PAC money is allowed, without regard to the ratio of PAC money to other contributions.
*1105While the district court found that the aggregate PAC contribution limit has lowered the amount of PAC money to about 29% of all contributions received by the average candidate, the same result can be achieved when a candidate receives more individual contributions or party contributions, rendering PAC contributions a smaller portion of the candidates’ campaign collection. Thus, Montana’s aggregate PAC contribution limit arbitrarily restricts more First Amendment rights than necessary without achieving any appreciable goal.
Veteran legislator Hal Harper testified that PACs funnel money into state legislative campaigns only during elections when their interests are at stake. It seems obvious to me that the perception of corruption would be lessened if legislative candidates accepted contributions from PACs on both sides of the issue. A candidate would then have a more diverse base of support to combat the perception of corruption, instead of being precluded from accepting contributions from a competing side because the candidate is “PAC’d out.”
The State’s chosen means is neither closely drawn, nor effective to achieve a diverse base of contributions. If a candidate “PACs out,” a candidate may collect PAC money “funneled” through a party, but from no or very few individuals. This leads to the same supposed corrosive effect of large amounts of special interest money in legislative elections (or the large proportion of PAC money that makes up a candidate’s contributions) which Montana seeks to eliminate. As referenced above, the memorandum by the Montana Republican legislator demonstrates that PAC money may allegedly corrupt or be perceived to corrupt candidates as well as political parties.5 Placing a limit on the amount PACs can contribute to an individual candidate, but not placing a limit on PAC contributions to a party and simultaneously increasing the amount a party can contribute to a candidate, will tend to encourage circumvention of the statute and possibly create the corruption that Montana seeks to prevent. As the majority noted, a PAC may donate to political parties without limitation. I submit that the aggregate restriction is little more than an arbitrary limitation which substantially impinges on PACs’ rights of association and free speech.
While I recognize that the aggregate limit allows candidates to return some PAC money to make room for other PAC contributions, the reality of the limitation is that candidates, such as Senator Ric Holden, are forced to reject contributions and to refrain from soliciting contributions from other PACs which would otherwise support a candidate. (See ER 511-12). While the candidate can return some money in order to receive other contributions, there is no incentive for the candidate to do so. A real risk exists that the candidate will not be able to replace the funds he returns, thereby creating even less incentive to refund PAC money already collected. (See ER 45, 795-96, 807). If the purpose of the restriction is to increase the *1106support base of candidates, why is the State making it more difficult for a candidate to receive contributions from multiple PACs?
While the majority applauds the restriction for not imposing a restraint on a candidate’s speech, I am not convinced that the associational rights of PACs have not been unnecessarily abridged. The majority’s example of this refund system demonstrates the likely encroachment. If a candidate for the senate has already accepted $100 from each of twenty different PACs, and wishes to accept a $100 contribution from another PAC, the candidate would need to return $5 to each of the other twenty PACs. The amount available to the candidate will not change regardless of whether he accepts the twenty-first PAC’s contribution. The act of refunding $5 to each of twenty PACs every time the candidate would like to receive another PAC’s contribution would cause the sheer logistics of the process to discourage the candidate from accepting that twenty-first contribution.
As a result, when a candidate rejects a contribution that he would otherwise have accepted but cannot because he has “PAC’d out” and does not or cannot refund money “to make room for the contribution,” the rejected PAC’s associational rights have been abridged.6 Although this may not be a direct result of the State’s action, an indirect abridgement of associational rights is just as obnoxious. See Buckley, 424 U.S. at 65, 96 S.Ct. 612 (the Court applies strict scrutiny even when “any deterrent effect on the exercise of First Amendment rights arises, not through direct government action, but indirectly as an unintended but inevitable result of the government’s conduct ... ”).
CONCLUSION
I do not question the legitimacy of the State’s interest to prevent corruption and perceived corruption. The State asserts that in order to prevent corruption or the perception of corruption, it must limit the amount of influence special interests exert over candidates. The State fails to provide evidence of corruption or a genuine threat of corruption by PACs as a monolithic group. Furthermore, the State asserts that the limit is closely drawn to its interest in preventing corruption by PACs; however, it increases the amount a political party may contribute to a candidate without imposing any restriction on the amount a PAC may contribute to a political party. In addition, the State seeks to prevent PAC contributions from forming a large portion of a candidate’s contributions without first distinguishing between direct monetary contributions and indirect contributions to a candidate’s campaign. What the majority fails to recognize is that influence over a candidate may come in forms other than monetary contributions. In addition, PACs may decide that they “may add more to political discourse by giving rather than spending, if the donee is able *1107to put the funds to more productive use than can the [PAC];” therefore, depriving PACs of that choice is an abridgement of associational freedoms. Shrink Missouri, 528 U.S. at 416-17, 120 S.Ct. 897 (Thomas, J., dissenting) (citing Colorado Republican, 518 U.S. at 636, 116 S.Ct. 2309 (Thomas, J., concurring in judgment and dissenting in part)); see also Federal Election Comm’n. v. Massachusetts Citizens for Life, Inc., 479 U.S. 238, 261, 107 S.Ct. 616, 93 L.Ed.2d 539 (1986). “The First Amendment mandates that we presume that speakers, not the government, know best both what they want to say and how to say it.” Shrink Missouri, 528 U.S.at 418, 120 S.Ct. 897 (Thomas, J., dissenting) (quoting Riley v. National Federation of Blind of N.C., Inc., 487 U.S. 781, 790-91, 108 S.Ct. 2667, 101 L.Ed.2d 669 (1988)).
. It is worth observing that the term "special interest” seems to be used in the pejorative *1100sense by Montana, even to the point of subtly equating it with corruption. In a democracy, every thoughtful voter and financial supporter (large or small) represents a “special interest.” It is only when a particular interest becomes a corrupting one that the state can claim an interest which justifies regulation.
. As noted by the majority, the Supreme Court’s recent decision in FEC v. Beaumont, - U.S. -, 123 S.Ct. 2200, 156 L.Ed.2d 179 (2003) reiterates many of the important principles set forth in Buckley and Shrink Missouri, including the standard for review. "[A] contribution limit involving significant interference with associational rights passes muster if it satisfies the lesser demand of being closely drawn to match a sufficiently important interest.” Beaumont, 123 S.Ct. at 2210 (internal quotation marks omitted).
. Conversely, in Daggett v. Commission on Governmental Ethics and Election Practices, 205 F.3d 445, 457 (1st Cir.2000), under the guidelines of Shrink Missouri, the First Circuit Court of Appeals relied on a poll as evidence of corruption and the appearance of corruption. Similarly, the district court in Landell v. Sorrell, 118 F.Supp.2d 459, 469, 478 (D.Vt.2000), relied on polling information to demonstrate an erosion of public confidence, noting that ''[tjypical barometers of citizen concern such as polls and media coverage” have been used by many other courts in reviewing the governmental interest in enacting contribution limits.
. The majority in Compassion in Dying v. State of Wash., 79 F.3d 790, amended by 85 F.3d 1440 (9th Cir.1996), cert. granted by Washington v. Glucksberg, 518 U.S. 1057, 117 S.Ct. 37, 135 L.Ed.2d 1128 (1996), rev’d by 521 U.S. 702, 117 S.Ct. 2258, 138 L.Ed.2d 772 (1997), relied on public opinion polls to determine the current societal attitudes as one factor in deciding whether a liberty interest in physician assisted suicide exists. That case is distinguishable. Here, the Supreme Court has already recognized an important state interest in preventing perceived corruption. See Buckley, 424 U.S. at 26-27, 96 S.Ct. 612. Rather than relying on public opinion polls to determine whether an important state interest exists to prevent perceived corruption, here the State seeks to offer public opinion polls as evidence of perceived corruption. As Judge Trott’s dissent notes, "[t]he Constitution’s explicit provisions for amendment rely on the government process, not on random sampling of public opinions. Polls are for the other branches of our government, not for the judiciary.” Compassion in Dying, 85 F.3d at 1449 (Trott, J., dissenting). Similarly here, such reliance on public opinion polls on issues of constitutional significance is improper and unnecessary.
. Montana cites as evidence of corruption or perceived corruption complaints from the 1988 election relating to alleged illegal transfers made by the national Republican Party and the Montana Republican Party. (ER 344, 830). The Supreme Court recognized that "political parties also share relevant features with many PAC’s, both having an interest in, and devoting resources to, the goal of electing candidates who will 'work to further' a particular 'political agenda,' which activity would benefit from coordination with those candidates.” Colorado Republican Fed. Campaign Comm. v. Federal Election Comm'n, 518 U.S. 604, 624, 116 S.Ct. 2309, 135 L.Ed.2d 795 (1996) (citation omitted). Thus, the aggregate PAC contribution limit is not narrowly tailored to achieve Montana’s goal of eliminating the influence of large amounts of money in Montana politics.
. Although the majority finds that PACs may contribute to a candidate's campaign through other means such as volunteering services, endorsing the candidate or independently buying advertising in support of the candidate, I do not find less corruption or less perceived corruption merely by defining what constitutes a "large portion” and instead chooses an arbitrary limit without regard to the ratio of PAC contributions as compared with other contributions. Finally, the State asserts that no associational freedoms have been unnecessarily abridged because if candidates “PAC out,” the candidate may refund some PAC money to make room for the other PAC money. In reality, the unintended result is that it limits the number of PACs who can participate in political speech without affecting the amount of PAC money involved in Montana’s legislative elections. Based on the foregoing, I conclude that the aggregate PAC limit fails the Buckley standard. Accordingly, I respectfully dissent.