Opinion by Judge SILVERMAN; Opinion dissenting in part by Judge TEILBORG.
SILVERMAN, Circuit Judge:In 1994, Montana voters passed various campaign finance reform measures contained in a ballot proposition known as Initiative 118. At issue in this case are two of the provisions contained in that initiative. The first lowers the maximum dollar amount both political action commit*1088tees and individuals may contribute to a political candidate; the second limits the aggregate dollar amount a candidate may receive from all PACs combined. Plaintiffs-appellants brought suit to invalidate some of the measures in Initiative 118, claiming they unduly burdened protected speech and associational rights. After a four-day bench trial, the district court made numerous factual findings and struck down portions of Initiative 118 not at issue here. As to the two provisions challenged on appeal, the district judge upheld them as sufficiently tailored to achieving Montana’s important interest in preventing corruption and the appearance of corruption in Montana politics.
We affirm. The district court’s factual findings are adequately supported by the record and are not clearly erroneous. Applying these facts to the analytical framework set forth in Buckley v. Valeo, 424 U.S. 1, 96 S.Ct. 612, 46 L.Ed.2d 659 (1976) and Nixon v. Shrink Missouri Gov’t PAC, 528 U.S. 377, 120 S.Ct. 897, 145 L.Ed.2d 886 (2000), we agree that the two challenged provisions do not violate the First Amendment.
I. Factual Background
In 1994, Montana voters passed Initiative 118, a campaign finance reform scheme containing, among other provisions, two sections that were subsequently enacted as MontCode Ann. (M.C.A.) §§ 18-37-216 and -218. The first provision at issue here, M.C.A. § 13-37-216, imposes limits on individual and political action committee contributions to state candidates, the amount of which varies with the office sought.
Aggregate contributions for each election in a campaign by a political committee or by an individual, other than a candidate, to a candidate are limited as follows:
(i) for candidates filed jointly for the office of governor and lieutenant governor, not to exceed $400;
(ii) for a candidate to be elected for state office in a statewide election, other than the candidates for governor and lieutenant governor, not to exceed $200;
(iii) for a candidate for any other public office, not to exceed $100.
MontCode Ann. § 13-37-216(l)(a). Because these limits apply to “each election in a campaign,” the amount an individual may contribute to a candidate doubles when the candidate participates in a contested primary. While M.C.A. § 13-37-216 lowered the amount of money that individuals and PACs can contribute to candidates, it increased the amount that political parties are permitted to contribute. Id. § 13 — 37—216(B).1
The second provision at issue in this appeal, M.C.A. § 13-37-218, limits the amount that a candidate for the state legislature may receive from all political action committees combined. It provides in pertinent part:
*1089Id. § 13-37-218. Adjusted for inflation, the PAC contribution ceiling at the time of trial was $2,000 for state senate candidates and $1,250 for state house candidates. Under M.C.A. § 13-37-218, a candidate is permitted to accept additional PAC contributions once the aggregate PAC contribution limit has been reached, provided that he returns funds to earlier PAC donors to make room for later-received contributions. It is important to note that M.C.A. § 13-37-218 does not prevent PACs from contributing to political parties, nor does it prevent PACs from spending money on independent political advertisements or otherwise engaging in political speech. Section 13-37-218 merely limits how much PACs as a group can donate to any one candidate.
The Montana Right to Life Association, Montana Right to Life Political Action Committee, and Julie Daffln, President of the Montana Right to Life Association (collectively, “MRLA”) have all made or attempted to make contributions to Montana legislative candidates. MRLA brought this lawsuit in 1996, challenging six of the campaign finance reform measures contained in Initiative 1-118.
The district court granted partial summary judgment to MRLA, declaring four of the initiative’s provisions unconstitutional, but left for trial the constitutionality of M.C.A. §§ 13-37-216 and -218. After a four-day bench trial, the district court issued findings of fact and conclusions of law, upholding the two provisions at issue here. The district court relied in part on the testimony of Jonathon Motl, the drafter of the ballot initiative, that 1-118 affects only the largest contributions to the various offices. The judge found that the limits imposed by M.C.A. § 13-37-216 “were in the upper 10% of contributions for the particular offices.” That is, nine out of ten donations to political candidates were unaffected by this measure.
The district court also found that M.C.A. § 13-37-218, the aggregate PAC contribution limit provision, had the effect of limiting the amount the average candidate received from PACs to about 29% of all contributions received. The court found that, at the time of trial, state house candidates continued to raise an average of $4,464.87, and state senate candidates continued to raise an average of $6,869.04, despite the limits imposed by M.C.A. § 13-37-218. The evidence further showed that the cost of a House race in Montana was between $3,000 and $7,000, and a Senate race between $6,000 and $9,000. The district court thus found that MRLA was unable to demonstrate that the limits imposed left candidates with insufficient funds to run an effective campaign: “[0]utside of bald, conclusory allegations that their campaigns would have been more ‘effective’ had they been able to raise more money, none of the witnesses offered any specifics as to why their campaigns were not effective.” It further found that “there is no indication that the contribution limitations imposed would have any dramatically adverse effect on the funding of campaigns and political associations .... ”
Applying the standards announced by the Supreme Court in Shrink Missouri, *1090the district court ultimately ruled that the State of Montana’s political contribution limits were “closely drawn to match the constitutionally sufficient interest in preventing campaign corruption and the appearance thereof.” The limits “are not so radical in effect as to render political association ineffective, drive the sound of a candidate’s voice below the level of notice, and render contributions pointless.” MRLA appeals this ruling.
II. Standard of Review
We review the constitutionality of state statutes de novo. California Democratic Party v. Jones, 169 F.3d 646, 647 (9th Cir.1999), rev’d on other grounds, 530 U.S. 567, 120 S.Ct. 2402, 147 L.Ed.2d 502 (2000); Arizona Right to Life Political Action Comm. v. Bayless, 320 F.3d 1002, 1007 (9th Cir.2003). We review the district court’s findings of fact for clear error. Montana Chamber of Commerce v. Argenbright, 226 F.3d 1049, 1054 (9th Cir.2000) (reviewing a district court’s findings of fact in a campaign contribution limit case under the “clearly erroneous” standard without discussion); Service Employees Int’l Union v. Fair Political Practices Comm’n, 955 F.2d 1312, 1316 (9th Cir.1992) (same). The district court’s application of the law to those facts is reviewed de novo. Bose Corp. v. Consumers Union of the United States, 466 U.S. 485, 499, 104 S.Ct. 1949, 80 L.Ed.2d 502 (1984).
III. Supreme Court Decisions Regarding the Constitutionality of Campaign Finance Restrictions
The starting place in the analysis of the constitutionality of campaign finance reform legislation is Buckley v. Valeo, 424 U.S. 1, 96 S.Ct. 612, 46 L.Ed.2d 659 (1976). Buckley involved a challenge to the Federal Election Campaign Act. The Act (1) limited individual contributions to any single candidate to $1,000 per election, with an overall annual limitation of $25,000 by any contributor; (2) limited independent expenditures by individuals and groups relative to a clearly identified candidate to $1,000 per year; (3) subjected campaign spending by candidates and political parties to prescribed limits; and (4) required public disclosure of all contributions and expenditures above defined limits.
The Buckley Court held that although the provisions limiting contributions to candidates were constitutional, the provisions limiting expenditures by candidates were invalid, violating candidates’ freedom of speech. Id. at 20-21, 96 S.Ct. 612. With respect to the contribution limitations, the Court made three important observations. First, regarding a contributor’s right to free speech, the effect of the contribution limitation was minimal:
A limitation upon the amount that any one person or group may contribute to a candidate or political committee entails only a marginal restriction upon the contributor’s ability to engage in free communication. A contribution serves as a general expression of support for the candidate and his views, but does not communicate the underlying basis for the support. The quantity of communication by the contributor does not increase perceptibly with the size of his contribution, since the expression rests solely on the undifferentiated, symbolic act of contributing.... 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, for it permits the symbolic expression of support evidenced by a contribution but does not in any way infringe the contributor’s freedom to discuss candidates and issues.
Id. (emphasis added).
Second, regarding the effect on a candidate’s free speech rights, the Buckley *1091Court held that contribution limits are constitutional as long as they do not prevent candidates from “amassing the resources necessary for effective advocacy.” Id. at 21, 96 S.Ct. 612. If a candidate is merely required “to raise funds from a greater number of persons and to compel people who would otherwise contribute amounts greater than the statutory limits to expend such funds on direct political expression,” the candidate’s freedom of speech is not impugned by limits on contributions. Id. at 21-22, 96 S.Ct. 612.
Finally, the Buckley Court observed that the main concern raised by contribution limitations was whether they interfered with a contributor’s right of association. Id. at 24-25, 96 S.Ct. 612. “Making a contribution, like joining a political party, serves to affiliate a person with a candidate.” Id. at 22, 96 S.Ct. 612. Recognizing that freedom of political association is a “basic constitutional freedom,” the Court held that restrictions on that right are subject to the “closest scrutiny.” Id. at 25, 96 S.Ct. 612. The Court was careful to note, however, that “neither the right to associate nor the right to participate in political activities is absolute.... Even a significant interference may be sustained if the State demonstrates a sufficiently important interest and employs a means closely drawn to avoid unnecessary abridgment of associational freedoms.” Id. at 25, 96 S.Ct. 612 (citations and internal quotation marks omitted).
The Supreme Court reaffirmed the principles announced in Buckley when it upheld a state campaign contribution limitation in Nixon v. Shrink Missouri Gov’t PAC, 528 U.S. 377, 120 S.Ct. 897, 145 L.Ed.2d 886 (2000).2 Shrink Missouri involved a Missouri statute that imposed contribution limits ranging from $275 to $1,075, depending on the office or size of the candidate’s constituency and accounting for inflation. Id. at 382, 120 S.Ct. 897. The Shrink Missouri Government PAC and an unsuccessful candidate for state auditor sued to enjoin enforcement of the statute, claiming that it violated the First and Fourteenth Amendments.
Upholding the statute as constitutional, the Shrink Missouri Court emphasized Buckley’s holding that “a contribution limit involving a ‘significant interference’ with associational rights could survive if the Government demonstrated that the regulation was ‘closely drawn’ to match a ‘sufficiently important interest,’ though the dollar amount of the limit need not be ‘fine-tuned.’ ” Id. (citations omitted). Shrink Missouri also stressed that courts considering contribution limits, as opposed to expenditure limits, need not be overly concerned with the precise standard of scrutiny to be applied because, in general, “limiting contributions [leaves] communications significantly unimpaired,” and “contribution limits ... more readily clear the hurdles before them” than would analogous expenditure limits. Id. at 387-88, 120 S.Ct. 897.
Shrink Missouri recognized that Buckley “specifically rejected the contention that $1,000, or any other amount, was a constitutional minimum below which legislatures could not regulate.” Id. at 397, 120 S.Ct. 897. Rather, the Court said that the outer limits of constitutional contribution limitations are defined by whether the limitation is so low as to impede a candidate’s ability to “amass the resources necessary for effective advocacy.” Id. at 397, 120 S.Ct. 897. The question to be asked in *1092evaluating laws that limit campaign contributions, then, is whether “the contribution limitation is so radical in effect as to render political association ineffective, drive the sound of a candidate’s voice below the level of notice, and render contributions pointless.” Id.
Recently, in FEC v. Beaumont, — U.S. -, 123 S.Ct. 2200, 156 L.Ed.2d 179 (2003), the Supreme Court reaffirmed the principles enunciated in Buckley. The Supreme Court noted that “[g]oing back to Buckley ... restrictions on political contributions have been treated as merely ‘marginal’ speech restrictions subject to relatively complaisant review under the First Amendment, because contributions he closer to the edges than to the core of political expression.” Id. at 2210.
The bottom line is this: After Buckley and Shrink Missouri, state campaign contribution limits will be upheld if (1) there is adequate evidence that the limitation furthers a sufficiently important state interest, and (2) if the limits are “closely drawn” — i.e., if they (a) focus narrowly on the state’s interest, (b) leave the contributor free to affiliate with a candidate, and (c) allow the candidate to amass sufficient resources to wage an effective campaign. With these principles in mind, we now turn to whether Montana’s campaign limits pass muster under Buckley and Shrink Missouri.
IV. M.C.A. § 13-37-216, limiting individual and PAC campaign contributions, is constitutional.
A. The State of Montana presented sufficient evidence of its asserted interest in avoiding corruption or the appearance of corruption.
Montana asserts that the campaign contribution limitation on individuals and PACs is necessary to avoid corruption or the appearance of corruption in Montana politics. MRLA does not dispute that this interest is sufficient to justify campaign contribution limits. Rather, it abgues that the limits imposed are unnecessarily stringent and there is no evidence that restricting contributions to such small amounts is needed to combat corruption.
This, however, is not the appropriate inquiry. The correct focus under Shrink Missouri is whether the state has presented sufficient evidence of a valid interest, not whether it has justified a particular dollar amount. The latter inquiry, if ever appropriate, occurs in the second part of our analysis, in examining whether the restriction is “closely drawn.” See, e.g., California Prolife Council Political Action Comm. v. Scully, 989 F.Supp. 1282, 1292 (E.D.Cal.1998) (“[A]s a general matter, the court will not second guess a legislative determination as to where the line for contribution limits shall be drawn.”). With respect to whether Montana has presented sufficient evidence of corruption or the appearance of corruption, we agree with the district court that it has.
A state’s interest in preventing corruption or the appearance of corruption is not confined to instances of bribery of public officials, but extends “to the broader threat from politicians too compliant with the wishes of large contributors.” Shrink Missouri, 528 U.S. at 389, 120 S.Ct. 897. With respect to the quantum of evidence necessary to justify this interest, the Supreme Court has required only that the perceived threat not be “illusory,” Buckley, 424 U.S. at 27, 96 S.Ct. 612, or “mere conjecture,” Shrink Missouri, 528 U.S. at 392, 120 S.Ct. 897. The amount of evidence needed will thus “vary up or down with the novelty and plausibility of the justification raised.” Id. at 391, 120 S.Ct. 897. The Shrink Missouri Court found sufficient evidence of the potential of contributions to corrupt simply in a state senator’s statement that contributions had the *1093“real potential to buy votes,” a smattering of newspaper articles reporting large contributions, and the fact that 74% of Missouri voters determined that contribution limits were necessary. Id.
The evidence presented by the State of Montana in this case is sufficient to justify the contribution limits imposed, and indeed carries more weight than that presented in Shrink Missouri. The record contains the testimony of a 30-year veteran of the Montana legislature who stated that special interests funnel more money into campaigns when particular issues approach a vote “because it gets results.” The state also pointed to a 1981 incident in which a Republican state senator dispatched a letter to his colleagues urging them to vote for passage of a bill favoring variable annual annuities to ensure that a highly disproportionate share of PAC contributions from the insurance industry continued to flow to the Republican party. The letter read in part:
Please destroy this letter after reading. Why? Because the Life Underwriters Association in Montana is one of the larger Political Action Committees in the state, and I don’t want the demos to know about it! In the last election they gave $8000 to state candidates.... Of this $8,000 — Republicans got $7000— you probably got something from them. This bill is important to the underwriters and I have been able to keep the contributions coming our way. In 1983, the PAC will be $15,000. Let’s keep it in our camp.
The Montana press published the contents of the senator’s letter. Although the author of the letter was ultimately cleared of wrongdoing, the letter and attendant publicity spawned five separate investigations.
A 1982 poll indicated that 78.3% of Montana voters believe money is synonymous with power. Another 69% of Montanans say that elected officials give special treatment to individuals and businesses that make large contributions. The district court found that MRLA had offered no evidence that Montana voter suspicion or perception was to the contrary. Moreover, MRLA is incorrect to suggest that our reliance on such evidence is impermissible. In Shrink Missouri, the Court relied, in part, on similar evidence: the result of a referendum election relating to contribution limits. See Shrink Missouri, 528 U.S. at 394, 120 S.Ct. 897. Cf. Atkins v. Virginia, 536 U.S. 304, 122 S.Ct. 2242, 2248-50, 153 L.Ed.2d 335 (2002) (relying on public consensus as evidence of what constitutes cruel and unusual punishment in a death penalty case). Taken together, the evidence presented below suffices under Shrink Missouri to establish Montana’s interest in avoiding corruption or the appearance of corruption. The state’s interest is neither illusory or conjectural.
B. M.C.A. § 13-37-216 is “closely drawn” to avoid unnecessary abridgment of associational freedoms.
MRLA also challenges M.C.A. § 13-37-216 as insufficiently tailored to the state’s interest in preventing corruption, arguing that it prevents candidates from amassing needed resources, discriminates against challengers, and unconstitutionally prohibits both small and large contributions. We disagree.
A campaign contribution limitation is “closely drawn” if it
focus[es] on the narrow aspect of political association where the actuality and potential for corruption have been identified — while leaving persons free to engage in independent political expression, to associate actively through volunteering their services, and to assist in a limited but nonetheless substantial ex*1094tent in supporting the candidates and committees with financial resources.
Buckley, 424 U.S. at 28, 96 S.Ct. 612. In examining whether a contribution limitation is sufficiently tailored to a state’s asserted interest, the focus is as much on those aspects of associational freedom unaffected by the law as the limitations that are imposed. We are mindful that the dollar amounts employed to prevent corruption should be upheld unless they are “so radical in effect as to render political association ineffective, drive the sound of a candidate’s voice beyond the level of notice, and render contributions pointless.” Shrink Missouri, 528 U.S. at 397, 120 S.Ct. 897. In making this determination, we look at all dollars likely to be forthcoming in a campaign, rather than the isolated contribution, id., and we also consider factors such as whether the candidate can look elsewhere for money, Buckley, 424 U.S. at 21-22, 96 S.Ct. 612, the percentage of contributions that are affected, Daggett, 205 F.3d at 461, the total cost of a campaign, id., and how much money each candidate would lose, id.
We agree with the district court that the state’s contribution limits are closely drawn to further its interest in preventing corruption and the appearance of corruption. The district court found that the contribution limits affect only the top 10% of contributions, and that the percentage affected includes the largest contributions. As the testimony of the statute’s drafter, Jonathon Motl, makes clear, this finding was not clearly erroneous. MRLA’s contention that M.C.A. § 13-37-216 unconstitutionally prohibits both small and large contributions is thus without merit.
In addition, M.C.A. § 13-37-216, while decreasing PAC and individual contributions, simultaneously increased the amount of money political parties may contribute to a candidate, almost doubling the amount that may be contributed in some races. The statute also did not limit the amount a candidate may give to himself, or the number of individuals from whom he can seek contributions. Candidates can therefore look elsewhere for forms of funding unaffected by the limitations imposed by M.C.A. § 13-37-216. Moreover, the statute in no way prevents PACs from affiliating with their chosen candidates in ways other than direct contributions, such as donating money to a candidate’s political party, volunteering individual members’ services, sending direct mail to their supporters, or taking out independent newspaper, radio, or television ads to convey their support.
The evidence before the district court showed that the State of Montana remains one of the least expensive states in the nation in which to run a political campaign. Montana’s 100 house districts average only 7,991 people, its 50 senate districts 15,981 people. Legislative candidates in Montana campaign primarily door-to-door, and only occasionally advertise on radio and television. It is undisputed that the total money contributed to political campaigns in the State of Montana has decreased considerably since the challenged measures went into effect. The parties agree that, of the money raised in the 1992 legislative election, before M.C.A. § 13-37-216 was enacted, 24% to 30% came from contributions that would now violate the new limits. That alone, however, does not make the contribution limits unconstitutional. Indeed, the Shrink Missouri Court upheld contributions limits despite a decrease of more than 50% in total spending in Missouri elections, nearly twice the decrease present here. See Shrink Missouri, 528 U.S. at 426 n. 10, 120 S.Ct. 897 (Thomas, J., dissenting). The district court found that the average amount raised by a Montana house candidate in 1998, with the challenged limits in effect, was $4,464.87, a *1095figure well within the range of money needed to run an effective house campaign. The same is true for the $6,869.00 average raised by state senate candidates. We cannot agree with MRLA that the challenged limits have impeded candidates’ campaigns to such an extent that speech and associational rights have been imper-missibly abridged.
As the district court found, Montana candidates remain able to mount effective campaigns, a primary concern in our inquiry. MRLA, however, presented the testimony of three candidates who claimed that the new limits preclude effective campaigning. We agree with the district court that this evidence is unpersuasive. Two of the witnesses raised more money after the enactment of M.C.A. § 13-37-216 than before, two were successfully elected to their positions, and the one losing candidate admitted that his absence during a pivotal campaign period prevented him from raising sufficient funds to win. Another MRLA witness, a campaign manager for a successfully elected Montana legislator, acknowledged that her candidate won with a $70,000.00 surplus of funds. We fully agree with the district court’s conclusion that, apart from “bald, conclusory allegations that their campaigns would have been more ‘effective’ had they been able to raise more money, none of the witnesses offered any specifics as to why their campaigns were not effective.”
It is true that the contribution limits imposed by M.C.A. § 13-37-216 are some of the lowest in the country. This is unsurprising in light of the fact that Montana is one of the least expensive states in the nation in which to mount a political campaign. As long as the limits are otherwise constitutional, it is not the prerogative of the courts to fine-tune the dollar amounts of those limits. See, e.g., Shrink Missouri, 528 U.S. at 388, 120 S.Ct. 897 (“[T]he dollar amount of the limit need not be fine tuned.”) (internal quotations omitted).
MRLA also claims that the state’s argument, that a candidate can campaign through less expensive means, is unpersuasive if the candidate is unable to mount the same type of campaign he could have run without the limit. This ignores the point emphasized in both Buckley and Shrink Missouri that a limit on what others can give a candidate is fundamentally different from a limit on what a candidate can spend. Limitations on candidates’ expenditures are viewed as direct restrictions on speech, while contribution limits are only rarely seen as restrictions on donors’ First Amendment rights. Under the standards articulated for contribution limits in Buckley and Shrink Missouri, MRLA cannot argue that Montana’s contribution limits impermissibly alter a candidate’s message, or result in a different kind of campaign as compared to when the limits did not exist. Rather, MRLA must show that limiting donations prevents candidates from amassing the resources necessary for effective advocacy, making a donee candidate’s campaign to be not merely different but ineffective. This MRLA has not done.
Finally, MRLA, noting that 40-50% of legislative seats went uncontested in the 1998 campaign, argues that the contribution limits unconstitutionally discriminate against challengers. This assertion fails for two reasons. First, while imposing contribution limits, M.C.A. § 13-37-216 also contains a provision preventing incumbents from using excess funds from one campaign in future campaigns. Such a provision keeps incumbents from building campaign war chests and gaining a fund-raising head start over challengers. The record shows that the average gap between the total amount of money raised by incumbents and challengers for all legislative races was only $65.00 per race. Sec*1096ond, Buckley squarely held that, without a record of “invidious discrimination against challengers as a class,” there is “no support for the proposition that an incumbent’s advantages [are] leveraged into something significantly more powerful by contribution limitations applicable to all candidates, whether veterans or upstarts.” Shrink Missouri, 528 U.S. at 389 n. 4, 120 S.Ct. 897. Accordingly, MRLA’s argument is without merit.
Because individuals and PACs remain “free to engage in independent political expression, to associate actively through volunteering their services, and to assist in a limited but nonetheless substantial extent in supporting the candidates and committees with financial resources,” Buckley, 424 U.S. at 28, 96 S.Ct. 612, we agree with the district court that M.C.A. § 13-37-216 is constitutional.
V. M.C.A. § 13-37-218, imposing an aggregate limit on PAC contributions to state legislative candidates, is constitutional.
We now turn to the aggregate limit on PAC contributions. M.C.A. § 13-37-218 provides:
A candidate for the state senate may receive no more than $1,000 [now increased to $2,000] in total combined monetary contributions from all political committees contributing to the candidate’s campaign, and a candidate for the state house of representatives may receive no more than $600 [now $1,250] in total combined monetary contributions from all political committees contributing to the candidate’s campaign. The limitations in this section must be multiplied by the inflation factor [defined elsewhere].
MRLA argues that the statutory limits on what a candidate may receive from all PACs unconstitutionally discriminates against PACs, unconstitutionally prohibits certain contributions entirely, impermissi-bly functions as a candidate spending limit, and is not tailored to any legitimate state interest.
A. The aggregate PAC limit is justified by a sufficiently important interest and does not unconstitutionally discriminate against PACs.
MRLA argues that M.C.A. § 13-37-218 unconstitutionally discriminates against PACs as opposed to individual donors. However, the differential treatment of PACs is constitutionally permissible where, as here, that treatment is necessary to serve a substantial government interest. See Police Dept, of Chicago v. Mosley, 408 U.S. 92, 95, 92 S.Ct. 2286, 33 L.Ed.2d 212 (1972) (“The crucial question is whether there is an appropriate government interest suitably furthered by the differential treatment.”). The State of Montana contends that the aggregate PAC limit is justified by the state’s concern over the corrupting influence of PAC money on campaigns. If the record demonstrates that the danger of corruption, or the appearance of such a danger, is greater when dealing with PAC money as opposed to other contributions, then the state’s justification is constitutionally sufficient. See, e.g., Austin v. Michigan Chamber of Comm., 494 U.S. 652, 658-60, 110 S.Ct. 1391, 108 L.Ed.2d 652 (1990) (prevention of corruption justification sufficient to justify differential treatment of corporations).
The district court found that the aggregate PAC limits are “essential” to preventing undue influence and the appearance of undue influence by special interest groups. As noted above, this finding is supported by a quantum of evidence that more than exceeds that found sufficient in Shrink Missouri. The most damning evidence was the letter from a state senator urging legislators to vote for a bill in order to *1097keep insurance industry PAC money in the Republican camp. It is true that the investigations spawned by the letter did not result in criminal charges, but that is not the test. The voters of Montana were entitled to view the widely-publicized letter as unwholesome and indicative of the corrosive influence of PAC money on the legislative process, as they apparently did.
This view was echoed by veteran legislator Hal Harper, who testified that, in general, PACs funnel money into state legislative campaigns only when their interests are at stake in order to “get results.” This testimony, like that of Missouri Senator Wayne Goodein, who stated that large contributions have “the real potential to buy votes,” Shrink Missouri, 528 U.S. at 398, 120 S.Ct. 897, speaks not to particular PACs but to the potentially corrosive effects of special interest groups in general, and the corresponding justification for government actions to counteract such effects.
Many courts have recognized that the danger of corruption in the political system is greater with respect to PAC contributions than it is for individuals. See, e.g., Kentucky Right to Life v. Terry, 108 F.3d 637 (6th Cir.1997) (upholding a law to combat corruption by placing greater restrictions upon direct corporate and PAC contributions to political candidates, and lesser restrictions upon individual contributions); Landell v. Sorrell, 118 F.Supp.2d 459, 489 (D.Vt.2000) (“[T]he anti-corruption rationale ... is arguably even stronger when applied to PAC contributions .... As their name suggests, PACs exist in order to affect certain political action. The likelihood of actual quid pro quo arrangements between PACs and candidates is high.”). One court has even explicitly held that, due in part to the disproportionate influence of special interests on a candidate’s campaign, aggregate PAC limits are constitutionally permissible. Gard v. Wisconsin State Elections Bd., 156 Wis.2d 28, 456 N.W.2d 809, 820 (1990).
MRLA argues that the State’s interest in preventing corruption and the appearance of corruption does not justify the aggregate PAC limit because M.C.A. § 13-37-218 does not differentiate on the basis of the size of any individual PAC’s contribution. This ignores the fact that the size of an individual PAC’s contribution is already limited by M.C.A. § 13-37-216. The two provisions before us work hand-in-glove to avoid corruption and the appearance of the same by reducing the impact of PAC money on Montana’s elections, thereby encouraging candidates to have a diverse base of support. The district court found that the aggregate PAC limit was justified in part because otherwise PACs could easily evade the individual contribution limits by contributing the statutory maximum through a multitude of individual committees. Like the district court, we find this justification persuasive. Accordingly, we hold that the aggregate PAC limit is justified by a sufficiently important state interest and does not unconstitutionally discriminate against PACs.
B. The aggregate PAC limit is closely drawn to serve the state’s anti-corruption purpose.
MRLA argues that M.C.A. § 13-37-218 is not sufficiently tailored to the state’s anti-corruption interest, preventing some PACs from contributing anything at all. The argument goes like this: If a candidate for, say, state senate has already accepted $2,000 in PAC money, he has “PAC’d out,” and other PACs are now prevented from contributing and can no longer express their support for the candidate. The flaw in this argument is that it fails to recognize that, under the Montana scheme, a candidate can return some money from one PAC to make room for other *1098PAC money. For example, in our hypothetical, suppose the senate candidate received twenty contributions of $100 each from twenty different PACs, thus reaching the aggregate $2,000 limit. If a twenty-first PAC wished to make a $100 contribution, and if the candidate wished to accept it, the candidate could refund $100 (for example, by returning five dollars to each of the other twenty PACs) to make room for the new contribution. What matters is that so long as a candidate wants a PAC involved in funding his campaign, Montana’s law does not infringe on the PACs’ associational freedoms. A candidate is free to manage his PAC contributions so as to be able to accept contributions from an unlimited number of PACs, allowing them to show their support for candidates they back and participate in the electoral process.
Furthermore, M.C.A. § 13-37-218 does not prevent PACs from otherwise affiliating with a candidate in ways other than direct contributions. PACs can continue, for example, to volunteer services to a candidate’s campaign, to endorse a candidate, to independently buy advertising in support of a candidate, etc.
It is important to recognize that the aggregate PAC limit does not directly affect a candidate’s speech. A candidate is free to obtain additional money from other sources, including an unlimited number of individual donors, the candidate’s own funds, and the candidate’s political party. Moreover, a PAC may still donate to political parties without limitation. As the district court found, even with the aggregate PAC limit imposed, PAC contributions comprised nearly a third of candidates’ total campaign funds in the last election. Clearly, PACs still play a significant role in Montana political campaigns, and MRLA’s attempts to characterize M.C.A. § 13-37-218 as stifling PACs’ voices in Montana elections are unconvincing. The limits imposed by Montana voters are not “so radical in effect as to render political association ineffective, drive the sound of a candidate’s voice beyond the level of notice, and render contributions pointless.” Shrink Missouri, 528 U.S. at 387-88, 397, 120 S.Ct. 897. We therefore agree with the district court that M.C.A. § 13-37-218 is constitutional.
VI. Conclusion
The voters of Montana are entitled to considerable deference when it comes to campaign finance reform initiatives designed to preserve the integrity of their electoral process. See FEC v. Beaumont, — U.S.-,-, 123 S.Ct. 2200, 2208, 2209, 156 L.Ed.2d 179 (2003). Our analysis in reviewing such initiatives is highly fact-intensive and relies heavily on the factual findings made by the district court in the wake of a four-day bench trial, findings that have ample support in the record and are not clearly erroneous. Applying these facts to the analytical framework set forth in Buckley and Missouri Shrink, we hold that Montana’s interest in purging corruption and the appearance of corruption from its electoral system is sufficiently important to withstand constitutional scrutiny, and that M.C.A. §§ 13-37-216 and -218 are closely tailored to achieving those ends. We therefore affirm the district court and hold that these statutes are constitutional and do not violate the First Amendment.
AFFIRMED.
. M.C.A. § 13-37-216(3) reads: "All political committees except those of political party organizations are subject to the provisions of subsections (1) and (2). For purposes of this subsection, 'political party organization' means any political organization that was represented on the official ballot at the most recent gubernatorial election. Political party organizations may form political committees that are subject to the following aggregate limitations from all political party committees: (a) for candidates filed jointly for the offices of governor and lieutenant governor, not to exceed $15,000; (b) for a candidate to be elected for state office in a statewide election, other than candidates for governor or lieutenant governor, not to exceed $5,000; (c) for a candidate for public service commissioner, not to exceed $2,000; (d) for a candidate for the state senate, not to exceed $800; (e) for a candidate for any other public office, not to exceed $500.”
. MRLA's reliance on VanNatta v. Keisling, 151 F.3d 1215 (9th Cir.1998); Service Employees Int’l Union, 955 F.2d at 1312; and other Ninth Circuit cases interpreting Buckley fails to recognize the impact of the Supreme Court’s superceding decision in Shrink Missouri.