Volume 76

Number 1

Spring 1998

RAISING A NEW FIRST AMENDMENT HURDLE FOR CAMPAIGN FINANCE "REFORM"

Cite as 76 Wash. U. L.Q. 217

D. BRUCE LA PIERRE^[*]

INTRODUCTION

THE 1994 MISSOURI CAMPAIGN FINANCE REFORM LEGISLATION
BEYOND BUCKLEY-ESTABLISHING A NEW, MORE RIGOROUS FIRST AMENDMENT STANDARD FOR CAMPAIGN FINANCE REGULATION
LESSONS FOR WOULD-BE REFORMERS
1. The New First Amendment Hurdle
2. Buckley-Which Way Will The Tree Fall?

INTRODUCTION

Proposals to regulate campaign contributions and candidates' spending invariably fly the banner of campaign finance "reform." The reformers, however, frequently have little or no evidence that particular campaign practices cause any real harm. Instead, they simply posit the existence of the disease-the corrosive effects of money on the political process-and assume that restrictions on the use of money will provide the cure. In Missouri, for example, both the legislature and the voters enacted laws in 1994 that set limits on political contributions to candidates and on candidates' campaign expenditures. These laws imposed substantial burdens on political speech and association, but they did little more than pander to public perceptions about the amorphous evils of "big money." The legislature and the voters had no evidence that contributions or expenditures in the prohibited amounts caused any identified harm, and any cure for problems in the State's elections was entirely serendipitous. The frank comment of a Missouri legislator that "[p]erception is more important than what's real" captures the substance of many measures that masquerade as campaign finance "reform."^[1]

The Supreme Court, unfortunately, left the door open in Buckley v. Valeo^[2] to campaign finance reform measures, like the 1994 Missouri laws, that are grounded on little more than speculation and the cynical assumption that money necessarily and inherently corrupts the political process. Although the Buckley Court held that campaign finance measures are subject to strict scrutiny, it accepted speculation about corruption and the appearance of corruption as a justification for regulating campaign contributions.^[3] Lawyers for the American Civil Liberties Union of Eastern Missouri ("ACLU/EM") discovered a means to close the door left open in Buckley and to provide greater First Amendment protection for political speech. They persuaded the courts in the Eighth Circuit to supplement Buckley and to impose a duty on the State to demonstrate that campaign finance regulations address a "real harm."^[4] Although the Supreme Court did not review these judgments, it endorsed the Eighth Circuit's new, more demanding First Amendment standard for campaign finance laws.^[5]

I. THE 1994 MISSOURI CAMPAIGN FINANCE REFORM LEGISLATION

In 1994, Missouri limited candidates' total campaign expenditures; it prohibited candidates from carrying over more than $1000, $2000, or $3000 of campaign funds from one election to another; and it prohibited candidates from spending more than $100, $200, or $300 of their personal resources to run for state and local office. Missouri also restricted campaign contributions. The State limited contributions to candidates for state and local office on a sliding scale from $300 to $200 to $100 per election cycle, and it prohibited elected officials holding certain offices, as well as candidates for these offices, from accepting contributions during regular sessions of the state legislature.

These campaign finance regulations were the product of two sets of amendments to Missouri's Campaign Finance Disclosure Law.^[6] In July 1994, the Missouri legislature enacted Senate Bill 650 and limited campaign contributions and expenditures.^[7] On November 8, 1994, the Missouri electorate approved Proposition A, a ballot initiative that also established campaign finance regulations.^[8] The Missouri Attorney General ruled that Proposition A, which was to become effective immediately, superseded Senate Bill 650 to the extent that its provisions were more restrictive and that, otherwise, Senate Bill 650 would become effective on January 1, 1995.^[9]

A. Limits on Candidates' Political Expenditures

Senate Bill 650 limited the total amount that candidates could spend to run for state and local office.^[10] Candidates had a duty to file an affidavit indicating whether they intended to comply with expenditure limits that ranged from $30,000 to $1,500,000.^[11] The legislature imposed two penalties on candidates who rejected these limits. First, they could accept contributions only from individuals, but opposing candidates-who agreed to abide by the expenditure limits-could accept contributions from political parties, political action committees, corporations, and labor unions, as well as from individuals.^[12] Second, candidates who rejected the spending limits had a duty to file daily reports of contributions and expenditures with the Missouri Ethics Commission after their expenditures exceeded the prescribed limit, but candidates who accepted the expenditure limits had no duty to file daily reports.^[13] Candidates who rejected the spending limits were subject to criminal sanctions for accepting contributions from sources other than individuals, and for failing to file daily disclosure reports.^[14]

In addition to Senate Bill 650's limits on candidates' total campaign expenditures, Proposition A established two other spending limits. Proposition A prohibited candidates for various elective offices in Missouri from spending more than $100, $200, or $300 of their own funds to support their campaigns.^[15] Proposition A also regulated the timing of campaign expenditures and restricted the amount of campaign funds that could be spent in subsequent elections. It provided that every candidate, within ninety days of an election, "shall" either turn over to Missouri Ethics Commission or return to the contributors "any balance of campaign funds in excess of expenses incurred for the campaign, except for an amount no greater than ten times the individual contribution limit" applicable to that office.^[16] Given individual contribution limits of $100, $200, or $300,^[17] candidates could not retain more than $1000, $2000, or $3000 of the funds raised in one campaign and could not carry these funds over for use in a subsequent campaign to express their political views.^[18]

B. Campaign Contribution Limits

Senate Bill 650 limited campaign contributions to candidates for office in Missouri on a sliding scale ranging from $1000 to $500 to $250.^[19] It permitted contributions of up to $1000 for governor and other statewide offices, as well as for candidates in districts with a population of at least 250,000;^[20] it permitted $500 contributions for candidates for state senate and candidates for any office in electoral districts with a population between 100,000 and 250,000;^[21] and it permitted contributions of $250 for candidates for state representative or for offices in districts with a population of under 100,000.^[22] Senate Bill 650 provided that these contribution limits "shall be increased" to take inflation into account.^[23]

Although the contribution limits of Senate Bill 650 were set to go into effect on January 1, 1995, they were superseded under a ruling of the Missouri Attorney General by the more restrictive contribution limits of Proposition A.^[24] Proposition A limited campaign contributions to candidates for office in Missouri on a sliding scale ranging from $300 to $200 to $100 per election cycle,^[25] and it drastically reduced the contribution limits set by Senate Bill 650 in two ways. First, it reduced the dollar amount of permissible contributions by as much as seventy percent. It reduced the contribution limit for governor and other statewide offices from $1000 to $300; it reduced the contribution limit for districts with a population over 100,000 from $500 to $200; and it reduced the limit for districts under 100,000 from $250 to $100.^[26] Second, Senate Bill 650 permitted contributions up to its limits in both primary and general elections, but Proposition A established a single cumulative limit on all contributions made in a primary election and the succeeding general election.^[27] Thus, for example, under Proposition A, a contributor could have given a gubernatorial candidate only $300 over the primary and general election combined. Under Senate Bill 650, however, the contributor could have given a gubernatorial candidate $1000 in the primary election and an additional $1000 in the general election, for a combined total of $2000.

C. Limits on Contributions During Legislative Sessions

Although both Senate Bill 650 and Proposition A set contribution limits, Proposition A did not regulate the timing of political contributions. Senate Bill 650, however, prohibited "statewide elected official[s]" and members of the general assembly, and candidates for these offices, from accepting campaign contributions "during any regular session of the general assembly."^[28]

II. BEYOND BUCKLEY--ESTABLISHING A NEW, MORE RIGOROUS FIRST AMENDMENT STANDARD FOR CAMPAIGN FINANCE REGULATION

Prospective candidates and potential contributors successfully challenged the 1994 amendments in three cases. In Shrink Missouri Government PAC v. Maupin (Shrink I),^[29] the Court of Appeals for the Eighth Circuit held that the Missouri limits on candidates' political expenditures were unconstitutional.^[30] In Carver v. Nixon,^[31] a district court upheld the $100 to $300 contribution limits,^[32] but the court of appeals held that these limits on political speech violated the First Amendment.^[33] Finally, in Shrink Missouri Government PAC v. Maupin (Shrink II),^[34] a district court held that the prohibition on campaign contributions during legislative sessions violated the First Amendment.^[35] The ACLU/EM represented the plaintiffs in Shrink I, and it supported the plaintiff in Carver with an amicus brief in the court of appeals.^[36]

All three cases in large part called for a relatively straightforward application of the First Amendment standards established in 1976 by the Supreme Court in Buckley v. Valeo.^[37] Nonetheless, the ACLU/EM persuaded the Court of Appeals for the Eighth Circuit to impose a new, more rigorous First Amendment standard on the states' power to regulate political expenditures and contributions.^[38] The court of appeals supplemented Buckley's strict scrutiny standard with a requirement that the state "must do more than simply `posit the existence of the disease sought to be cured.'"^[39] The state "must demonstrate that the recited harms are real, . . . and that the regulation will in fact alleviate these harms in a direct and material way."^[40]

A. The Buckley Framework--Limited Protection of Political Speech

In Buckley v. Valeo, the Supreme Court set the First Amendment measure of campaign finance regulation. Although the Court applied strict scrutiny to both expenditure limits and contribution limits, it drew a fundamental distinction between governmental power to regulate political expenditures and governmental power to regulate political contributions.^[41] The Court held that (1) limits on independent expenditures made by individuals or groups to oppose or advocate the election of candidates, (2) limits on candidates' expenditures from personal and family funds, and (3) limits on candidates' overall campaign expenditures violate the First Amendment, but it upheld restrictions on political contributions as a means of preventing corruption or the appearance of corruption.^[42] The Court did not require any showing that the prohibited contributions caused any harm, and it recognized, in effect, substantial government power to regulate political contributions and to limit political speech and association.

The Supreme Court held that campaign expenditure limits are subject to strict scrutiny because they burden fundamental First Amendment rights of political association and political expression.^[43] A limit on campaign spending "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."^[44] Spending limits impose "substantial . . . restraints on the quantity and diversity of political speech" because "virtually every means of communicating ideas in today's mass society requires the expenditure of money."^[45] Given these burdens, the Court held that limits on overall campaign expenditures by federal candidates^[46] and limits on candidates' personal expenditures^[47] violate the First Amendment. The Court also held that limits on independent expenditures made by individuals or groups to oppose or advocate the election of candidates violate the First Amendment.^[48]

Although the Court also applied strict scrutiny to campaign contribution limits,^[49] it upheld the Federal Election Campaign Act's $1000 limit on contributions to federal candidates.^[50] The Court held that this $1000 limit on campaign contributions served the government's compelling interest in "limit[ing] the actuality and appearance of corruption resulting from large individual financial contributions" and that it "focuse[d] precisely on the problem of large campaign contributions-the narrow aspect of political association where the actuality and potential for corruption have been identified."^[51] Congress, however, had not in fact identified any particular level of political contributions, much less contributions in excess of $1000, that caused corruption or the appearance of corruption.

Congress, at most, had some basis for concern about problems created by "large" contributions in unspecified amounts. Congress set the $1000 limit in response to reports that in the 1972 national elections "large contributions" had been "given to secure a political quid pro quo from current and potential office holders" and in response to "the appearance of corruption stemming from public awareness of the opportunities for abuse inherent in a regime of large individual financial contributions."^[52] Congress did not have any evidence that contributions in excess of $1000 were large enough either to corrupt the recipients or to create the appearance of corruption.

The Court, nonetheless, did not look for any evidence in the legislative record to support the specific $1000 limit, and it did not require the government to develop any after-the-fact evidence to justify the legislature's decision. Indeed, the Court disclaimed any concern whether the $1000 limit was "unrealistically low because much more than that amount would still not be enough to enable an unscrupulous contributor to exercise improper influence . . ."^[53] It was enough that Congress thought that some limit on political contributions was necessary; there would be no judicial "fine tuning" of legislative choices between a $1000 limit and a $2000 limit.^[54] In short, the Court permitted Congress to limit political contributions without any evidence that contributions in any particular amounts in excess of $1000 caused the harms, corruption, or the appearance of corruption, that inspired the restriction.^[55]

B. First Amendment Supplement To Buckley

The Buckley Court applied strict scrutiny to campaign finance regulation, but it also accepted speculation about actual corruption or the appearance of corruption as a justification for limiting political speech and association. The Court endorsed Congress' power to prohibit contributions in "large" but unspecified amounts, and then upheld a $1000 contribution limit without any evidence that contributions in excess of $1000 caused the problem or appearance of corruption. This judgment created a risk that speculation about campaign finance problems might be sufficient to justify other restrictions on political speech and association.^[56] Indeed, the Missouri limits on political expenditures and contributions, adopted in 1994 by the legislature in Senate Bill 650 and by the electorate in Proposition A, were based on just such speculation and conjecture about problems in the electoral process.

Fortunately, however, a significant evolution of First Amendment law provided a means to close the door left open in Buckley and to protect political speech. In United States v. National Treasury Employees Union ("NTEU"),^[57] the Supreme Court imposed a heavy burden on government to justify regulation of federal employees' nonpolitical speech and writing. The government must offer more "than mere speculation about serious harms."^[58] It must demonstrate that "the recited harms are real . . . and that the regulation will in fact alleviate these harms in a direct and material way."^[59] The challenges to the 1994 Missouri statutes created an opportunity to transfer this First Amendment requirement to campaign finance regulation and to force government to justify regulation of political speech, not on the basis of public perceptions and fears, but instead on the basis of actual, real world, harms.^[60]

1. National Treasury Employees Union: Duty To "Demonstrate That The Recited Harms Are Real"

The Supreme Court held in NTEU that a prohibition against federal employees accepting honoraria for making appearances, giving speeches, or writing articles violated the First Amendment. The government's interest "that federal officers not misuse or appear to misuse power by accepting compensation for their unofficial and nonpolitical writing and speaking activities" was "undeniably powerful."^[61] Nonetheless, regulation of federal employees' speech required some evidence that accepting honoraria caused the harm that the government postulated:

[w]hen the Government defends a regulation on speech as a means to redress past harms or prevent anticipated harms, it must do more than simply "posit the existence of the disease sought to be cured." . . . It must demonstrate that the recited harms are real, not merely conjectural, and that the regulation will in fact alleviate these harms in a direct and material way.^[62]

Although the government posited the disease (misuse of power), it could not demonstrate that the acceptance of honoraria by low-level federal employees caused any real harm. The government had "no evidence of misconduct related to honoraria in the vast rank and file of [low-level] federal employees," and it had only "limited evidence of actual or apparent impropriety by legislators and high-level executives."^[63] Accordingly, the Court held that Congress could not prohibit lower ranking federal employees from accepting honoraria.^[64]

As Justice O'Connor noted, Congress did not have any evidence that the acceptance of honoraria by low-level federal employees was a problem. There was no showing "that Congress considered empirical or anecdotal data pertaining to abuses by lower-echelon executive employees."^[65] The government did not have any other evidence of harm. The government's lawyers cited an official 1992 report, made three years after the ban on honoraria was imposed, but the Court found that "[i]ts 112 pages contain not one mention of any real or apparent impropriety related to a lower level employee."^[66] The Court refused to defer to the government's speculation about the problems caused by low-level federal employees accepting honoraria.^[67] It insisted that even burdens on nonpolitical expression require "a justification far stronger than mere speculation about serious harms."^[68]

2. "Real Harm" As A Prerequisite For Campaign Finance Regulation

The ACLU/EM lawyers concluded that NTEU could be used to establish a new First Amendment requirement for regulation of political expenditures and contributions: the State must demonstrate that campaign finance regulations address real harms. If the State could not demonstrate that its regulations addressed real harms, then there was no warrant for any restriction on political speech or association. If, however, the State demonstrated that its regulations addressed real harms, then the court should apply Buckley's strict scrutiny test.^[69] The State must show that it has a compelling interest in rectifying these real harms; that its regulations are "narrowly tailored;" and that it has considered alternative means of regulation that would impose fewer burdens on First Amendment interests. In short, the State "must demonstrate that the recited harms are real, not merely conjectural, and that the regulation will in fact alleviate these harms in a direct and material way."^[70]

We argued in Shrink I and in Carver that the State must demonstrate that the legislature had considered carefully evidence of real problems and that it had "substantial evidence" of those problems. The State could not satisfy NTEU's evidentiary requirement with after-the-fact rationalizations advanced by its counsel at trial. Justice O'Connor had examined the evidence before Congress to determine in NTEU whether the acceptance of honoraria by low-level federal employees caused the harm of misuse of power that the government recited.^[71] Similarly, when Justice Kennedy originally stated the test adopted in NTEU, he turned to the evidence of real harms considered by the legislature.^[72] Justice Kennedy recognized that courts must accord substantial deference to the legislature's predictive judgments and findings. Nonetheless, courts must also "exercise independent judgment when First Amendment rights are implicated," and they must determine that the legislature "has drawn reasonable inferences based on substantial evidence."^[73]

We also argued that the State must meet this evidentiary burden for measures like Proposition A, enacted directly by the voters, as well as for measures, like Senate Bill 650, enacted by the legislature. It is well-settled that "voters may no more violate the Constitution by enacting a ballot measure than a legislative body may do so by enacting legislation."^[74] Therefore, the State must demonstrate that the electorate, like the legislature, had considered carefully evidence of real harms. The fact that a super-majority of the voters had adopted Proposition A did not in itself satisfy this evidentiary burden.^[75] A law enacted by popular vote may in fact be more suspect than a law enacted by the legislature. The electorate does not have the benefit of public hearings and legislative deliberations to analyze proposals. Moreover, voters, unlike legislators, do not take an oath to uphold the Constitution.^[76]

We were confident that Missouri could not meet its burden of demonstrating "real harm" under NTEU; the problems in the financing of state and local elections in Missouri were, at best, speculative. The state legislature and electorate had not considered any evidence of specific problems. Indeed, neither Senate Bill 650 nor Proposition A even "recited" any particular harms allegedly to be redressed by their provisions.^[77] Even if courts were willing to consider after-the-fact rationalizations, it seemed unlikely that counsel would be able to fill the evidentiary void at trial. It would not be sufficient for the State to submit books of election data from which a court or others might hypothesize "problems" in the Missouri electoral process.^[78] It would not be sufficient for the State to recite a brief summary of recent so-called "campaign reform" efforts and to make vague assertions of popular dissatisfaction with Missouri politics.^[79] It would not be sufficient for the State's counsel to assert broadly and without any factual support that "such large sums of money are now being spent on Missouri campaigns that the vast majority of Missouri citizens have seen the need to take affirmative steps to refocus those campaigns away from money and toward the kind of grass-roots campaigning that is more likely to make the process open, fair, and informative."^[80] The State's interests were entirely hypothetical because there was no evidence, much less any substantial evidence, of any real harms in the Missouri electoral process.

C. Adoption Of The "Real Harm" Standard In The Eighth Circuit

The ACLU/EM made two efforts to supplement Buckley's First Amendment standards. In Shrink I, we argued that some evidence of a "real harm" was a prerequisite for regulation of candidates' political expenditures.^[81] Although the district court gave cautious consideration to this argument, both the district court and the court of appeals ultimately held that Missouri's spending limits, under a direct application of Buckley, were invalid.^[82] Our second effort, however, was successful. Appearing as an amicus in Carver, we argued again that the State had a duty under NTEU to demonstrate that campaign contribution regulations addressed a "real harm."^[83] The court of appeals adopted the NTEU test, and it found no evidence that contributions to candidates in excess of the prohibited amounts caused any harm.^[84] Although Missouri protested that the Eighth Circuit had established a new First Amendment hurdle for campaign finance regulation, the Supreme Court denied review.^[85] In Shrink II, a district court gave full effect to the Eighth Circuit's new First Amendment standard for regulation of political speech.^[86] It held that Missouri's limits on campaign contributions during legislative sessions violated the First Amendment because the State had no evidence that such contributions caused any real harm.^[87]

1. Shrink I--Cautious Consideration

In its first case, Shrink I, the ACLU/EM argued that the 1994 Missouri spending limits violated the First Amendment because the State had no evidence that any specific amount of candidates' total campaign expenditures, of candidates' personal expenditures, or of candidates' expenditure of carryover funds caused any "real harm." The district court expressly noted our argument that under NTEU the State must "demonstrate that the recited harms are real."^[88] The district court did not, however, require the State to make any showing that the legislature or the electorate had any evidence of harm. It did not even require the State to make any after-the-fact showing that the prohibited political expenditures caused any real harm. Although we argued "that the state must prove that the legislature and the electorate explicitly considered the specific wrongs now claimed to be remedied by the statutes,"^[89] the district court refused to require "specific, historical factual proof."^[90] In fact, the district court considered certain election reports attached as exhibits to the one of the State's briefs even though there was "no evidence provided that the legislature or the electorate actually considered or even had available" these reports.^[91] The election reports, however, did "nothing to prove or disprove either defendants' or plaintiffs' case."^[92] Thus, the State, notwithstanding its reliance on "after-the-fact rationalizations,"^[93]did not make any showing that the three campaign expenditure limits addressed any real harms in the Missouri electoral process.

Instead of requiring the State to make some showing of a "real harm" under NTEU, the district court gave the State "the benefit of the doubt" and assumed "that in fact the justifications offered [by counsel were] the reasons for the passage of the laws."^[94] Although the district court tested the statutes in light of the harms "recited" by the State on cross motions for summary judgment, it held that the three limits on candidates' spending were invalid under Buckley.^[95] Limits on direct expenditures by candidates with their own money were "clearly unconstitutional" under Buckley.^[96] Limits on candidates' total campaign expenditures "clearly violate the rule of Buckley, and cannot withstand strict scrutiny."^[97] Missouri's spending limits "restrict the quantity of political speech" and, like the spending limits held invalid in Buckley, do not promote any state interest "in preventing quid pro quo corruption."^[98] Finally, the restrictions on the expenditure of carryover campaign funds could not be justified under Buckley.^[99] The requirements that candidates "spend down" their campaign funds rather than carrying them forward for use in a subsequent election "limit political speech by telling a candidate when he or she must speak," and they "do nothing to further the goal of preventing corruption."^[100] The court of appeals affirmed the district court's judgment and its application of Buckley, and it did not discuss our NTEU argument that the State must "demonstrate that the recited harms are real."^[101]

2. Carver--Adoption

The NTEU argument that the district court cautiously skirted in Shrink I, carried the day in Carver. The court of appeals adopted our argument that the State has a duty to demonstrate that its campaign finance regulations address real harms.^[102] It found that the State had no evidence that the Proposition A contribution limits addressed any "real" harm or disease. Although the court of appeals noted the State's after-the-fact arguments at trial regarding the harms allegedly caused by campaign contributions,^[103] it expressly determined that judicial deference to predictive judgments about harms would turn on the evidence before the legislature or the electorate.^[104] Moreover, the court accepted our argument^[105] that laws, like Proposition A, adopted directly by the electorate through initiatives or referenda were not entitled to the same level of deference as legislative measures.^[106]

The court of appeals recognized that "[t]he question [was] not simply that of some limits or none at all"^[107] on campaign contributions. Both Proposition A and Senate Bill 650 limited campaign contributions,^[108] but "Proposition A limits [were] only ten to twenty percent of the higher limits in Senate Bill 650."^[109] The narrow question was, then, whether lowering the contribution limits set by Senate Bill 650 to the levels set by Proposition A addressed any "real harm."^[110] The court of appeals, invoking the NTEU standard, found that "[t]he record is barren of any evidence of a harm or disease that needed to be addressed between the limits of Senate Bill 650 and those enacted in Proposition A."^[111] The State had "no evidence as to why the Proposition A limits of $100, $200, and $300 were selected;"^[112] it had "no evidence to demonstrate that the limits were narrowly tailored to combat corruption or the appearance of corruption associated with large campaign contributions."^[113]

Although the State had "no evidence" that contributions above Proposition A's limits caused any "real harm," the State, ironically, demonstrated that these contribution limits would have a substantial adverse effect on candidates' ability to communicate their political views. Proposition A's $100, $200, and $300 contribution limits per election cycle were "dramatically lower than the $2000 limit per election cycle approved in Buckley."^[114] Although the contribution limits approved in Buckley would have barred only about 5.1% of the contributions made in the preceding 1974 election, the Proposition A limits would have effected "a much higher percentage of contributors."^[115] Given "no evidence" that contributions in excess of Proposition A's limits caused any "real harm" and the substantial evidence of their adverse effects on political speech, the court of appeals concluded that the State had not carried its burden of justification: "the State has failed to carry its burden of demonstrating that Proposition A will alleviate the harms in a direct and material way."^[116]

The court of appeals also rejected the State's contention that it should defer to the voters' judgments about the harms in the political process to be cured by contribution limits.^[117] The Supreme Court in NTEU had adopted verbatim Justice Kennedy's original statement of the "real harm" test in Turner Broadcasting System, Inc. v. FCC,^[118] and it was true, as the State argued, that Justice Kennedy had recognized that courts "must accord substantial deference to the predictive judgments" of the legislature.^[119] The court of appeals, however, refused to accord the "same deference to Proposition A adopted through the initiative process by the citizens of Missouri."^[120] Deference "requires that courts ascertain that the legislative body `has drawn reasonable inferences based on substantial evidence.'"^[121] The voters, however, had no evidence, much less any substantial evidence, to support their predictive judgments.^[122] Moreover, the reasons for judicial deference to "legislative enactments" do not apply to "proposals adopted by initiative."^[123]

3. Shrink II--Application

Shortly after the court of appeals adopted the NTEU test in Carver, a district court held that Missouri's limits on campaign contributions during legislative sessions violated the First Amendment. In Shrink II,^[124] the court insisted that the State must do more than posit the existence of some problem or disease and must do more than recite some harm; the State must demonstrate that the prohibited campaign contributions cause some "real harm." The district court reviewed the State's after-the-fact evidence at trial, and it also considered the evidence available to the legislature. Missouri, however, simply had no evidence of any contributions, either to incumbents or to challengers, during the regular legislative sessions that were actually corrupt or that created the appearance of corruption.

In Senate Bill 650, the Missouri legislature prohibited statewide elected officials, state senators, state representatives, and candidates for those offices from accepting campaign contributions during any regular session of the general assembly.^[125] The district court found that this "temporal ban on contributions . . . effectively eliminates a candidate's contributions intake for four and one-half . . . months."^[126] Given this "severe impact" on political association and communication, the State had to prove that the limit on political contributions was "narrowly tailored to serve a compelling state interest."^[127] Although Buckley held that avoiding corruption or the appearance of corruption was a compelling interest,^[128] the State had a duty under NTEU to show some "real harm" that triggered this compelling interest.^[129]

Missouri, however, had no evidence that the acceptance of contributions during the legislative sessions caused any "real harm." At the one day trial, the State's attorneys had not presented any after-the-fact evidence of actual corruption caused by political contributions made during a legislative session or that such contributions created the appearance of corruption.^[130] Although "two witnesses testified in general terms of their belief that the public perceives the acceptance of contributions during the legislative session as `inappropriate,'" the State did not provide any "factual basis . . . for these witnesses' perception. . . ."^[131] The State did not have any "examples or incidents of actual corruption linked to such contributions nor incidents wherein `innocent' contributions were perceived by the public as being given and accepted for a corruptive intent."^[132] Newspaper articles about political contributions "fail[ed] to demonstrate a pervasive problem with contributions during the legislative session and the public's negative viewpoint of such contributions."^[133]

More importantly, the legislature did not have any evidence that contributions made during its sessions caused any "real harm." The district court found that the State had not demonstrated that "any evidence was presented to [a joint legislative campaign finance reform committee] regarding actual corruption or the appearance of corruption and the eradication of this harm by an in-session ban on campaign contributions."^[134] The court could not find any evidence that the legislature made "any attempt . . . to access and analyze the public's views on acceptance of contributions during the general assembly's regular session."^[135] In short, Missouri had "wholly failed to establish a constitutional nexus between the interest they [sought] to further, i.e. [prevention of] actual corruption or the appearance of corruption, and the vast prohibition on acceptance of contributions during the general assembly's regular session."^[136] Given the absence of any evidence that the prohibited contributions caused any "real harm," the State had "failed to carry [its] burden of demonstrating that [the statute would] alleviate actual corruption or the appearance of corruption in a direct and material way."^[137]

III. LESSONS FOR WOULD-BE REFORMERS

Shrink I, Carver, and Shrink II-the three Missouri campaign finance cases-have established a new First Amendment requirement. In the Eighth Circuit, campaign finance reformers will have to do more than recite a list of harms in order to regulate political expenditures or contributions. Instead, they must "demonstrate that the recited harms are real."^[138] This NTEU standard protects political speech against legislation and initiatives that merely pander to unproved specters of campaign ills, and it leaves government free to address actual campaign finance problems. Although courts and commentators have, as yet, taken little notice, the Supreme Court in Colorado Republican Federal Campaign Committee v. Federal Election Commission applied the same standard to a national campaign finance regulation.^[139] If, as many campaign finance reformers hope, the Court reconsiders Buckley, the three Missouri cases and Colorado Republican suggest strongly that any new restrictions on political speech must be narrowly tailored responses, not to public clamoring, but to real harms.

A. The New First Amendment Hurdle

In its petitions for review in Shrink I and Carver, Missouri recognized correctly that the Eighth Circuit had established a new standard for campaign finance laws: the states must justify regulation of political speech, not on the basis of public perceptions and fears, but instead on the basis of actual, real world harms.^[140] The State complained that the Eighth Circuit's adoption of NTEU "impose[d] a substantial obstacle to campaign finance reform efforts" because states would not be able to show that either the legislature or the electorate had considered evidence of "real" campaign finance harms.^[141] Missouri, and most other states, would not be able to demonstrate real harm because they do not make formal legislative history.^[142] Moreover, if "providing evidence to support an act of the state legislature may be difficult, proving the `initiative history' for measures adopted through the initiative process is virtually impossible."^[143] Missouri does, indeed, have reason to be concerned about the need, not for any particular type of legislative history or "initiative history," but for evidence that the legislature or the electorate actually considered "real" campaign finance harms. Although the district court in Shrink I expressly refused to impose this requirement, the court of appeals in Carver and another district court in Shrink II considered the question whether the legislature or the electorate had any evidence of actual harm, as well as after-the-fact rationalizations offered by the state's counsel at trial.^[144]

Regardless whether the court of appeals ultimately requires the State to demonstrate that the legislature or the electorate had some evidence of real harm or permits the State, through counsel at trial, to build an after-the-fact case that campaign finance regulations address a real harm, it has established an important new First Amendment bulwark for political speech and association. The State's failure to demonstrate that Proposition A's contribution limits addressed any "real harm" proved fatal in Carver. Similarly, the State's failure to demonstrate that the acceptance of contributions during legislative sessions caused any "real harm" proved fatal in Shrink II.

To say that the Missouri legislature in Senate Bill 650 and voters in Proposition A had no evidence of any harm caused by the campaign practices that they prohibited is not to deny that some practices may be harmful; that the harm might be identified; and that narrowly tailored regulations might be devised. If, for example, campaign contributions during legislative sessions cause some "real harm," then surely it is not too much to require the state to make the type of showing that the district court found wanting in Shrink II.^[145] If, however, the State can make no showing that such contributions cause any "real harm," then it is hard to understand how the State could have any interest, compelling or otherwise, in prohibiting them. Absent some evidence of "real harm," would-be campaign finance reformers should address public perceptions of corruption or other harms with narrowly tailored regulations, like disclosure provisions, that respect our nation's commitment, under the First Amendment, to open, robust political debate.^[146]

B. Buckley--Which Way Will The Tree Fall?

Although the ACLU/EM successfully supplemented Buckley's First Amendment protection of political speech with the NTEU requirement that states must "demonstrate" that campaign finance regulations alleviate "real harms," others are busy laying the groundwork to expand governmental regulation of political speech. The Brennan Center for Justice at New York University School of Law, to take but one prominent example, is now waging a "full-court press" to force the Supreme Court to overrule Buckley's First Amendment protection of candidates' political expenditures.^[147] Of course, given broad agreement that Buckley's distinction between contributions and expenditures is false and that both contributions and expenditures are political speech,^[148] there is some risk that the Court might instead overrule Buckley's approval of Congress' power to regulate contributions. It is, in short, not clear which way the tree will fall--governmental power to regulate "both contributions and expenditures-or neither."^[149]

Fortunately, there are strong reasons to believe that the tree will fall on the First Amendment side of protecting both political contributions and political expenditures. Indeed, although some courts continue to assess campaign finance regulations exclusively under the Buckley framework,^[150] a few other courts have now followed the Eighth Circuit's lead and have augmented First Amendment protection of both political expenditures and contributions. The Court of Appeals for the District of Columbia Circuit, for example, held that regulation of political contributions requires "evidence" of harm.^[151] In a similar vein, the Oregon Supreme Court held that $100 and $500 contribution limits were invalid under the state constitution in part on the ground that the state had no evidence that campaign contributions caused corruption.^[152]

Most importantly, the Supreme Court, albeit with little fanfare, has adopted the same evidentiary requirement for campaign finance regulations as the Eighth Circuit. In Colorado Republican Federal Campaign Committee v. Federal Election Commission, the Court followed Buckley and held that limits on "independent" expenditures by political parties in national elections violate the First Amendment.^[153] The Court, however, did more than simply reaffirm and apply Buckley's First Amendment standard; it imposed a duty on the government to demonstrate that independent expenditures by political parties caused some harm or problem.^[154] The Court found that neither Congress, nor the government's lawyers at trial, had any evidence that independent expenditures by political parties caused any problem of corruption.^[155] This evidentiary requirement, largely ignored by most lower courts and commentators,^[156] might well be read as implicitly overruling Buckley's approval of Congress' power to regulate contributions on the basis of speculation about corruption or the appearance of corruption.^[157] Even if it is premature to conclude that the Buckley tree has already fallen on the First Amendment side, Colorado Republican confirms the Eighth Circuit's decision to supplement Buckley's First Amendment standards and to require that restrictions on political speech are a response to real problems, not merely to public passions.^[158]

[*] Professor of Law, Washington University. A.B., Princeton University, 1969; J.D., Columbia University, 1973. The author was co-counsel for the American Civil Liberties Union of Eastern Missouri ("ACLU/EM") in two of the three principal cases discussed below. See infra note 36 and accompanying text.

[1.] Missouri State Senator Jim Mathewson, testifying before a state legislative committee, supported campaign contribution limits because the perception of the voters who had overwhelmingly favored such limits in an initiative was more important than the reality that the limits were "silly." See Virginia Young, Campaign Money Limits Look Good, ST. LOUIS POST-DISPATCH, Apr. 10, 1997, at 4B.

[2.] 424 U.S. 1 (1976) (per curiam).

[3.] See infra text at notes 41-55.

[4.] See infra text at notes 69-137.

[5.] See Colorado Republican Fed. Campaign Comm. v. Federal Election Comm'n, 116 S. Ct. 2309 (1996) (discussed infra at notes 153-58 and accompanying text).

[6.] See MO. ANN. STAT. § 130.011 et seq. (West 1997 & Supp. 1998).

[7.] See Carver v. Nixon, 72 F.3d 633, 634-35 (8th Cir. 1995), cert. denied, 116 S. Ct. 2579 (1996); Shrink Missouri Gov't PAC v. Maupin, 71 F.3d 1422, 1423 (8th Cir. 1995), cert. denied, 116 S. Ct. 2579 (1996).

[8.] Id.

[9.] See 94 Op. Mo. Att'y Gen 218 (Dec. 6, 1994), 1994 MO. AG LEXIS 24.

[10.] See MO. REV. STAT. § 130.052 (1996).

[11.] See id. § 130.052.8 (1996).

[12.] See id. § 130.052.3 (1996).

[13.] See id.

[14.] See MO. ANN. STAT. § 130.081 (West 1997).

[15.] Proposition A limited "contributions" to candidates for office in Missouri on a sliding scale from $100 to $200 to $300 per election cycle (see infra text at notes 25-27), and it had the effect of limiting candidates' personal expenditures because the term "contribution" was defined to included "[a] candidate's own money or property used in support of the person's candidacy . . ." MO. ANN. STAT. § 130.011(12)(a) (West Supp. 1998). As a result, with the exception of expenses for "food, lodging, travel, and payment of any fee necessary to the filing for public office," candidates for state and local office in Missouri could not spend more than the prescribed amounts of their own funds in their efforts to be elected. Id.

Senate Bill 650 also established contribution limits. See MO. REV. STAT. § 130.032 (1996); see infra text at notes 19-23. These limits, however, were superseded by the more restrictive contribution limits of Proposition A. After the Proposition A contribution limits were held unconstitutional (see infra text at notes 102-23), the contribution limits set by Senate Bill 650 took effect and limited contributions by candidates to their own campaigns. See MO. ANN. STAT. §§ 130.011(12)(a), 130.032.1 (West Supp. 1998). The Missouri legislature amended Senate Bill 650's contribution limits in 1997 to remove the limits on the amounts that candidates could contribute to their own campaigns. See S.B. 16, 89th Gen. Ass., § A; see MO. ANN. STAT. § 130.032.1 (West Supp. 1998) (contribution limits apply to "any person other than the candidate").

[16.] MO. REV. STAT. § 130.130 (1996).

[17.] See id. § 130.100 (1996).

[18.] Senate Bill 650 also limited candidates' expenditure of carryover campaign funds. See MO. REV. STAT. § 130.038 (1996). This provision, which was initially superseded by the more restrictive provision of Proposition A, was repealed in 1997. See S.B. 16, 89th Gen. Ass., § A.

[19.] See MO. ANN. STAT. § 130.032.1 (West Supp. 1998).

[20.] See id. § 130.032.1(1), (6) (West Supp. 1998).

[21.] See id. § 130.032.1(2), (5) (West Supp. 1998).

[22.] See id. § 130.032.1(3), (4) (West Supp. 1998).

[23.] See id. § 130.032.2 (West Supp. 1998).

[24.] See supra note 9 and accompanying text. After Proposition A's contribution limits were held unconstitutional (see infra text at notes 102-23), Senate Bill 650's contribution limits, which had been superseded under the ruling of the Missouri Attorney General, took effect. See MO. ANN. STAT. § 130.032.1 (West Supp. 1998).

[25.] See MO. REV. STAT. § 130.100 (1996).

[26.] The categories for different contribution limits in Proposition A and Senate Bill 650 were not identical. Senate Bill 650 permitted a higher level of contribution ($1000) for candidates in districts over 250,000; Proposition A subjected candidates in all districts over 100,000 to a $200 limit. Compare MO. ANN. STAT. § 130.032.1 (1), (6) (West Supp. 1998), with MO. REV. STAT. § 130.100 (1996).

[27.] Proposition A limited contributions "per election cycle," and Senate Bill 650 limited contributions "in any one election." Compare MO. REV. STAT. § 130.100 (1996), with MO. ANN. STAT. §§ 130.032, 130.032.4 (West Supp. 1998); see also MO. ANN. STAT. § 130.011(15) (West Supp. 1998) (definition of "election"); id. § 130.011(16) (West 1997) (definition of "election cycle") (the 1997 amendment deleted the definition of "election cycle"). See generally MO. ANN. STAT. § 130.011 (West Supp. 1998).

[28.] MO. ANN. STAT. § 130.032(4) (West 1997). After a district court held that this provision violated the First Amendment (see infra text at notes 124-37), the Missouri legislature repealed it. See S.B. 16, 89th Gen. Ass., § A; see MO. ANN. STAT. § 130.032 (West Supp. 1998).

[29.] 892 F. Supp. 1246 (E.D. Mo.), aff'd, 71 F.3d 1422 (8th Cir. 1995), cert. denied, 116 S. Ct. 2579 (1996).

[30.] See id. at 1429.

[31.] 882 F. Supp. 901 (W.D. Mo.), rev'd, 72 F.3d 633 (8th Cir. 1995), cert. denied, 116 S. Ct. 2579 (1996).

[32.] See Carver, 882 F. Supp. at 906.

[33.] See Carver v. Nixon, 72 F.3d 633, 645 (8th Cir. 1995).

[34.] 922 F. Supp. 1413 (E.D. Mo. 1996).

[35.] See id. at 1425.

[36.] The author of this Article and Frank Susman were co-counsel for the ACLU/EM, and Denise Field worked extensively on both cases. This Article states the opinions and analysis of the author, and it does not purport to present the views of Mr. Susman, Ms. Field, the ACLU/EM, or the national ACLU.

[37.] 424 U.S. 1 (1976) (per curiam).

[38.] See Carver, 72 F.3d at 638.

[39.] Id. (quoting United States v. National Treasury Employees Union ("NTEU"), 513 U.S. 454, 475 (1995), quoting Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 664 (1994)).

[40.] Id. (quoting NTEU, 513 U.S. at 475 (emphasis added)).

[41.] See Buckley, 424 U.S. at 14-23 ("expenditure ceilings impose significantly more severe restrictions on protected freedoms of political expression and association than . . . limitations on financial contributions").

[42.] Compare id. at 23-38 with id. at 39-59.

[43.] See id. at 14-23, 39-59.

[44.] Id. at 19 (footnote omitted).

[45.] Id.

[46.] See id. at 54-59. The Court found that there was "[n]o governmental interest . . . sufficient to justify the restriction on the quantity of political expression" imposed by campaign expenditure limits, and it specifically rejected the arguments that any interest "in equalizing the financial resources of candidates" or "in reducing the allegedly skyrocketing costs of political campaigns" could justify limits on candidates' campaign spending. Id. at 55-57.

Although the Court upheld a public financing scheme for presidential elections that included an expenditure limit, it expressly noted that this expenditure limit was valid only because it was tied directly to an offer of public financing. See id. at 57 n.65, 90-109.

[47.] See id. at 51-54. The Court held that a "ceiling on personal expenditures by candidates on their own behalf" interferes with the candidate's "right to engage in the discussion of public issues and vigorously and tirelessly to advocate his own election." Id. at 52. This restraint could not be justified as an attempt to curb the corruption or appearance of corruption of political candidates. See id. at 53. In fact, "the use of personal funds reduces the candidate's dependence on outside contributions and thereby counteracts the coercive pressures and attendant risks of abuse to which . . . contribution limitations are directed." Id. (footnote omitted).

[48.] See id. at 39-51. Although the Court subsequently upheld one very narrowly tailored restriction on independent expenditures by corporations in support of candidates, it has never wavered from the fundamental proposition that limits on candidates' campaign spending violate the First Amendment. Compare Austin v. Michigan Chamber of Commerce, 494 U.S. 652, 655-61 (1990), with Buckley, 424 U.S. at 39-57.

[49.] Campaign contribution limits are subject to strict scrutiny because they burden fundamental First Amendment rights of political association and political expression. See Buckley, 424 U.S. at 14-29; accord Citizens Against Rent Control v. Berkeley, 454 U.S. 290, 298 (1981) ("Contributions by individuals . . . [are] beyond question a very significant form of political expression . . . [and] regulation of First Amendment rights is always subject to exacting judicial scrutiny.").

[50.] See Buckley, 424 U.S. at 14-29. The Court also upheld a $5000 limit on contributions by political committees to candidates and a $25,000 limit on total contributions by an individual during any calendar year. See id. at 35-36, 38.

[51.] Id. at 26, 28 (emphasis added). The Court found that there was no connection between the compelling interest in avoiding corruption and limits on candidates' campaign expenditures. See id. at 55 ("The interest in alleviating the corrupting influence of large contributions is served by . . . contribution limitations and disclosure provisions rather than by . . . campaign expenditure ceilings."). In Austin v. Michigan Chamber of Commerce, the Court did recognize a connection between limits on independent expenditures by corporations in support of candidates and corruption. 494 U.S. 652, 660 (1990). The Austin Court, however, did not question the understanding in Buckley that limits on candidates' expenditures are not related to the government's interest in avoiding corruption.

[52.] Buckley, 424 U.S. at 26-27. The Supreme Court did not discuss any evidence that contributions of any particular amount had been given to secure a political quid pro quo. It simply noted that "the deeply disturbing examples surfacing after the 1972 election demonstrate that the problem is not an illusory one," and cited two pages and three footnotes in the court of appeals' opinion that "discussed a number of the abuses uncovered after the 1972 elections." Id. at 27 & n.28. The court of appeals identified the amount of only one contribution, a $2,000,000 contribution from the dairy industry to President Richard M. Nixon's 1972 re-election campaign. See Buckley v. Valeo, 519 F.2d 821, 839-40 & nn.36-40 (D.C. Cir. 1975).

[53.] Buckley, 424 U.S. at 30.

[54.] See id. at 30.

[55.] Although the Buckley Court did not address the empirical question whether contributions over $1000 caused any identifiable harm, it did consider evidence of the actual effect of this contribution limit on political speech. The Court recognized that "[g]iven the important role of contributions in financing political campaigns, contribution restrictions could have a severe impact on political dialogue if the limitations prevented candidates and political committees from amassing the resources necessary for effective advocacy." Id. at 21. There was, however, "no indication" that the $1000 limit would have "any dramatic adverse effect" because only 5.1% of all contributions to congressional candidates in 1974 had been made in contributions exceeding $1000. Id. at 21 & n.23, 26 n.27. Moreover, it seemed likely that "some or all" of those funds "could have been replaced through efforts to raise additional contributions from persons giving less than $1000." Id. at 26 n.27. In addition to considering the actual effect of the $1000 contribution limit on candidates' ability to communicate their views, the Court also considered the effect of this contribution limit on incumbents and challengers. The record, however, did not provide any basis for concluding that the $1000 limit favored incumbents over major-party, minor-party, or independent challengers. See id. at 30-35.

[56.] See Citizens Against Rent Control v. Berkeley, 454 U.S. 290, 296-97 (1981) (Buckley authorizes limits on political activity to address the "perception of undue influence of large contributors to a candidate").

[57.] 513 U.S. 454 (1995).

[58.] Id. at 475.

[59.] Id. (quoting Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 664 (1994)).

[60.] The Court had suggested in one post-Buckley decision that mere speculation about harms would not be sufficient to justify a $1000 limit on a political action committee's independent expenditures in support of a presidential candidate. See Federal Election Comm'n v. National Conservative Political Action Comm., 470 U.S. 480, 498 (1985) ("On this record, . . . an exchange of political favors for uncoordinated expenditures remains a hypothetical possibility and nothing more.").

[61.] NTEU, 513 U.S. at 472.

[62.] Id. at 475 (quoting Turner Broad. Sys., Inc., 512 U.S. at 664).

[63.] NTEU, 513 U.S. at 472.

[64.] See id. at 476-78.

[65.] Id. at 485 (O'Connor, J., concurring in the judgment in part and dissenting in part); see id. at 483 (Congress relied on reports that did not note "any problems, anecdotal or otherwise, stemming from the receipt of honoraria by rank-and-file Executive Branch employees.")

[66.] Id. at 472 n.18.

[67.] See id. at 467 n.11 ("The honoraria ban . . . deters an enormous quantity of speech before it is uttered, based only on speculation that the speech might threaten the Government's interests.").

[68.] Id. at 475; see id. at 475-76 n.21 ("Deferring to the Government's speculation about the pernicious effects of thousands of articles and speeches yet to be written or delivered would encroach unacceptably on the First Amendment's protections.").

[69.] Courts in the Eighth Circuit had recognized that, under Buckley v. Valeo, 424 U.S. 1 (1976) (per curiam), and its progeny, regulations of political speech must be "narrowly tailored" to serve a "compelling" state interest. See e.g., Day v. Holahan, 34 F.3d 1356, 1361 (8th Cir. 1994), cert. denied, 115 S. Ct. 936 (1995).

[70.] NTEU, 513 U.S. at 475 (emphasis added).

[71.] See supra note 65 and accompanying text.

[72.] The NTEU test quoted directly Justice Kennedy's plurality opinion in Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 664 (1994). See supra note 62 and accompanying text.

[73.] Turner Broad. Sys., Inc., 512 U.S. at 666 (Kennedy, J. plurality opinion).

[74.] Citizens Against Rent Control v. Berkeley, 454 U.S. 290, 295 (1981); see U.S. Term Limits, Inc. v. Thornton, 514 U.S. 779 (1995) (holding unconstitutional congressional term limits adopted directly by state voters).

[75.] Seventy-four percent of the voters in the November 1994 elections approved Proposition A. See Carver v. Nixon, 72 F.3d 633, 640 (8th Cir. 1995). There were many reasons for voter approval of Proposition A, some of which, like a desire to level the playing field, are themselves unconstitutional, and others, like controlling spiraling campaign costs, that are not sufficiently compelling to justify restrictions on political speech. See Buckley v. Valeo, 424 U.S. 1, 48-49, 54, 56, 57 (1976) (per curiam). Thus, the mere fact that a majority of the electorate adopted Proposition A could not satisfy the State's duty to show that it had a compelling interest.

[76.] See U.S. CONST. art. VI, cl. 3; MO CONST. art. III, § 15.

[77.] Senate Bill 650 did not include any statement of legislative findings. Similarly, Proposition A did not include a preamble identifying and describing the problems that the various provisions of this initiative were intended to address, and it offered no explanation how the various regulations were tailored to accomplish these objectives.

[78.] In Shrink I, Missouri filed a stack of election reports with their reply brief. See Memorandum and Order at 2, Shrink Missouri Government PAC v. Maupin, 892 F. Supp. 1246 (E.D. Mo. 1995) (No. 4:CV815 CDP). The district court ultimately concluded that this raw data did "nothing to prove or disprove either defendants' or plaintiffs' case." Shrink I, 892 F. Supp. at 1250.

[79.] See Memorandum In Support of Defendants' Motion For Summary Judgment at 2-3, 4-5, Shrink Missouri Government PAC v. Maupin, 892 F. Supp. 1246 (E.D. Mo. 1995) (No. 4:95CV815 CDP).

[80.] Id. at 5 (footnote omitted). Although counsel intoned this hypothetical state interest in lofty terms, the Supreme Court had previously held that any interest "in reducing the allegedly skyrocketing costs of political campaigns" could not justify limits on candidates' campaign spending. Buckley, 424 U.S. at 57.

[81.] See Shrink I, 892 F. Supp. at 1249.

[82.] See infra notes 96-101 and accompanying text.

[83.] See Brief of The American Civil Liberties Union of Eastern Missouri as Amicus Curiae in Support of Appellant at 7, Carver v. Nixon, 72 F.3d 633 (8th Cir. 1995) (No. 95-2608).

[84.] See Carver, 72 F.3d at 638, 644.

[85.] See infra notes 140-42 and accompanying text.

[86.] See Shrink II, 922 F. Supp. 1413 (E.D. Mo. 1996).

[87.] See id. at 1420-24, 1425.

[88.] Shrink I, 892 F. Supp. 1246, 1249 (E.D. Mo. 1995) (quoting NTEU, 513 U.S. 454, 475 (1995), citing Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 664 (1994)).

[89.] Shrink I, 892 F. Supp. at 1250.

[90.] Id. The district court rejected any contention that the "wrongs" to be remedied must by identified "by including `findings' in the laws themselves or by, for example, presenting evidence of legislative history or testimony about the intent of the electorate or of the legislature . . ." Id. Missouri, and other states "where the legislature does not collect or preserve legislative history, could never meet [this] burden . . ." Id.

[91.] Id. The State did not present any other evidence. Id.

[92.] Id.

[93.] Id. at 1249.

[94.] Id. at 1250; see id. at 1249 (State had not "provided any evidence that either the state legislature or the electorate actually considered those justifications [advanced by the State at trial] in enacting Senate Bill 650 or Proposition A.").

[95.] See id. at 1251-54. The court also held that a statute regulating candidates' campaign advertisements violated the First Amendment. See id. at 1254-56. The State did not appeal this judgment. See Shrink I, 71 F.3d 1422, 1423 n.2 (8th Cir. 1995).

[96.] See Shrink I, 892 F. Supp. at 1251. The district court also found that Missouri had made "no attempt to provide any justification" for "this direct limitation on a candidate's own political speech." Id.

[97.] Id. at 1253.

[98.] Id. at 1252. The district court held that the State's statutory scheme is "coercive[] because it withdraws an important source of private campaign funding otherwise available to candidates." Id. It can not be justified under "Buckley's rationale for upholding spending limits in return for public financing" because "Missouri provides no matching or public funds to any candidates." Id.

[99.] See id. at 1254.

[100.] Id. at 1254 (emphasis in original). Even if limiting the power of incumbency was a compelling interest under Buckley, the district court found that the Missouri statute "is not narrowly tailored to serve that interest." Id. The State's interest in "`equaliz[ing]' the spending power of the rich and poor" was not compelling under Buckley. Id.

[101.] Shrink I, 71 F.3d 1422 (8th Cir. 1995).

[102.] See Carver v. Nixon, 72 F.3d 633, 638 (8th Cir. 1995) (quoting the NTEU test as set out in the ACLU/EM amicus brief); see supra text accompanying notes 69-80.

[103.] In Shrink I, we argued that the State had a duty under NTEU to show that the legislature, or the electorate, had considered real harms. See supra text accompanying notes 89-90. In Carver, we suggested that the question whether the state could rely on "after-the-fact rationalizations" offered at trial as evidence of these harms or whether there had to be some evidence that the legislature or the electorate had considered these harms was moot. The district court had given the State an opportunity at trial to make some showing that Proposition A's contribution limits addressed some "real harm," and the State had come up empty handed. See Brief of the American Civil Liberties Union of Eastern Missouri as Amicus Curiae in Support of Appellant at 14-15 n.7, Carver v. Nixon, 72 F.3d 633 (8th Cir. 1995) (No. 95-2608).

[104.] See Carver, 72 F.3d at 644.

[105.] See supra text at notes 74-76.

[106.] See Carver, 72 F.3d at 644-45.

[107.] Id. at 642.

[108.] See supra text at notes 19-27.

[109.] Carver, 72 F.3d at 642 (footnote omitted).

[110.] Id. ("The question is . . Proposition A [contribution limits] as compared to those in Senate Bill 650, which was to become effective January 1, 1995.").

[111.] Id. at 643 (citing NTEU, 513 U.S. 454, 474 (1995).

[112.] Carver, 72 F.3d at 642-43.

[113.] Id. at 643 (citation omitted). The court of appeals noted that "[t]he State presented testimony at trial about a $420,000 contribution from a Morgan Stanley political action committee to various races in north Missouri." Id. at 642. It concluded, however, that "[a] $42,000 contribution is a far cry from the limits [$100 to $300] in Proposition A" and that the state's other examples "involve[d] individual conduct leading to criminal prosecution." Id.

[114.] Id. at 641-42; see Buckley v. Valeo, 424 U.S. 1, 24 (1976) ($1000 limit "applies to aggregate amounts contributed to the candidate for each election-with primaries, run-off elections, and general elections counted separately."). The court of appeals noted the ACLU/EM's determination that "after adjusting for inflation, Proposition A's $300 limit is 6 percent of the limit per election cycle considered in Buckley, the $200 limit is 4 percent of the Buckley limit per election cycle, and the $100 limit is only 2 percent of the Buckley limit per election cycle." Carver, 72 F.3d at 642 n.8.

[115.] Carver, 72 F.3d at 643; see supra note 55. The "State's own evidence" showed that 19% to 35.6% of the contributors in various 1994 Missouri elections gave more than the amounts permitted by Proposition A. See Carver, 72 F.3d at 643-44.

[116.] Carver, 72 F.3d at 644 (citation omitted); see id. ("The State made no showing as to why it was necessary to adopt the lowest contribution limits in the nation and restrict the First Amendment rights of so many contributors in order to prevent corruption or the appearance of corruption associated with large campaign contributions.").

[117.] See id. at 644.

[118.] 512 U.S. 622, 664 (1994); see supra note 62 and accompanying text.

[119.] Turner Broad. Sys., Inc., 512 U.S. at 665 (1994); see Carver, 72 F.3d at 644.

[120.] Carver, 72 F.3d at 644.

[121.] Id. at 644 (quoting Turner Broad. Sys., Inc., 512 U.S. at 666).

[122.] The court of appeals found that "[t]here is simply no evidence in the record identifying the source of Proposition A, whether it was an individual or a group, the process of its development, nor the reasons for the particular dollar limits" and that "there is no evidence of the details of the campaign waged in support of the initiative." Carver, 72 F.3d at 644.

[123.] Id. at 645. The court of appeals found:

Legislative bodies consist of elected representatives sworn to be bound by the United States Constitution, and their legislative product is subject to veto by the elected executive, either President or Governor. The process of enactment, while perhaps not always perfect, includes deliberation and an opportunity for compromise and amendment, and usually committee studies and hearings. These are substantial reasons for according deference to legislative enactments that do not exist with respect to proposals adopted by initiative.

Id. at 644-45. See supra notes 74-76 and accompanying text (additional reasons why popular measures are entitled to less deference than legislative measures).

[124.] 922 F. Supp. 1413 (E.D. Mo. 1996).

[125.] See supra text at note 28.

[126.] Shrink II, 922 F. Supp. at 1418 (footnote omitted).

[127.] Id. at 1420.

[128.] See id.; see supra text at note 51.

[129.] See Shrink II, 922 F. Supp. at 1420 ("[t]he harm that the [state] seek[s] to eradicate must exist"); see id. at 1421 (quoting the NTEU "real harm" test).

[130.] See id. at 1420-21.

[131.] Id. at 1421.

[132.] Id.

[133.] Id.

[134.] Id. at 1423; see id. ("no record of any [legislative] discussion or debate about a prevailing problem with actual corruption or the appearance of corruption necessitating such a ban").

[135.] Id.

[136.] Id. at 1422.

[137.] Id. at 1424. The court also held that the Missouri statute was "unconstitutionally vague." Id. at 1424-25.

[138.] Carver v. Nixon, 72 F.3d 633, 638 (8th Cir. 1995) (quoting NTEU, 513 U.S. 454, 457 (1995)).

[139.] 116 S. Ct. 2309 (1996); see infra notes 153-58 and accompanying text.

[140.] Missouri sought review in Shrink I in part on the ground that the court of appeals had adopted a new "evidentiary requirement." Petition For Writ Of Certiorari at 26-29, Nixon v. Shrink Missouri Government PAC, 71 F.3d 1422 (8th Cir. 1995) (No. 95-1524); see Reply In Support Of Petition For Writ Of Certiorari at 4-5, Nixon v. Shrink Missouri Government PAC, 71 F.3d 1422 (8th Cir. 1995) (No. 95-1524) (complaining about the "Eighth Circuit's decision to demand some unspecified kind of additional evidence"). Similarly, the State sought review in Carver in part on the ground that the Eighth Circuit had imposed "an evidentiary requirement that is entirely absent from Buckley." Petition For Writ Of Certiorari, Nixon v. Carver, 72 F.3d 633 (8th Cir. 1995) (No. 95-1258); see id. at 25-26; see also Brief of the States of Kentucky As Amici Curiae On Petition For Writ Of Certiorari at 20, Nixon v. Carver, 72 F.3d 633 (8th Cir. 1995) (No. 95-1258) (amicus brief for nine states and two cities arguing that, contrary to Buckley, the Eighth Circuit had imposed a "burden upon the State or local government to prove why it selected a particular contribution limit lower than other available alternatives.").

[141.] Petition For Writ Of Certiorari at 25, Nixon v. Carver, 72 F.3d 633 (8th Cir. 1995) (No. 95-1258).

[142.] Petition For Writ Of Certiorari at 27-28, Nixon v. Shrink Missouri Government PAC, 71 F.3d 1422 (8th Cir. 1995) (No. 95-1524).

[143.] Id. at 28. Missouri noted that, under state statutes and case law, the text of an initiative does not identify its purpose, and that "nothing in the formal initiative process creates a record of the findings or goals of either the proponents or circulators of the measure, or of the citizens who vote for its passage." Id.

[144.] Compare supra text at notes 88-90 with text at notes 102-04, 124-36.

[145.] See supra text at notes 130-37. A recent Arkansas district court decision suggests the ways in which a state might satisfy the Eight Circuit's "real harm" standard. See Russell v. Burris, 978 F. Supp. 1211 (E.D. Ark. 1997) (upholding a $100 per election contribution limit for certain non-statewide elections on grounds that the State had "substantial evidence" of the need for this limit).

[146.] The Missouri legislature amended the State's campaign finance disclosure requirement in 1997 and established an electronic reporting system. See MO. ANN. STAT. §§ 130.046, 130.057 (West Supp. 1998).

[147.] See, e.g., Rosenkranz, Campaign Reform: The Hidden Killers, THE NATION, May 5, 1997, at 5; Jack W. Germond & Jules Witcover, A Full-Court Press On Campaign Cash, NAT'L J., Dec. 7, 1996, at 2656.

[148.] See Colorado Republican Fed. Campaign Comm. v. Federal Election Comm'n, 116 S. Ct. 2309, 2325-28 (1996) (Thomas, J., concurring in the judgment and dissenting in part) ("[a] contribution is simply an indirect expenditure; though contributions and expenditures may thus differ in form, they do not differ in substance"); Buckley v. Valeo, 424 U.S. 1, 241-42 (1976) (per curiam) (Burger, C.J.) ("contributions and expenditures are two sides of the same First Amendment coin"); Lillian R. BeVier, Money and Politics: A Perspective on the First Amendment and Campaign Finance Reform, 73 CAL. L. REV. 1045, 1062-65 (1985).

[149.] See Germond & Witcover, supra note 147.

[150.] See e.g., Kentucky Right to Life, Inc. v. Terry, 108 F.3d 637, 648-51 (6th Cir. 1997), cert. denied, 118 S. Ct. 162 (1997) (upholding campaign contribution limits under Buckley without addressing the question whether the State had any evidence that contributions in excess of the prohibited amounts caused any real harm).

[151.] See Blount v. SEC, 61 F.3d 938, 944 (D.C. Cir. 1995), cert. denied, 116 S. Ct. 1351 (1996) (recognizing government's duty to demonstrate that "the ills it claims the rule addresses in fact exist and the rule will materially reduce them" (citing Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 664 (1994)) and finding that the particular harm, conflict of interest, was self-evident). The District Court for the District of Columbia subsequently held that $50 and $100 campaign contribution limits violated the First Amendment. National Black Police Ass'n v. District of Columbia, 924 F. Supp. 270 (D.D.C. 1996), vacated and remanded, 108 F.3d 346 (D.C. Cir. 1997). There was no evidence to support these contribution limits, and the court, following Carver, expressly refused to defer to the predictive judgments of the legislature, which had not considered any "solid evidence" about the need for and the potential effects of the $50 and $100 contribution limits. See id. at 281-82, 285. The court of appeals found that the enactment of legislation increasing the contribution limits mooted the case, and it vacated the district court's judgment. See 108 F.3d at 348, 354.

The Carver/NTEU requirement of "real harm" will be raised to challenge both Alaska and Colorado campaign finance regulations. See Alaska Civil Liberties Union v. Alaska, No. 3AN-97-CI (Superior Court, 3d Judicial District); Durham v. Colorado (D. Colo.), No. 96-WY-2973).

[152.] See Vanatta v. Kiesling, 931 P.2d 770, 785-87 (Or. 1997).

[153.] See 116 S. Ct. 2309 (1996). An expenditure made by a political party and that is not coordinated with a candidate is an "independent" expenditure. See id. at 2313. Seven members of the Court, in three opinions, supported the Court's judgment that limits on a political party's "independent" expenditures violated the First Amendment. Justice Breyer announced the judgment of the Court and delivered an opinion, joined by Justice O'Connor and Justice Souter, that limits on "independent" expenditures violate the First Amendment. Four other members of the Court would have held that limits on both "independent" and "coordinated" expenditures violate the First Amendment. Justice Kennedy, joined by The Chief Justice and Justice Scalia, supported this broader holding. See id. at 2321 (Kennedy, J., Rehnquist, C.J., & Scalia, J., concurring in the judgment and dissenting in part). Justice Thomas, also joined by the Chief Justice and Justice Scalia, supported this broader holding. See id. at 2323 (Thomas, J., Rehnquist, C.J., & Scalia, J., concurring in the judgment and dissenting in part).

[154.] In two of the three opinions supporting the Court's judgment (see supra note 153), six members of the Court expressly adopted Justice Kennedy's statement for a plurality in Turner Broadcasting System, Inc. v. FCC of the government's duty to demonstrate that regulations of speech address a real harm. 116 S. Ct. at 2317 (Breyer, J., O'Connor, & Souter, JJ.) (citing Turner Broad. Sys., Inc., 512 U.S. at 661-63 (Kennedy, J.)), see Colorado Republican, 116 S. Ct. at 2331 (Thomas, J., Rehnquist, C.J., & Scalia, J., concurring in the judgment and dissenting in part) (citing Turner Broad. Sys., Inc., 512 U.S. at 664 (Kennedy J.)). Although Justice Kennedy did not cite his Turner Broadcasting System test in his Colorado Republican opinion, there is no reason to believe that he would object to the application of his test to campaign finance regulations. See Colorado Republican, 116 S. Ct. at 2322-23 (Kennedy, J., Rehnquist, C.J., & Scalia, J., concurring in the judgment and dissenting in part). There would seem to be only two members of the Court who disagree with the imposition of a duty on the government to demonstrate that campaign finance regulations address a real problem. See id. at 2332 (Stevens, J., and Ginsburg, J.) (urging "special deference to [Congress'] judgment on questions related to the extent and nature of limits on campaign spending."); see also id. at 2329 n.9 (Thomas, J.) (rejecting deference to Congress as amounting to "letting the fox stand watch over the henhouse."). In NTEU, a majority of the Court adopted the test stated by Justice Kennedy for a plurality in Turner Broadcasting System, and the courts in the Eighth Circuit subsequently invoked the NTEU statement of the government's duty to show some "real harm." See supra notes 88, 102 & 129 and accompanying text.

[155.] See Colorado Republican, 116 S. Ct. at 2317 (Breyer, J., O'Connor, & Souter, JJ.) (noting that "[t]he Government does not point to record evidence or legislative findings suggesting any special corruption problem in respect to independent party expenditures"); see id. at 2331 (Thomas, J., Rehnquist, C.J., & Scalia, J., concurring in the judgment and dissenting in part) (noting that "[t]he Government . . . has identified no more proof of the corrupting dangers of coordinated expenditures than it has of independent expenditures").

[156.] The lower courts, with only one exception, have not yet recognized that Colorado Republican imposes a duty on the states and the national government to demonstrate that campaign finance regulations address a real problem. See American Constitutional Law Found. v. Meyer, 120 F.3d 1092, 1105 (10th Cir. 1997) (state disclosure provisions not necessary to address any problem of fraud because state had not made any showing that existing regulations did not adequately cure this problem).

[157.] See supra text at notes 49-55.

[158.] The Court, albeit in the context of the Fourth Amendment, has recently reaffirmed the requirement that government must act on the basis of real, not hypothetical, problems. See Chandler v. Miller, 117 S. Ct. 1295, 1303 (1997) (holding that compulsory drug testing of candidates for certain state offices violated the Fourth Amendment where "[n]othing in the record hints that the hazards [the state] broadly describe[s] are real and not simply hypothetical."); see also Glickman v. Wileman Bros. & Elliot, Inc., 117 S. Ct. 2130, 2150 (1997) (Souter, J., dissenting) (under Turner Broadcasting System, Inc., 512 U.S. at 664, the government has an "obligation to establish the empirical reality of the problems it purports to be addressing" in order to justify an underinclusive regulation of commercial speech); cf. Timmons v. Twin Cities Area New Party, 117 S. Ct. 1364, 1372 (1997) (noting that "elaborate, empirical verification of the weightiness of the State's asserted justifications" is not required where an election law does not impose severe burdens on associational interests).