McConnell v. Federal Election Comm'n
540 U.S. 93 (2003)

Annotate this Case

OPINION OF KENNEDY, J.
MCCONNELL V. FEDERAL ELECTION COMM'N
540 U. S. ____ (2003)
SUPREME COURT OF THE UNITED STATES
NOS. 02-1674, 02-1675, 02-1676, 02-1702, 02-1727, 02-1733, 02-1734;02-1740, 02-1747, 02-1753, 02-1755, AND 02-1756

MITCH McCONNELL, UNITED STATES SENATOR, et al., APPELLANTS

02–1674   v.

FEDERAL ELECTION COMMISSION, et al.;

NATIONAL RIFLE ASSOCIATION, et al., APPELLANTS

02–1675   v.

FEDERAL ELECTION COMMISSION, et al.;

FEDERAL ELECTION COMMISSION, et al., APPELLANTS

02–1676   v.

MITCH McCONNELL, UNITED STATES SENATOR, et al.;

JOHN McCAIN, UNITED STATES SENATOR, et al., APPELLANTS

02–1702   v.

MITCH McCONNELL, UNITED STATES SENATOR, et al.;

REPUBLICAN NATIONAL COMMITTEE, et al., APPELLANTS

02–1727   v.

FEDERAL ELECTION COMMISSION, et al.;

NATIONAL RIGHT TO LIFE COMMITTEE, INC., et al., APPELLANTS

02–1733   v.

FEDERAL ELECTION COMMISSION, et al.;

AMERICAN CIVIL LIBERTIES UNION, APPELLANTS

02–1734   v.

FEDERAL ELECTION COMMISSION, et al.;

VICTORIA JACKSON GRAY ADAMS, et al., APPELLANTS

02–1740   v.

FEDERAL ELECTION COMMISSION, et al.;

RON PAUL, UNITED STATES CONGRESSMAN, et al., APPELLANTS

02–1747   v.

FEDERAL ELECTION COMMISSION, et al.;

CALIFORNIA DEMOCRATIC PARTY, et al., APPELLANTS

02–1753   v.

FEDERAL ELECTION COMMISSION, et al.;

AMERICAN FEDERATION OF LABOR AND CONGRESS OF INDUSTRIAL ORGANIZATIONS, et al., APPELLANTS

02–1755   v.

FEDERAL ELECTION COMMISSION, et al.;

CHAMBER OF COMMERCE OF THE UNITED STATES, et al., APPELLANTS

02–1756   v.

FEDERAL ELECTION COMMISSION, et al.

on appeals from the united states district court for the district of columbia

[December 10, 2003]

   Justice Kennedy, concurring in the judgment in part and dissenting in part with respect to BCRA Titles I and II.*

   The First Amendment guarantees our citizens the right to judge for themselves the most effective means for the expression of political views and to decide for themselves which entities to trust as reliable speakers. Significant portions of Titles I and II of the Bipartisan Campaign Reform Act of 2002 (BCRA or Act) constrain that freedom. These new laws force speakers to abandon their own preference for speaking through parties and organizations. And they provide safe harbor to the mainstream press, suggesting that the corporate media alone suffice to alleviate the burdens the Act places on the rights and freedoms of ordinary citizens.

   Today’s decision upholding these laws purports simply to follow Buckley v. Valeo,424 U. S. 1 (1976) (per curiam), and to abide by stare decisis, see ante, at 27 (joint opinion of Stevens and O’Connor, JJ. (hereinafter Court or majority)); but the majority, to make its decision work, must abridge free speech where Buckley did not. Buckley did not authorize Congress to decide what shapes and forms the national political dialogue is to take. To reach today’s decision, the Court surpasses Buckley’s limits and expands Congress’ regulatory power. In so doing, it replaces discrete and respected First Amendment principles with new, amorphous, and unsound rules, rules which dismantle basic protections for speech.

   A few examples show how BCRA reorders speech rights and codifies the Government’s own preferences for certain speakers. BCRA would have imposed felony punishment on Ross Perot’s 1996 efforts to build the Reform Party. Compare Federal Election Campaign Act of 1971 (FECA) §§309(d)(1)(A), 315(a)(1)(B), and 323(a)(1) (prohibiting, by up to five years’ imprisonment, any individual from giving over $25,000 annually to a national party), with Spending By Perot, The Houston Chronicle, Dec. 13, 1996, p. 43 (reporting Perot’s $8 million founding contribution to the Reform Party). BCRA makes it a felony for an environmental group to broadcast an ad, within 60 days of an election, exhorting the public to protest a Congressman’s impending vote to permit logging in national forests. See BCRA §203. BCRA escalates Congress’ discrimination in favor of the speech rights of giant media corporations and against the speech rights of other corporations, both profit and nonprofit. Compare BCRA §203, with Austin v. Michigan Chamber of Commerce,494 U. S. 652, 659–660 (1990) (first sanctioning this type of discrimination).

   To the majority, all this is not only valid under the First Amendment but also is part of Congress’ “steady improvement of the national election laws.” Ante, at 6. We should make no mistake. It is neither. It is the codification of an assumption that the mainstream media alone can protect freedom of speech. It is an effort by Congress to ensure that civic discourse takes place only through the modes of its choosing. And BCRA is only the beginning, as its congressional proponents freely admit:

“This is a modest step, it is a first step, it is an essential step, but it does not even begin to address, in some ways, the fundamental problems that exist with the hard money aspect of the system.” 148 Cong. Rec. S2101 (Mar. 20, 2002) (statement of Sen. Feingold).

Id., at S2097 (statement of Sen. Wellstone) (“[P]assing this legislation … will whet people’s appetite for more”); id., at S2101 (statement of Sen. Boxer) (“[T]his bill is not the be-all or the end-all, but it is a strong start”); id., at S2152 (statement of Sen. Corzine) (“[T]his should not and will not be the last time campaign finance reform is debated on the Senate floor. We have many more important campaign finance issues to explore”); id., at S2157 (statement of Sen. Torricelli) (“Make [BCRA] the beginning of a reform, not the end of reform”); id., at H442 (Feb. 13, 2002) (statement of Rep. Doggett) (“Mr. Chairman, if [BCRA] has any defect, it is that it does too little, not too much”).

   Our precedents teach, above all, that Government cannot be trusted to moderate its own rules for suppression of speech. The dangers posed by speech regulations have led the Court to insist upon principled constitutional lines and a rigorous standard of review. The majority now abandons these distinctions and limitations.

   With respect, I dissent from the majority opinion upholding BCRA Titles I and II. I concur in the judgment as to BCRA §213 and new FECA §323(e) and concur in the judgment in part and dissent in part as to BCRA §§201, 202, and 214.

I. TITLE I AND COORDINATION PROVISIONS

   Title I principally bans the solicitation, receipt, transfer and spending of soft money by the national parties (new FECA §323(a), 2 U. S. C. A. §441i(a) (Supp. 2003)). It also bans certain uses of soft money by state parties (new FECA §323(b)); the transfer of soft money from national parties to nonprofit groups (new FECA §323(d)); the solicitation, receipt, transfer, and spending of soft money by federal candidates and officeholders (new FECA §323(e)); and certain uses of soft money by state candidates (new FECA §323(f)). These provisions, and the other provisions with which this opinion is principally concerned, are set out in full, see Appendix, infra. Even a cursory review of the speech and association burdens these laws create makes their First Amendment infirmities obvious:

   Title I bars individuals with shared beliefs from pooling their money above limits set by Congress to form a new third party. See new FECA §323(a).

   Title I bars national party officials from soliciting or directing soft money to state parties for use on a state ballot initiative. This is true even if no federal office appears on the same ballot as the state initiative. See new FECA §323(a).

   A national party’s mere involvement in the strategic planning of fundraising for a state ballot initiative risks a determination that the national party is exercising “indirect control” of the state party. If that determination is made, the state party must abide by federal regulations. And this is so even if the federal candidate on the ballot, if there is one, runs unopposed or is so certain of election that the only voter interest is in the state and local campaigns. See new FECA §323(a).

   Title I compels speech. Party officials who want to engage in activity such as fundraising must now speak magic words to ensure the solicitation cannot be interpreted as anything other than a solicitation for hard, not soft, money. See ibid.

   Title I prohibits the national parties from giving any sort of funds to nonprofit entities, even federally regulated hard money, and even if the party hoped to sponsor the interest group’s exploration of a particular issue in advance of the party’s addition of it to their platform. See new FECA §323(d).

   By express terms, Title I imposes multiple different forms of spending caps on parties, candidates, and their agents. See new FECA §§323(a), (e), and (f).

   Title I allows state parties to raise quasi-soft money Levin funds for use in activities that might affect a federal election; but the Act prohibits national parties from assisting state parties in developing and executing these fundraising plans, even when the parties seek only to advance state election interests. See new FECA §323(b).

   Until today’s consolidated cases, the Court has accepted but two principles to use in determining the validity of campaign finance restrictions. First is the anticorruption rationale. The principal concern, of course, is the agreement for a quid pro quo between officeholders (or candidates) and those who would seek to influence them. The Court has said the interest in preventing corruption allows limitations on receipt of the quid by a candidate or officeholder, regardless of who gives it or of the intent of the donor or officeholder. See Buckley, 424 U. S., at 26– 27, 45–48; infra, at  7–10. Second, the Court has analyzed laws that classify on the basis of the speaker’s corporate or union identity under the corporate speech rationale. The Court has said that the willing adoption of the entity form by corporations and unions justifies regulating them differently: Their ability to give candidates quids may be subject not only to limits but also to outright bans; their electoral speech may likewise be curtailed. See Austin, 494 U. S., at 659–660; Federal Election Comm’n v. National Right to Work Comm.,459 U. S. 197, 201–211 (1982).

   The majority today opens with rhetoric that suggests a conflation of the anticorruption rationale with the corporate speech rationale. See ante, at 3–6 (hearkening back to, among others, Elihu Root and his advocacy against the use of corporate funds in political campaigning). The conflation appears designed to cast the speech regulated here as unseemly corporate speech. The effort, however, is unwarranted, and not just because money is not per se the evil the majority thinks. Most of the regulations at issue, notably all of the Title I soft money bans and the Title II coordination provisions, do not draw distinctions based on corporate or union status. Referring to the corporate speech rationale as if it were the linchpin of the case, when corporate speech is not primarily at issue, adds no force to the Court’s analysis. Instead, the focus must be on Buckley’s anticorruption rationale and the First Amendment rights of individual citizens.

A. Constitutionally Sufficient Interest

   In Buckley, the Court held that one, and only one, interest justified the significant burden on the right of association involved there: eliminating, or preventing, actual corruption or the appearance of corruption stemming from contributions to candidates.

“It is unnecessary to look beyond the Act’s primary purpose—to limit the actuality and appearance of corruption resulting from large individual financial contributions—in order to find a constitutionally sufficient justification for the $1,000 contribution limitation.” 424 U. S., at 26.

See also ibid. (concluding this corruption interest was sufficiently “significant” to sustain “closely drawn” interference with protected First Amendment rights).

   In parallel, Buckley concluded the expenditure limitations in question were invalid because they did not advance that same interest. See id., at 47–48 (“[T]he independent expenditure ceiling thus fails to serve any substantial governmental interest in stemming the reality or appearance of corruption in the electoral process”); see also id., at 45, 46.

   Thus, though Buckley subjected expenditure limits to strict scrutiny and contribution limits to less exacting review, itheld neither could withstand constitutional challenge unless it was shown to advance the anticorruption interest. In these consolidated cases, unless Buckley is to be repudiated, we must conclude that the regulations further that interest before considering whether they are closely drawn or narrowly tailored. If the interest is not advanced, the regulations cannot comport with the Constitution, quite apart from the standard of review.

   Buckley made clear, by its express language and its context, that the corruption interest only justifies regulating candidates’ and officeholders’ receipt of what we can call the “quids” in the quid pro quo formulation. The Court rested its decision on the principle that campaign finance regulation that restricts speech without requiring proof of particular corrupt action withstands constitutional challenge only if it regulates conduct posing a demonstrable quid pro quo danger:

“To the extent that large contributions are given to secure a political quid pro quo from current and potential office holders, the integrity of our system of representative democracy is undermined.” Id., at 26–27.

See also id., at 45 (“[A]ssuming, arguendo, that large independent expenditures pose the same dangers of actual or apparent quid pro quo arrangements as do large contributions …”). That Buckley rested its decision on this quid pro quo standard is not a novel observation. We have held this was the case:

“The exception [of contribution limits being justified under the First Amendment] relates to the perception of undue influence of large contributions to a candidate: ‘ To the extent that large contributions are given to secure a political quid pro quo from current and potential office holders, the integrity of our system of representative democracy is undermined.’ ” Citizens Against Rent Control/Coalition for Fair Housing v. Berkeley,454 U. S. 290, 297 (1981) (quoting Buckley, supra, at 26–27).

See also Federal Election Comm’n v. Beaumont, 539 U. S. ___ (2003) (furthering this anticorruption rationale by upholding limits on contributions given directly to candidates); Nixon v. Shrink Missouri Government PAC, 528 U. S. 377 (2000) (same).

   Despite the Court’s attempt to rely on language from cases like Shrink Missouri to establish that the standard defining corruption is broader than conduct that presents a quid pro quo danger, see ante, at 43, n. 48, in those cases the Court in fact upheld limits on conduct possessing quid pro quo dangers, and nothing more. See also infra, 12. For example, the Shrink Missouri Court’s distinguishing of what was at issue there and quid pro quo, in fact, shows only that it used the term quid pro quo to refer to actual corrupt, vote-buying exchanges, as opposed to interactions that possessed quid pro quo potential even if innocently undertaken. Thus, the Court said:

“[W]e spoke in Buckley of the perception of corruption ‘inherent in a regime of large individual financial contributions’ to candidates for public office . . . as a source of concern “almost equal” to quid pro quo improbity.” 528 U. S., at 390 (citations omitted).

Thus, the perception of corruption that the majority now asserts is somehow different from the quid pro quo potential discussed in this opinion, was created by an exchange featuring quid pro quo potential—contributions directly to a candidate.

   In determining whether conduct poses a quid pro quo danger the analysis is functional. In Buckley, the Court confronted an expenditure limitation provision that capped the amount of money individuals could spend on any activity intended to influence a federal election (i.e., it reached to both independent and coordinated expenditures). See 424 U. S., at 46–47. The Court concluded that though the limitation reached both coordinated and independent expenditures, there were other valid FECA provisions that barred coordinated expenditures. Hence, the limit at issue only added regulation to independent expenditures. On that basis it concluded the provision was unsupported by any valid corruption interest. The conduct to which it added regulation (independent expenditures) posed no quid pro quo danger. See ibid.

   Placing Buckley’s anticorruption rationale in the context of the federal legislative power yields the following rule: Congress’ interest in preventing corruption provides a basis for regulating federal candidates’ and officeholders’ receipt of quids, whether or not the candidate or officeholder corruptly received them. Conversely, the rule requires the Court to strike down campaign finance regulations when they do not add regulation to “actual or apparent quid pro quo arrangements.” Id., at 45.

   The Court ignores these constitutional bounds and in effect interprets the anticorruption rationale to allow regulation not just of “actual or apparent quid pro quo arrangements,” ibid., but of any conduct that wins goodwill from or influences a Member of Congress. It is not that there is any quarrel between this opinion and the majority that the inquiry since Buckley has been whether certain conduct creates “undue influence.” See ante, at 40–41. On that we agree. The very aim of Buckley’s standard, however, was to define undue influence by reference to the presence of quid pro quo involving the officeholder. The Court, in contrast, concludes that access, without more, proves influence is undue. Access, in the Court’s view, has the same legal ramifications as actual or apparent corruption of officeholders. This new definition of corruption sweeps away all protections for speech that lie in its path.

   The majority says it is not abandoning our cases in this way, but its reasoning shows otherwise:

   “More importantly, plaintiffs conceive of corruption too narrowly. Our cases have firmly established that Congress’ legitimate interest extends beyond preventing simple cash-for-votes corruption to curbing ‘undue influence on an officeholder’s judgment, or the appearance of such influence.’ [Federal Election Comm’n v. Colorado Republican Federal Campaign Comm.,533 U. S. 431, 441 (2001) (Colorado II)]. Many of the ‘deeply disturbing examples’ of corruption cited by this Court in Buckley to justify FECA’s contribution limits were not episodes of vote buying, but evidence that various corporate interests had given substantial donations to gain access to high-level government officials. Even if that access did not secure actual influence, it certainly gave the ‘appearance of such influence.’ Colorado II, supra, at 441; see also [Buckley v. Valeo, 519 F. 2d 821, 838 (CADC 1975)].

   “The record in the present case is replete with similar examples of national party committees peddling access to federal candidates and officeholders in exchange for large soft-money donations. See [251 F. Supp. 2d 176, 492–506 (DC 2003) (Kollar-Kotelly, J.)].” Ante, at 40–41.

   The majority notes that access flowed from the regulated conduct at issue in Buckley and its progeny, then uses that fact as the basis for concluding that access peddling by the parties equals corruption by the candidates. That conclusion, however, is tenable only by a quick and subtle shift, and one that breaks new ground: The majority ignores the quid pro quo nature of the regulated conduct central to our earlier decisions. It relies instead solely on the fact that access flowed from the conduct.

   To ignore the fact that in Buckley the money at issue was given to candidates, creating an obvious quid pro quo danger as much as it led to the candidates also providing access to the donors, is to ignore the Court’s comments in Buckley that show quid pro quo was of central importance to the analysis. See 424 U. S., at 26–27, 45. The majority also ignores that in Buckley, and ever since,those party contributions that have been subject to congressional limit were not general party-building contributions but were only contributions used to influence particular elections. That is, they were contributions that flowed to a particular candidate’s benefit, again posing a quid pro quo danger. And it ignores that in Colorado II, the party spending was that which was coordinated with a particular candidate, thereby implicating quid pro quo dangers. In all of these ways the majority breaks the necessary tether between quid and access and assumes that access, all by itself, demonstrates corruption and so can support regulation. See also ante, at 47 (“[L]arge soft-money donations to national party committees are likely to buy donors preferential access to federal officeholders no matter the ends to which their contributions are eventually put”).

   Access in itself, however, shows only that in a general sense an officeholder favors someone or that someone has influence on the officeholder. There is no basis, in law or in fact, to say favoritism or influence in general is the same as corrupt favoritism or influence in particular. By equating vague and generic claims of favoritism or influence with actual or apparent corruption, the Court adopts a definition of corruption that dismantles basic First Amendment rules, permits Congress to suppress speech in the absence of a quid pro quo threat, and moves beyond the rationale that is Buckley’s very foundation.

   The generic favoritism or influence theory articulated by the Court is at odds with standard First Amendment analyses because it is unbounded and susceptible to no limiting principle. Any given action might be favored by any given person, so by the Court’s reasoning political loyalty of the purest sort can be prohibited. There is no remaining principled method for inquiring whether a campaign finance regulation does in fact regulate corruption in a serious and meaningful way. We are left to defer to a congressional conclusion that certain conduct creates favoritism or influence.

   Though the majority cites common sense as the foundation for its definition of corruption, see ante, at 35, 43, in the context of the real world only a single definition of corruption has been found to identify political corruption successfully and to distinguish good political responsiveness from bad—that is quid pro quo. Favoritism and influence are not, as the Government’s theory suggests, avoidable in representative politics. It is in the nature of an elected representative to favor certain policies, and, by necessary corollary, to favor the voters and contributors who support those policies. It is well understood that a substantial and legitimate reason, if not the only reason, to cast a vote for, or to make a contribution to, one candidate over another is that the candidate will respond by producing those political outcomes the supporter favors. Democracy is premised on responsiveness. Quid pro quo corruption has been, until now, the only agreed upon conduct that represents the bad form of responsiveness and presents a justiciable standard with a relatively clear limiting principle: Bad responsiveness may be demonstrated by pointing to a relationship between an official and a quid.

   The majority attempts to mask its extension of Buckley under claims that BCRA prevents the appearance of corruption, even if it does not prevent actual corruption, since some assert that any donation of money to a political party is suspect. See ante, at 40–42. Under Buckley’s holding that Congress has a valid “interest in stemming the reality or appearance of corruption,” 424 U. S., at 47–48, however, the inquiry does not turn on whether some persons assert that an appearance of corruption exists. Rather, the inquiry turns on whether the Legislature has established that the regulated conduct has inherent corruption potential, thus justifying the inference that regulating the conduct will stem the appearance of real corruption. Buckley was guided and constrained by this analysis. In striking down expenditure limits the Court in Buckley did not ask whether people thought large election expenditures corrupt, because clearly at that time many persons, including a majority of Congress and the President, did. See id., at 25 (“According to the parties and amici, the primary interest served . . . by the Act as a whole, is the prevention of corruption and the appearance of corruption”). Instead, the Court asked whether the Government had proved that the regulated conduct, the expenditures, posed inherent quid pro quo corruption potential. See id., at 46.

   The Buckley decision made this analysis even clearer in upholding contribution limitations. It stated that even if actual corrupt contribution practices had not been proved, Congress had an interest in regulating the appearance of corruption that is “inherent in a regime of large individual financial contributions.” Id., at 27 (discussing contributions tocandidates). See also id., at 28, 30. The quid pro quo nature of candidate contributions justified the conclusion that the contributions pose inherent corruption potential; and this in turn justified the conclusion that their regulation would stem the appearance of real corruption.

   From that it follows that the Court today should not ask, as it does, whether some persons, even Members of Congress, conclusorily assert that the regulated conduct appears corrupt to them. Following Buckley, it should instead inquire whether the conduct now prohibited inherently poses a real or substantive quid pro quo danger, so that its regulation will stem the appearance of quid pro quo corruption.

1. New FECA §§323(a), (b), (d), and (f)

   Sections 323(a), (b), (d), and (f), 2 U. S. C. A. §§441i(a), (b), (d), and (f ) (Supp. 2003), cannot stand because they do not add regulation to conduct that poses a demonstrable quid pro quo danger. They do not further Buckley’s corruption interest.

   The majority, with a broad brush, paints §323(a) as aimed at limiting contributions possessing federal officeholder corruption potential. From there it would justify §323’s remaining provisions as necessary complements to ensure the national parties cannot circumvent §323(a)’s prohibitions. The broad brush approach fails, however, when the provisions are reviewed under Buckley’s proper definition of corruption potential.

   On its face §323(a) does not regulate federal candidates’ or officeholders’ receipt of quids because it does not regulate contributions to, or conduct by, candidates or officeholders. See BCRA §101(a) (setting out new FECA §323(a): National parties may not “solicit, receive, or direct to another person … or spend any [soft money]”).

   The realities that underlie the statute, furthermore, do not support the majority’s interpretation. Before BCRA’s enactment, parties could only use soft money for a candidate’s “benefit” (e.g., through issue ads, which all parties now admit may influence elections) independent of that candidate. And, as discussed later, §323(e) validly prohibits federal candidate and officeholder solicitation of soft money party donations. See infra, at 31. Section 323(a), therefore, only adds regulation to soft money party donations not solicited by, or spent in coordination with, a candidate or officeholder.

   These donations (noncandidate or officeholder solicited soft money party donations that are independently spent) do not pose the quid pro quo dangers that provide the basis for restricting protected speech. Though the government argues §323(a) does regulate federal candidates’ and officeholders’ receipt of quids, it bases its argument on this flawed reasoning:

   (1) “[F]ederal elected officeholders are inextricably linked to their political parties,” Brief for Appellees/Cross Appellants FEC et al. in No. 02–1674 et al., p. 21; cf. Colorado Republican Federal Campaign Comm. v. Federal Election Comm’n,518 U. S. 604, 626 (1996) (Colorado I) (Kennedy, J., concurring in judgment and dissenting in part).

   (2) All party receipts must be connected to, and must create, corrupt donor favoritism among these officeholders.

   (3) Therefore, regulation of party receipts equals regulation of quids to the party’s officeholders.

   The reasoning is flawed because the Government’s reliance on reasoning parallel to the Colorado I concurrence only establishes the first step in its chain of logic: that a party is a proxy for its candidates generally. It does not establish the second step: that as a proxy for its candidates generally, all moneys the party receives (not just candidate solicited-soft money donations, or donations used in coordinated activity) represent quids for all the party’s candidates and officeholders. The Government’s analysis is inconsistent with what a majority of the Justices, in different opinions, have said.

   Justice Thomas’ dissent in Colorado II, 533 U. S., at 476–477, taken together with Justice Breyer’s opinion announcing the judgment of the Court in Colorado I, rebuts the second step of the Government’s argument. Justice Thomas demonstrated that a general party-candidate corruption linkage does not exist. As he pointed out:

“The dearth of evidence [of such corruption] is unsurprising in light of the unique relationship between a political party and its candidates: ‘The very aim of a political party is to influence its candidate’s stance on issues and, if the candidate takes office or is reelected, his votes.’ If coordinated expenditures help achieve this aim, the achievement ‘does not … constitute “a subversion of the political process.” ’ ” Colorado II, supra, at 476–477 (citations omitted).

   Justice Breyer reached the same conclusion about the corrupting effect general party receipts could have on particular candidates, though on narrower grounds. He concluded that independentparty conduct lacks quid pro quo corruption potential. See Colorado I, 518 U. S., at 617–618; id., at 617 (“If anything, an independent [party] expenditure made possible by a $20,000 donation, but controlled and directed by a party rather than the donor, would seem less likely to corrupt than the same (or a much larger) independent expenditure made directly by that donor”); id., at 616 (“[T]he opportunity for corruption posed by [soft money] contributions is, at best, attenuated” because they may not be used for the purposes of influencing a federal election under FECA).

   These opinions establish that independent party activity, which by definition includes independent receipt and spending of soft money, lacks a possibility for quid pro quo corruption of federal officeholders. This must be all the more true of a party’s independent receipt and spending of soft money donations neither directed to nor solicited by a candidate.

   The Government’s premise is also unsupported by the record before us. The record confirms that soft money party contributions, without more, do not create quid pro quo corruption potential. As a conceptual matter, generic party contributions may engender good will from a can- didate or officeholder because, as the Government says: “[A] Member of Congress can be expected to feel a natural temptation to favor those persons who have helped the ‘team,’ ” Brief for Appellees/Cross-Appellants FEC et al. in No. 02–1674 et al., p. 33. Still, no Member of Congress testified this favoritism changed voting behavior.

   The piece of record evidence the Government puts forward on this score comes by way of deposition testimony from former Senator Simon and Senator Feingold. See251 F. Supp. 2d 176, 482 (DC 2003) (Kollar-Kotelly, J.). Senator Simon reported an unidentified colleague indicated frustration with Simon’s opposition to legislation that would benefit a party contributor on the grounds that “ ‘we’ve got to pay attention to who is buttering our bread’ ” and testified he did not think there was any question “ ‘this’ ” (i.e., “donors getting their way”) was why the legislation passed. See App. 805. Senator Feingold, too, testified an unidentified colleague suggested he support the legislation because “ ‘they [i.e., the donor] just gave us [i.e., the party] $100,000.’ ” 251 F. Supp. 2d, at 482 (Kollar-Kotelly, J.).

   That evidence in fact works against the Government. These two testifying Senators expressed disgust toward the favoring of a soft money giver, and not the good will one would have expected under the Government’s theory. That necessarily undercuts the inference of corruption the Government would have us draw from the evidence.

   Even more damaging to the Government’s argument from the testimony is the absence of testimony that the Senator who allegedly succumbed to corrupt influence had himself solicited soft money from the donor in question. Equally, there is no indication he simply favored the company with his vote because it had, without any involvement from him, given funds to the party to which he belonged. This fact is crucial. If the Senator himself had been the solicitor of the soft money funds in question, the incident does nothing more than confirm that Congress’ efforts at campaign finance reform ought to be directed to conduct that implicates quid pro quo relationships. Only if there was some evidence that the officeholder had not solicited funds from the donor could the Court extrapolate from this episode that general party contributions function as quids, inspiring corrupt favoritism among party members. The episode is the single one of its type reported in the record and does not seem sufficient basis for major incursions into settled practice. Given the Government’s claim that the corrupt favoritism problem is widespread, its inability to produce more than a single instance purporting to illustrate the point demonstrates the Government has not fairly characterized the general attitudes of Members towards soft money donors from whom they have not solicited.

   Other aspects of the record confirm the Government has not produced evidence that Members corruptly favor soft money donors to their party as a per se matter. Most testimony from which the Government would have the Court infer corruption is testimony that Members are rewarded by their parties for soliciting soft money. See id., at 438–521 (Kollar-Kotelly, J.). This says nothing about how Members feel about a party’s soft money donors from whom they have not solicited. Indeed, record evidence on this point again cuts against the Government:

“ ‘As a Member of the Senate Finance Committee, I experienced the pressure first hand. On several occasions when we were debating important tax bills, I needed a police escort to get into the Finance Committee hearing room because so many lobbyists were crowding the halls, trying to get one last chance to make their pitch to each Senator. Senators generally knew which lobbyist represented the interests of which large donor. I was often glad that I limited the amount of soft money fundraising I did and did not take PAC contributions, because it would be extremely difficult not to feel beholden to these donors otherwise.’ ” Id., at 482 (testimony of former Senator Boren; see 6–R Defs. Exhs., Tab 8, ¶ ;8).

Thus, one of the handful of Senators on whom the Government relies to make its case candidly admits the pressure of appeasing soft money donors derives from the Members’ solicitation of donors, not from those donors’ otherwise giving to their party.

   In light of all this, §323(a) has no valid anticorruption interest. The anticircumvention interests the Government offers in defense of §§323(b), (d), and (f) must also fall with the interests asserted to justify §323(a). Any anticircumvention interest can be only as compelling as the interest justifying the underlying regulation.

   None of these other sections has an independent justifying interest. Section 323(b), for example, adds regulation only to activity undertaken by a state party. In the District Court two of the three judges found as fact that particular state and local parties exist primarily to participate in state and local elections, that they spend the majority of their resources on those elections, and that their voter registration and Get Out The Vote (GOTV) activities, in particular, are directed primarily at state and local elections. See 251 F. Supp. 2d, at 301–302 (Henderson, J., concurring in judgment in part and dissenting in part); id., at 837–840 (Leon, J.). These findings, taken together with BCRA’s other, valid prohibitions barring coordination with federal candidates or officeholders and their soft money solicitation, demonstrate that §323(b) does not add regulation to conduct that poses a danger of a federal candidate’s or officeholder’s receipt of quids.

   Even §323(b)’s narrowest regulation, which bans state party soft money funded ads that (1) refer to a clearly identified federal candidate, and (2) either support or attack any candidate for the office of the clearly mentioned federal candidate, see new FECA §301(20)(A)(iii), fails the constitutional test. The ban on conduct that by the statute’s own definition may serve the interest of a federal candidate suggests to the majority that it is conduct that poses quid pro quo danger for federal candidates or officeholders. Yet, even this effect—considered after excising the coordination and candidate-solicited funding aspects elsewhere prohibited by BCRA §§202 and 214(a) and new FECA §323(a)—poses no danger of a federal candidate’s or officeholder’s receipt of a quid. That conduct is no different from an individual’s independent expenditure referring to and supporting a clearly identified candidate—and this poses no regulable danger.

   Section 323(d), which governs relationships between the national parties and nonprofit groups, fails for similar reasons. It is worth noting that neither the record nor our own experience tells us how significant these funds transfers are at this time. It is plain, however, that the First Amendment ought not to be manipulated to permit Congress to forbid a political party from aiding other speakers whom the party deems more effective in addressing discrete issues. One of the central flaws in BCRA is that Congress is determining what future course the creation of ideas and the expression of views must follow. Its attempt to foreclose new and creative partnerships for speech, as illustrated here, is consistent with neither the traditions nor principles of our Free Speech guarantee, which insists that the people, and not the Congress, decide what modes of expression are the most legitimate and effective.

   The majority’s upholding §323(d) is all the more unsettling because of the way it ignores the Act as Congress wrote it. Congress said national parties “shall not solicit any funds for, or make or direct any donations to” §501(c) nonprofit organizations that engage in federal election activity or to §527 political committees. The Court, however, reads out the word “any” and construes the words “funds” and “donations” to mean “soft money funds” and “soft money donations.” See ante, at 72 (“This construction is consistent with the concerns animating Title I, whose purpose is to plug the soft-money loophole”). The Court’s statutory amendment may be consistent with its anti-soft-money rationale; it is not, however, consistent with the plain and unavoidable statutory text Congress has given us. Even as construed by the Court, moreover, it is invalid.

   The majority strains to save the provision from what must seem to it an unduly harsh First Amendment. It does so by making a legislative determination Congress chose not to make: to prefer hard money to soft money within the construct of national party relationships with nonprofit groups. Congress gave no indication of a preference to regulate either hard money or soft in this context. Rather, it simply proscribed all transfers of money between the two organizations and all efforts by the national parties to raise any money on the nonprofit groups’ behalf. The question the Court faces is not which part of a text to sever and strike, but whether Congress can prohibit such transfers altogether. The answer, as the majority recognizes, is no. See ante, at 71 (“[P]rohibiting parties from donating funds already raised in compliance with FECA does little to further Congress’ goal of preventing corruption or the appearance of corruption of federal candidates and officeholders”).

   Though §323(f) in effect imposes limits on candidate contributions, it does not address federal candidate and officeholder contributions. Yet it is the possibility of federal officeholder quid pro quo corruption potential that animates Buckley’s rule as it relates to Acts of Congress (as opposed to Acts of state legislatures). See 424 U. S., at 13 (“The constitutional power of Congress to regulate federal elections is well established”).

   When one recognizes that §§323(a), (b), (d), and (f) do not serve the interest the anticorruption rationale contemplates, Title I’s entirety begins to look very much like an incumbency protection plan. See J. Miller, Monopoly Politics 84–101 (1999) (concluding that regulations limiting election fundraising and spending constrain challengers more than incumbents). That impression is worsened by the fact that Congress exempted its officeholders from the more stringent prohibitions imposed on party officials. Compare new FECA §323(a) with new FECA §323(e). Section 323(a) raises an inflexible bar against soft money solicitation, in any way, by parties or party officials. Section 323(e), in contrast, enacts exceptions to the rule for federal officeholders (the very centerpiece of possible corruption), and allows them to solicit soft money for various uses and organizations.

   The law in some respects even weakens the regulation of federal candidates and officeholders. Under former law, officeholders were understood to be limited to receipt of hard money by their campaign committees. See 2 U. S. C. §§431, 441a (setting out the pre-BCRA FECA regime). BCRA, however, now allows them and their campaign committees to receive soft money that fits the hard money source and amount restrictions, so long as the officeholders direct that money on to other nonfederal candidates. See new FECA §323(e)(1)(B). The majority’s characterization of this weakening of the regime as “tightly constrain[ing]” candidates, ante, at 73, n. 70, is a prime example of its unwillingness to confront Congress’ own interest or the persisting fact that the regulations violate First Amendment freedoms. The more lenient treatment accorded to incumbency-driven politicians than to party officials who represent broad national constituencies must render all the more suspect Congress’ claim that the Act’s sole purpose is to stop corruption.

   The majority answers this charge by stating the obvious, that “§323(e) applies to both officeholders and candidates.” Ante, at 78, n. 72. The controlling point, of course, is the practical burden on challengers. That the prohibition applies to both incumbents and challengers in no way establishes that it burdens them equally in that regard. Name recognition and other advantages held by incumbents ensure that as a general rule incumbents will be advantaged by the legislation the Court today upholds.

   The Government identifies no valid anticorruption interest justifying §§323(a), (b), (d), and (f). The very nature of the restrictions imposed by these provisions makes one all the more skeptical of the Court’s explanation of the interests at stake. These provisions cannot stand under the First Amendment.

2. New FECA §323(e)

   Ultimately, only one of the challenged Title I provisions satisfies Buckley’s anticorruption rationale and the First Amendment’s guarantee. It is §323(e). This provision is the sole aspect of Title I that is a direct and necessary regulation of federal candidates’ and officeholders’ receipt of quids. Section 323(e) governs “candidate[s], individual[s] holding Federal office, agent[s] of a candidate or an individual holding Federal office, or an entity directly or indirectly established, financed, maintained or controlled by or acting on behalf of 1 or more candidates or individuals holding Federal office.” 2 U. S. C. A. §441i(e) (Supp. 2003). These provisions, and the regulations that follow, limit candidates’ and their agents’ solicitation of soft money. The regulation of a candidate’s receipt of funds furthers a constitutionally sufficient interest. More difficult, however, is the question whether regulation of a candidate’s solicitation of funds also furthers this interest if the funds are given to another.

   I agree with the Court that the broader solicitation regulation does further a sufficient interest. The making of a solicited gift is a quid both to the recipient of the money and to the one who solicits the payment (by granting his request). Rules governing candidates’ or officeholders’ solicitation of contributions are, therefore, regulations governing their receipt of quids. This regulation fits under Buckley’s anticorruption rationale.

B. Standard of Review

   It is common ground between the majority and this opinion that a speech-suppressing campaign finance regulation, even if supported by a sufficient Government interest, is unlawful if it cannot satisfy our designated standard of review. See ante, at 24–27. In Buckley, we applied “closely drawn” scrutiny to contribution limitations and strict scrutiny to expenditure limitations. Compare 424 U. S., at 25, with id., at 44–45. Against that backdrop, the majority assumes that because Buckley applied the rationale in the context of contribution and expenditure limits, its application gives Congress and the Court the capacity to classify any challenged campaign finance regulation as either a contribution or an expenditure limit. Thus, it first concludes Title I’s regulations are contributions limits and then proceeds to apply the lesser scrutiny.

“Complex as its provisions may be, §323, in the main, does little more than regulate the ability of wealthy individuals, corporations, and unions to contribute large sums of money to influence federal elections, federal candidates, and federal officeholders.” Ante, at 28.

   Though the majority’s analysis denies it, Title I’s dynamics defy this facile, initial classification.

   Title I’s provisions prohibit the receipt of funds; and in most instances, but not all, this can be defined as a contribution limit. They prohibit the spending of funds; and in most instances this can be defined as an expenditure limit. They prohibit the giving of funds to nonprofit groups; and this falls within neither definition as we have ever defined it. Finally, they prohibit fundraising activity; and the parties dispute the classification of this regulation (the challengers say it is core political association, while the Government says it ultimately results only in a limit on contribution receipts).

   The majority’s classification overlooks these competing characteristics and exchanges Buckley’s substance for a formulaic caricature of it. Despite the parties’ and the majority’s best efforts on both sides of the question, it ignores reality to force these regulations into one of the two legal categories as either contribution or expenditure limitations. Instead, these characteristics seem to indicate Congress has enacted regulations that are neither contribution nor expenditure limits, or are perhaps both at once.

   Even if the laws could be classified in broad terms as only contribution limits, as the majority is inclined to do, that still leaves the question what “contribution limits” can include if they are to be upheld under Buckley. Buckley’s application of a less exacting review to contribution limits must be confined to the narrow category of money gifts that are directed, in some manner, to a candidate or officeholder. Any broader definition of the category contradicts Buckley’s quid pro quo rationale and overlooks Buckley’s language, which contemplates limits on contributions to a candidate or campaign committee in explicit terms. See424 U. S., at 13 (applying less exacting review to “contribution … limitations in the Act prohibit[ing] individuals from contributing more than $25,000 in a single year or more than $1,000 to any single candidate for an election campaign”); id., at 45 (“[T]he contribution limitations’ [apply a] total ban on the giving of large amounts of money to candidates”). See also id., at 20, 25, 28.

   The Court, it must be acknowledged, both in Buckley and on other occasions, has described contribution limits due some more deferential review in less than precise terms. At times it implied that donations to political parties would also qualify as contributions whose limitation too would be subject to less exacting review. See id., at 23–24, n. 24 (“[T]he general understanding of what constitutes a political contribution[:] Funds provided to a candidate or political party or campaign committee either directly or indirectly through an intermediary constitute a contribution”). See also Federal Election Comm’n v. Beaumont, 539 U. S., at ___ (2003) (slip op., at 14) (“ ‘[C]ontributions may result in political expression if spent by a candidate or an association’ ”).

   These seemingly conflicting statements are best reconciled by reference to Buckley’s underlying rationale for applying less exacting review. In a similar, but more imperative, sense proper application of the standard of review to regulations that are neither contribution nor expenditure limits (or which are both at once) can only be determined by reference to that rationale.

   Buckley’s underlying rationale is this: Less exacting review applies to Government regulations that “significantly interfere” with First Amendment rights of association. But any regulation of speech or associational rights creating “markedly greater interference” than such significant interference receives strict scrutiny. Unworkable and ill advised though it may be, Buckley unavoidably sets forth this test:

“Even a ‘ “significant interference” with protected rights of political association’ may be sustained if the State demonstrates [1] a sufficiently important interest and [2] employs means closely drawn to avoid unnecessary abridgment of associational freedoms. Cousins v. Wigoda, [419 U. S. 477, 488 (1975)]; NAACP v. Button, [371 U. S. 415, 438 (1963)]; Shelton v. Tucker [364 U. S. 479, 488 (1960)].” 424 U. S., at 25.

“The markedly greater burden on basic freedoms [referring to ‘the freedom of speech and association’] caused by [expenditure limits] thus cannot be sustained simply by invoking the interest in maximizing the effectiveness of the less intrusive contribution limitations. Rather, the constitutionality of [the expenditure limits] turns on whether the governmental interests advanced in its support satisfy the exacting scrutiny applicable to limitations on core First Amendment rights of political expression.” Id., at 44–45.†