FEDERAL ELECTION COMMISSION, APPELLANT
WISCONSIN RIGHT TO LIFE, INC.
SENATOR JOHN McCAIN, et al., APPELLANTS
WISCONSIN RIGHT TO LIFE, INC.
on appeals from the united states district court forthe district of columbia
Justice Scalia, with whom Justice Kennedy and Justice Thomas join, concurring in part and concurring in the judgment.
A Moroccan cartoonist once defended his criticism of the Moroccan monarch (lse majesté being a serious crime in Morocco) as follows: “ ‘I’m not a revolutionary, I’m just defending freedom of speech… . I never said we had to change the king—no, no, no, no! But I said that some things the king is doing, I do not like. Is that a crime?’ ”1 Well, in the United States (making due allowance for the fact that we have elected representatives instead of a king) it is a crime, at least if the speaker is a union or a corporation (including not-for-profit public-interest corporations) and if the representative is identified by name within a certain period before a primary or congressional election in which he is running. That is the import of §203 of the Bipartisan Campaign Reform Act of 2002 (BCRA), the constitutionality of which we upheld three Terms ago in McConnell v. Federal Election Comm’n, 540 U. S. 93 (2003) . As an element essential to that determination of constitutionality, our opinion left open the possibility that a corporation or union could establish that, in the particular circumstances of its case, the ban was unconstitutional because it was (to pursue the analogy) only the king’s policies and not his tenure in office that was criticized. Today’s cases present the question of what sort of showing is necessary for that purpose. For the reasons I set forth below, it is my view that no test for such a showing can both (1) comport with the requirement of clarity that unchilled freedom of political speech demands, and (2) be compatible with the facial validity of §203 (as pronounced in McConnell). I would therefore reconsider the decision that sets us the unsavory task of separating issue-speech from election-speech with no clear criterion.
Today’s cases originated in the efforts of Wisconsin Right to Life, Inc. (WRTL), a Wisconsin nonprofit, nonstock ideological advocacy corporation, to lobby Wisconsin voters concerning the filibustering of the President’s judicial nominees. The problem for WRTL was that, under §203 of BCRA, it would have been unlawful to air its television and radio ads within 30 days before the September 14, 2004, primary or within 60 days before the November 2, 2004, general election because the ads named Senator Feingold, who was then seeking reelection. Section 203(a) of BCRA amended §316(b)(2) of the Federal Election Campaign Act Amendments of 1974, which prohibited corporations and unions from “mak[ing] a contribution or expenditure in connection with any election to any political office, or in connection with any primary election … for any political office.” 2 U. S. C. §441b(a). Prior to BCRA, that section covered only expenditures for communications that expressly advocated the election or defeat of a candidate (in campaign-finance speak, so-called “express advocacy”). McConnell, supra, at 204. As amended, however, that section was broadened to cover “electioneering communication[s],” §441b(b)(2) (2000 ed., Supp. IV), which include “any broadcast, cable, or satellite communication” that “refers to a clearly identified candidate for Federal office” and that is aired within 60 days before a general election, or 30 days before a primary election, in the jurisdiction in which the candidate is running. §434(f)(3) (2000 ed., Supp. IV).2 Under the new law, a corporation or union wishing to air advertisements covered by the definition of “electioneering communication” is prohibited by §203 from doing so unless it first creates a separate segregated fund run by a “political action committee,” commonly known as a “PAC.” §441b(b)(2)(C) (2000 ed., Supp. IV). Three Terms ago, in McConnell, supra, this Court upheld most of BCRA’s provisions against constitutional challenge, including §203. The Court found that the “vast majority” of ads aired during the 30-day and 60-day periods before elections were “the functional equivalent of express advocacy,” id., at 206, but suggested that “pure issue ads,” id., at 207, or “genuine issue ads,” id., at 206, would be protected.
The question in these cases is whether §203 can be applied to WRTL’s ads consistently with the First Amendment . Last Term, this Court unanimously held, in Wisconsin Right to Life, Inc. v. Federal Election Comm’n, 546 U. S. 410, 411–412 (2006) (per curiam) (WRTL I), that as-applied challenges to §203 are available. The District Court in these cases subsequently held that §203 is unconstitutional as applied to the three ads at issue. The Court today affirms the judgment of the District Court. While I agree with that result, I disagree with the principal opinion’s reasons.
A proper explanation of my views in these cases requires some discussion of the case law leading up to McConnell. I begin with the seminal case of Buckley v. Valeo, 424 U. S. 1 (1976) (per curiam), wherein this Court considered the constitutionality of various political contribution and expenditure limitations contained in the Federal Election Campaign Act of 1971 (FECA), 86 Stat. 3, as amended, 88 Stat. 1263. Buckley set forth a now-familiar framework for evaluating the constitutionality of campaign-finance regulations. The Court began with the recognition that contributing money to, and spending money on behalf of, political candidates implicates core First Amendment protections, and that restrictions on such contributions and expenditures “operate in an area of the most fundamental First Amendment activities.” 424 U. S., at 14. The Court also recognized, however, that the Government has a compelling interest in “prevention of corruption and the appearance of corruption.” Id., at 25. The “corruption” to which the Court repeatedly referred was of the “quid pro quo” variety, whereby an individual or entity makes a contribution or expenditure in exchange for some action by an official. Id., at 26, 27, 45, 47.
The Court then held that FECA’s contribution limitations passed constitutional muster because they represented a “marginal restriction upon the contributor’s ability to engage in free communication,” id., at 20–21, and were thus subject to a lower level of scrutiny, id., at 25. The Court invalidated, however, FECA’s limitation on independent expenditures (i.e., expenditures made to express one’s own positions and not in coordination with a campaign). Id., at 39–51. In the Court’s view, expenditure limitations restrict speech that is “ ‘at the core of our electoral process and of the First Amendment freedoms,’ ” id., at 39, and require the highest scrutiny, id., at 44–45.
The independent-expenditure restriction at issue in Buckley limited the amount of money that could be spent “ ‘relative to a clearly identified candidate.’ ” Id., at 41 (quoting 18 U. S. C. §608(e)(1) (1970 ed., Supp. IV) (repealed 1976)). Before striking down the expenditure limitation, the Court narrowly construed §608(e)(1), in light of vagueness concerns, to cover only express advocacy—that is, advertising that “in express terms advocate[s] the election or defeat of a clearly identified candidate for federal office” by use of such words of advocacy “as ‘vote for,’ ‘elect,’ ‘support,’ ‘cast your ballot for,’ ‘Smith for Congress,’ ‘vote against,’ ‘defeat,’ ‘reject.’ ” 424 U. S., at 44, and n. 52. This narrowing construction excluded so-called “issue advocacy”—for example, an ad that refers to a clearly identified candidate’s position on an issue, but does not expressly advocate his election or defeat. Even as narrowly construed to cover only express advocacy, however, §608(e)(1) was held to be unconstitutional because the narrowed prohibition was too narrow to be effective and (quite apart from that shortcoming) independent expenditures did not pose a serious enough threat of corruption. Id., at 45–46. Notably, the Court also found the Government’s interest in “equalizing the relative ability of individuals and groups to influence the outcome of elections” insufficient to support limitations on independent expenditures. Id., at 48.
Buckley might well have been the last word on limitations on independent expenditures. Some argued, however, that independent expenditures by corporations should be treated differently. That argument should have been foreclosed by Buckley for several reasons: (1) the particular provision at issue in Buckley, §608(e)(1) of FECA, was directed to expenditures not just by “individuals,” but by “persons,” with “ ‘persons’ ” specifically defined to include “ ‘corporation[s],’ ” id., at 23, 39, n. 45; (2) the plaintiffs in Buckley included corporations, id., at 8; and (3) Buckley, id., at 50–51, cited a case that involved limitations on corporations in support of its striking down the restriction at issue, Miami Herald Publishing Co. v. Tornillo, 418 U. S. 241 (1974) . Moreover, pre-Buckley cases had accorded corporations full First Amendment protection. See, e.g., NAACP v. Button, 371 U. S. 415, 428–429, 431 (1963) (holding that the corporation’s activities were “modes of expression and association protected by the First and Fourteenth Amendment s”); Grosjean v. American Press Co., 297 U. S. 233, 244 (1936) (holding that corporations are guaranteed the “freedom of speech and of the press … safeguarded by the due process of law clause of the Fourteenth Amendment ”). See also Pacific Gas & Elec. Co. v. Public Util. Comm’n of Cal., 475 U. S. 1, 8 (1986) (plurality opinion) (“The identity of the speaker is not decisive in determining whether speech is protected”; “[c]orporations and other associations, like individuals, contribute to the ‘discussion, debate, and the dissemination of information and ideas’ that the First Amendment seeks to foster”).
Indeed, one would have thought the coup de gr<fs fs="11.5">â<fs fs="11">ce to the argument that corporations can be treated differently for these purposes was dealt by First Nat. Bank of Boston v. Bellotti, 435 U. S. 765 (1978) , decided just two years after Buckley. In that case, the Court struck down a Massachusetts statute that prohibited corporations from spending money in connection with a referendum unless the referendum materially affected the corporation’s property, business, or assets. As the Court explained: The principle that such advocacy is “at the heart of the First Amendment ’s protection” and is “indispensable to decisionmaking in a democracy” is “no less true because the speech comes from a corporation rather than an individual.” 435 U. S., at 776–777. And the Court rejected the arguments that corporate participation “would exert an undue influence on the outcome of a referendum vote”; that corporations would “drown out other points of view” and “destroy the confidence of the people in the democratic process,” id., at 789; and that the prohibition was needed to protect corporate shareholders “by preventing the use of corporate resources in furtherance of views with which some shareholders may disagree,” id., at 792–793.3
The Court strayed far from these principles, however, in one post-Buckley case: Austin v. Michigan Chamber of Commerce, 494 U. S. 652 (1990) . This was the only pre-McConnell case in which this Court had ever permitted the Government to restrict political speech based on the corporate identity of the speaker. Austin upheld state restrictions on corporate independent expenditures in support of, or in opposition to, any candidate in elections for state office. 494 U. S., at 654–655. The statute had been modeled after the federal statute that BCRA §203 amended, which had been construed to reach only express advocacy, id., at 655, n. 1. And the ad at issue in Austin used the magical and forbidden words of express advocacy: “Elect Richard Bandstra.” Id., at 714 (App. to opinion of Kennedy, J., dissenting). How did the Court manage to reach this result without overruling Bellotti? It purported to recognize a different class of corruption: “the corrosive and distorting effects of immense aggregations of wealth that are accumulated with the help of the corporate form and that have little or no correlation to the public’s support for the corporation’s political ideas.” Austin, supra, at 660.
Among the many problems with this “new” theory of corruption was that it actually constituted “the same ‘corrosive and distorting effects of immense aggregations of wealth,’ found insufficient to sustain a similar prohibition just a decade earlier,” in Bellotti. McConnell, 540 U. S., at 325 (opinion of Kennedy, J.) (quoting Austin, supra, at 660; citation omitted). Indeed, Buckley itself had cautioned that “[t]he First Amendment ’s protection against governmental abridgment of free expression cannot properly be made to depend on a person’s financial ability to engage in public discussion.” 424 U. S., at 49. However, two Members of Austin’s 6-to-3 majority appear to have thought it significant that Austin involved express advocacy whereas Bellotti involved issue advocacy. 494 U. S., at 675–676 (Brennan, J., concurring); id., at 678 (Stevens, J., concurring).4
Austin was a significant departure from ancient First Amendment principles. In my view, it was wrongly decided. The flawed rationale upon which it is based is examined at length elsewhere, including in a dissenting opinion in Austin that a Member of the 5-to-4 McConnell majority had joined, see Austin, 494 U. S., at 695–713 (opinion of Kennedy, J., joined by O’Connor, J.). See also id., at 679–695 (Scalia, J., dissenting); McConnell, 540 U. S., at 257–259 (opinion of Scalia, J.); id., at 325–330 (opinion of Kennedy, J.); id., at 273–275 (opinion of Thomas, J.). But at least Austin was limited to express advocacy, and nonexpress advocacy was presumed to remain protected under Buckley and Bellotti, even when engaged in by corporations.
Three Terms ago the Court extended Austin’s flawed rationale to cover an even broader class of speech. In McConnell, the Court rejected a facial overbreadth challenge to BCRA §203’s restrictions on corporate and union advertising, which were not limited to express advocacy but covered vast amounts of nonexpress advocacy (embraced within the term “electioneering communications”). 540 U. S., at 203–209. The Court held that, at least in light of the availability of the PAC option, the compelling governmental interest that supported restrictions on corporate expenditures for express advocacy also justified the extension of those restrictions to “electioneering communications,” the “vast majority” of which were intended to influence elections. Id., at 206. Of course, the compelling interest to which the Court referred was “ ‘the corrosive and distorting effects of immense aggregations of [corporate] wealth,’ ” id., at 205 (quoting Austin, supra, at 660). “The justifications for the regulation of express advocacy,” the Court explained, “apply equally” to ads run during the BCRA blackout period “to the extent … [those ads] are the functional equivalent of express advocacy.” 540 U. S., at 206 (emphasis added). The Court found that the “vast majority” of ads aired during the 30- and 60-day periods before elections fit that description. Finally, the Court concluded that, “[e]ven … assum[ing] that BCRA will inhibit some constitutionally protected corporate and union speech” (i.e., “pure issue ads,” id., at 207, or “genuine issue ads,” id., at 206, and n. 88), its application to such ads was insubstantial, and thus the statute was not overbroad, id., at 207. But McConnell did not foreclose as-applied challenges to §203, WRTL I, 546 U. S., at 411–412, which brings me back to the present cases.
The question is whether WRTL meets the standard for prevailing in an as-applied challenge to BCRA §203. Answering that question obviously requires the Court to articulate the standard. The most obvious one, and the one suggested by the Federal Election Commission (FEC) and intervenors, is the standard set forth in McConnell itself: whether the advertisement is the “functional equivalent of express advocacy.” McConnell, supra, at 206. See also Brief for Appellant FEC 18 (arguing that WRTL’s “advertisements are the functional equivalent of the sort of express advocacy that this Court has long recognized may be constitutionally regulated”); Reply Brief for Appellant Sen. John McCain et al. in No. 06–970, p. 14 (“[C]ourts should apply the standard articulated in McConnell; Congress may constitutionally restrict corporate funding of ads that are the ‘functional equivalent of express advocacy’ for or against a candidate”). Intervenors flesh out the standard somewhat further: “[C]ourts should ask whether the ad’s audience would reasonably understand the ad, in the context of the campaign, to promote or attack the candidate.” Id., at 15. The District Court instead articulated a five-factor test that looks to whether the ad under review “(1) describes a legislative issue that is either currently the subject of legislative scrutiny or likely to be the subject of such scrutiny in the near future; (2) refers to the prior voting record or current position of the named candidate on the issue described; (3) exhorts the listener to do anything other than contact the candidate about the described issue; (4) promotes, attacks, supports, or opposes the named candidate; and (5) refers to the upcoming election, candidacy, and/or political party of the candidate.” 466 F. Supp. 2d 195, 207 (DC 2006). The backup definition of “electioneering communications” contained in BCRA itself, see n. 2, supra, offers another possibility. It covers any communication that “promotes or supports a candidate for that office … (regardless of whether the communication expressly advocates a vote for or against a candidate) and which also is suggestive of no plausible meaning other than an exhortation to vote for or against a specific candidate.” And the principal opinion in this case offers a variation of its own (one bearing a strong likeness to BCRA’s backup definition): whether “the ad is susceptible of no reasonable interpretation other than as an appeal to vote for or against a specific candidate.” Ante, at 16.
There is a fundamental and inescapable problem with all of these various tests. Each of them (and every other test that is tied to the public perception, or a court’s perception, of the import, the intent, or the effect of the ad) is impermissibly vague and thus ineffective to vindicate the fundamental First Amendment rights of the large segment of society to which §203 applies. Consider the application of these tests to WRTL’s ads: There is not the slightest doubt that these ads had an issue-advocacy component. They explicitly urged lobbying on the pending legislative issue of appellate-judge filibusters. The question before us is whether something about them caused them to be the “functional equivalent” of express advocacy, and thus constitutionally subject to BCRA’s criminal penalty. Does any of the tests suggested above answer this question with the degree of clarity necessary to avoid the chilling of fundamental political discourse? I think not.
The “functional equivalent” test does nothing more than restate the question (and make clear that the electoral advocacy need not be express). The test which asks how the ad’s audience “would reasonably understand the ad” provides ample room for debate and uncertainty. The District Court’s five-factor test does not (and could not possibly) specify how much weight is to be given to each factor—and includes the inherently vague factor of whether the ad “promotes, attacks, supports, or opposes the named candidate.” (Does attacking the king’s position attack the king?) The tests which look to whether the ad is “susceptible of no plausible meaning” or “susceptible of no reasonable interpretation” other than an exhortation to vote for or against a specific candidate seem tighter. They ultimately depend, however, upon a judicial judgment (or is it—worse still—a jury judgment?) concerning “reasonable” or “plausible” import that is far from certain, that rests upon consideration of innumerable surrounding circumstances which the speaker may not even be aware of, and that lends itself to distortion by reason of the decisionmaker’s subjective evaluation of the importance or unimportance of the challenged speech. In this critical area of political discourse, the speaker cannot be compelled to risk felony prosecution with no more assurance of impunity than his prediction that what he says will be found susceptible of some “reasonable interpretation other than as an appeal to vote for or against a specific candidate.” Under these circumstances, “[m]any persons, rather than undertake the considerable burden (and sometimes risk) of vindicating their rights through case-by-case litigation, will choose simply to abstain from protected speech—harming not only themselves but society as a whole, which is deprived of an uninhibited marketplace of ideas.” Virginia v. Hicks, 539 U. S. 113, 119 (2003) (citation omitted).
It will not do to say that this burden must be accepted—that WRTL’s antifilibustering, constitutionally protected speech can be constrained—in the necessary pursuit of electoral “corruption.” We have rejected the “can’t-make-an-omelet-without-breaking-eggs” approach to the First Amendment , even for the infinitely less important (and less protected) speech category of virtual child pornography. In Ashcroft v. Free Speech Coalition, 535 U. S. 234 (2002) , the Government argued:
“the possibility of producing images by using computer imaging makes it very difficult for it to prosecute those who produce pornography by using real children. Experts … may have difficulty in saying whether the pictures were made by using real children or by using computer imaging. The necessary solution … is to prohibit both kinds of images.” Id., at 254–255.
The Court rejected the principle that protected speech may be banned because it is difficult to distinguish from unprotected speech. Ibid. “[T]hat protected speech may be banned as a means to ban unprotected speech,” it said, “turns the First Amendment upside down.” Id., at 255. The same principle must be applied here. Indeed, it must be applied a fortiori, since laws targeting political speech are the principal object of the First-Amendment guarantee. The fact that the line between electoral advocacy and issue advocacy dissolves in practice is an indictment of the statute, not a justification of it.
Buckley itself compels the conclusion that these tests fall short of the clarity that the First Amendment demands. Recall that Buckley narrowed the ambiguous phrase “any expenditure … relative to a clearly identified candidate” to mean any expenditure “advocating the election or defeat of a candidate.” 424 U. S., at 42 (internal quotation marks omitted). But that construction alone did not eliminate the vagueness problem because “the distinction between discussion of issues and candidates and advocacy of election or defeat of candidates may often dissolve in practical application.” Ibid. Any effort to distinguish between the two based on intent of the speaker or effect of the speech on the listener would “ ‘pu[t] the speaker … wholly at the mercy of the varied understanding of his hearers,’ ” would “ ‘offe[r] no security for free discussion,’ ” and would “ ‘compe[l] the speaker to hedge and trim.’ ” Id., at 43 (quoting Thomas v. Collins, 323 U. S. 516, 535 (1945) ). In order to avoid these “constitutional deficiencies,” the Court was compelled to narrow the statutory language even further to cover only advertising that used the magic words of express advocacy. 424 U. S., at 43–44.
If a permissible test short of the magic-words test existed, Buckley would surely have adopted it. Especially since a consequence of the express-advocacy interpretationwas the invalidation ofthe entire limitation on independent expenditures, in part because the statute (as thus narrowed) could not be an effective limitation on expenditures for electoral advocacy. (It would be “naiv[e],” Buckley said, to pretend that persons and groups would have difficulty “devising expenditures that skirted the restriction on express advocacy of election or defeat but nevertheless benefited the candidate’s campaign.” Id., at 45.) Why did Buckley employ such a “highly strained” reading of the statute, McConnell, 540 U. S., at 280 (opinion of Thomas, J.), when broader readings, more faithful to the text, were available that might not have resulted in such underinclusiveness? In particular, after going to the trouble of narrowing the statute to cover “advocacy of [the] election or defeat of a candidat[e],” why not do what the principal opinion in these cases does, which is essentially to preface that phrase with the phrase “susceptible of no reasonable interpretation other than as”? Ante, at 16. There is only one plausible explanation: The Court eschewed narrowing constructions that would have been more faithful to the text and more effective at capturing campaign speech because those tests were all too vague. We cannot now adopt a standard held to be facially vague on the theory that it is somehow clear enough for constitutional as-applied challenges. If Buckley foreclosed such vagueness in a statutory test, it also must foreclose such vagueness in an as-applied test.
Though the principal opinion purports to recognize the “imperative for clarity” in this area of First Amendment law, its attempt to distinguish its test from the test found to be vague in Buckley falls far short. It claims to be “not so sure” that Buckley rejected its test because Buckley’s holding did not concern “what the constitutional standard was in the abstract, divorced from specific statutory language.” Ante, at 21, n. 7. Forget about abstractions: Thespecific statutory language at issue in Buckley was interpreted to mean “ ‘advocating the election or defeat of a candidate,’ ” and that is materially identical to the operative language in the principal opinion’s test. The principal opinion’s protestation that Buckley’s vagueness holding “d[id] not dictate a constitutional test,” ante, at 21, n. 7, is utterly compromised by the fact that the principal opinion itself relies on the very same vagueness holding to reject an intent-and-effect test in this case. See ante, at 13–14 (citing Buckley, supra, at 43–44). It is the same vagueness holding, and the principal opinion cannot invoke it on page 13 of its opinion and disclaim it on page 22. Finally, the principal opinion quotes McConnell for the proposition that “[t]he Buckley Court’s ‘express advocacy restriction was an endpoint of statutory interpretation, not a first principle of constitutional law.’ ” Ante, at 21, n. 7 (quoting McConnell, 540 U. S., at 190). I am not sure why this cryptic statement is at all relevant, since we are discussing here the principle of constitutional law that underlay Buckley’s express-advocacy restriction. In any case, the statement is assuredly not a repudiation of Buckley’s vagueness holding, since overbreadth and not vagueness was the issue in McConnell.<footcall num="5">
What, then, is to be done? We could adopt WRTL’s proposed test, under which §203 may not be applied to any ad (1) that “focuses on a current legislative branch matter, takes a position on the matter, and urges the public to ask a legislator to take a particular position or action with respect to the matter,” and (2) that “does not mention any election, candidacy, political party, or challenger, or the official’s character, qualifications, or fitness for office,” (3) whether or not it “say[s] that the public official is wrong or right on the issue,” so long as it does not expressly say he is “wrong for [the] office.” Brief for Appellee 56–57 (footnote omitted).6 Or we could of course adopt the Buckley test of express advocacy. The problem is that, although these tests are clear, they are incompatible with McConnell’sholding that §203 is facially constitutional, which was premised on the finding that a vast majority of ads proscribed by §203 are “sham issue ads,” 540 U. S., at 185, that fall outside the First Amendment ’s protection. Indeed, any clear rule that would protect all genuine issue ads would cover such a substantial number of ads prohibited by §203 that §203 would be rendered substantially overbroad. The Government claims that even the amorphous test adopted by the District Court “call[s] into question a substantial percentage of the statute’s applications,” Tr. of Oral Arg. 4,7 and that any test providing relief to WRTL is incompatible with McConnell’s facial holding because WRTL’s ads are in the “heartland” of what Congress meant to prohibit, Brief for Appellant FEC 18, 28, 36, n. 9. If that is so, then McConnell cannot be sustained.
Like the Buckley Court and the parties to these cases, I recognize the practical reality that corporations can evade the express-advocacy standard. I share the instinct that “[w]hat separates issue advocacy and political advocacy is a line in the sand drawn on a windy day.” See McConnell, supra, at 126, n. 16 (internal quotation marks omitted); Brief for Appellant FEC 30; Brief for Appellant Sen. John McCain et al. in No. 06–970, p. 35. But the way to indulge that instinct consistently with the First Amendment is either to eliminate restrictions on independent expenditures altogether or to confine them to one side of the traditional line—the express-advocacy line, set in concrete on a calm day by Buckley, several decades ago. Section 203’s line is bright, but it bans vast amounts of political advocacy indistinguishable from hitherto protected speech.
The foregoing analysis shows that McConnell was mistaken in its belief that as-applied challenges could eliminate the unconstitutional applications of §203. They can do so only if a test is adopted which contradicts the holding of McConnell—that §203 is facially valid because the vast majority of pre-election issue ads can constitutionally be proscribed. In light of the weakness in Austin’s rationale, and in light of the longstanding acceptance of the clarity of Buckley’s express-advocacy line, it was adventurous for McConnell to extend Austin beyond corporate speech constituting express advocacy. Today’s cases make it apparent that the adventure is a flop, and that McConnell’s holding concerning §203 was wrong.8
Which brings me to the question of stare decisis. “Stare decisis is not an inexorable command” or “ ‘a mechanical formula of adherence to the latest decision.’ ” Payne v. Tennessee, 501 U. S. 808, 828 (1991) (quoting Helvering v. Hallock, 309 U. S. 106, 119 (1940) ). It is instead “ ‘a principle of policy,’ ” Payne, supra, at 828, and this Court has a “considered practice” not to apply that principle of policy “as rigidly in constitutional as in nonconstitutional cases.” Glidden Co. v. Zdanok, 370 U. S. 530, 543 (1962) . This Court has not hesitated to overrule decisions offensive to the First Amendment (a “fixed star in our constitutional constellation,” if there is one, West Virginia Bd. of Ed. v. Barnette, 319 U. S. 624, 642 (1943) )—and to do so promptly where fundamental error was apparent. Just three years after our erroneous decision in Minersville School Dist. v. Gobitis, 310 U. S. 586 (1940) , the Court corrected the error in Barnette. Overruling a constitutional case decided just a few years earlier is far from unprecedented.9
Of particular relevance to the stare decisis question in these cases is the impracticability of the regime created by McConnell. Stare decisis considerations carry little weight when an erroneous “governing decisio[n]” has created an “unworkable” legal regime. Payne, supra, at 827. As described above, the McConnell regime is unworkable because of the inability of any acceptable as-applied test to validate the facial constitutionality of §203—that is, its inability to sustain proscription of the vast majority of issue ads. We could render the regime workable only by effectively overruling McConnell without saying so—adopting a clear as-applied rule protective of speech in the “heartland” of what Congress prohibited. The promise of an administrable as-applied rule that is both effective in the vindication of First Amendment rights and consistent with McConnell’s holding is illusory.
It is not as though McConnell produced a settled body of law. Indeed, it is far more accurate to say that McConnell unsettled a body of law. Not until 1947, with the enactment of the Taft-Hartley amendments to the Federal Corrupt Practices Act, 1925, did Congress even purport to regulate campaign-related expenditures of corporations and unions. See United States v. CIO, 335 U. S. 106, 107, 113–115 (1948) . In the three decades following, this Court expressly declined to pronounce upon the constitutionality of such restrictions on independent expenditures. See Pipefitters v. United States, 407 U. S. 385, 400 (1972) ; United States v. Automobile Workers, 352 U. S. 567, 591–592 (1957) ; CIO, supra, at 110, 124. When the Court finally did turn to that question, it struck them down. See Buckley, 424 U. S. 1 . Our subsequent pre-McConnell decisions, with the lone exception of Austin, disapproved limits on independent expenditures. The modest medicine of restoring First Amendment protection to nonexpress advocacy—speech that was protected until three Terms ago—does not unsettle an established body of law.
Neither do any of the other considerations relevant to stare decisis suggest adherence to McConnell. These cases do not involve property or contract rights, where reliance interests are involved. Payne, supra, at 828. And McConnell’s §203 holding has assuredly not become “embedded” in our “national culture.” Dickerson v. United States, 530 U. S. 428, 443–444 (2000) (declining to overrule Miranda v. Arizona, 384 U. S. 436 (1966) , in part because it had become embedded in our national culture). If §203 has had any cultural impact, it has been to undermine the traditional and important role of grassroots advocacy in American politics by burdening the “budget-strapped nonprofit entities upon which many of our citizens rely for political commentary and advocacy.” McConnell, 540 U. S., at 340 (opinion of Kennedy, J.).
Perhaps overruling this one part of McConnell with respect to one part of BCRA would not “ai[d] the legislative effort to combat real or apparent corruption.” Id., at 194.But the First Amendment was not designed to facilitate legislation, even wise legislation. Indeed, the assessment of former House Minority Leader Richard Gephardt, a proponent of campaign-finance reform, may well be correct. He said that “ ‘[w]hat we have is two important values in direct conflict: freedom of speech and our desire for healthy campaigns in a healthy democracy,’ ” and “ ‘[y]ou can’t have both.’ ” Gibbs, The Wake-Up Call, Time, Feb. 3, 1997, pp. 22, 25. (He was referring, presumably, to incumbents’ notions of healthy campaigns.) If he was wrong, however, and the two values can coexist, it is pretty clear which side of the equation this institution is primarily responsible for. It is perhaps our most important constitutional task to assure freedom of political speech. And when a statute creates a regime as unworkable and unconstitutional as today’s effort at as-applied review proves §203 to be, it is our responsibility to decline enforcement.
* * *
There is wondrous irony to be found in both the genesis and the consequences of BCRA. In the fact that the institutions it was designed to muzzle—unions and nearly all manner of corporations—for all the “corrosive and distorting effects” of their “immense aggregations of wealth,” were utterly impotent to prevent the passage of this legislation that forbids them to criticize candidates (including incumbents). In the fact that the effect of BCRA has been to concentrate more political power in the hands of the country’s wealthiest individuals and their so-called 527 organizations, unregulated by §203. (In the 2004 election cycle, a mere 24 individuals contributed an astounding total of $142 million to 527s. S. Weissman & R. Hassan, BCRA and the 527 Groups, in The Election After Reform 79, 92—96 (M. Malbin ed. 2006).) And in the fact that while these wealthy individuals dominate political discourse, it is this small, grass-roots organization of Wisconsin Right to Life that is muzzled.
I would overrule that part of the Court’s decision in McConnell upholding §203(a) of BCRA. Accordingly, I join Parts I and II of today’s principal opinion and otherwise concur only in the judgment.
1 Whitlock, Satirist Continues to Prove Himself a Royal Pain, Washington Post, Apr. 26, 2005, pp. C1, C8.
2 BCRA also includes a backup definition of “electioneering communication” that will take effect in the event the primary definition is “held to be constitutionally insufficient … to support the regulation provided herein.” 2 U. S. C. §434(f)(3)(A)(ii) (2000 ed., Supp. IV). This defines “electioneering communication” as “any broadcast, cable, or satellite communication which promotes or supports a candidate for [a federal] office, or attacks or opposes a candidate for that office (regardless of whether the communication expressly advocates a vote for or against a candidate) and which also is suggestive of no plausible meaning other than an exhortation to vote for or against a specific candidate.” Ibid.
3 In Federal Election Comm’n v. Massachusetts Citizens for Life, Inc., 479 U. S. 238, 248 (1986) (MCFL), we addressed the pre-BCRA version of 2 U. S. C. §441b, which was interpreted to ban corporate treasury expenditures for express advocacy in connection with federal elections. We held that, “[r]egardless of whether th[e] concern [for unfair advantage to organizations that amass great wealth] is adequate to support application of §441b to commercial enterprises, a question not before us, that justification” did not support application of the statute to the nonprofit organization that brought the challenge in MCFL. 479 U. S., at 263 (emphasis added).
4 The dissent asserts that Austin was faithful to Bellotti’s principles, to prove which it quotes a footnote in Bellotti leaving open the possibility that independent expenditures by corporations might someday be demonstrated to beget quid-pro-quo corruption. Post, at 12, n. 6 (opinion of Souter, J.) (quoting Bellotti, 435 U. S., at 788, n. 26). That someday has never come. No one seriously believes that independent expenditures could possibly give rise to quid-pro-quo corruption without being subject to regulation as coordinated expenditures.
5 Justice Alito’s concurrence at least hints that the principal opinion’s test may impermissibly chill speech, and offers to reconsider McConnell’s holding “[i]f it turns out that the implementation of the as-applied standard set out in the [principal opinion] impermissibly chills political speech.” Post, at 1 (emphasis added). The wait-and-see approach makes no sense and finds no support in our cases. How will we know that would-be speakers have been chilled and have not spoken? If a tree does not fall in the forest, can we hear the sound it would have made had it fallen? Our normal practice is to assess ex ante the risk that a standard will have an impermissible chilling effect on First Amendment protected speech. Justice Alito seemed to recognize that as recently as, well, today. In another opinion released this morning, he finds that a proposed test for censoring student speech “can easily be manipulated in dangerous ways,” wherefore he “would reject it before such abuse occurs.” Morse v. Frederick, ante, at 2 (concurring opinion) (emphasis added). I would accord the core First Amendment speech at issue here at least the same respect he accords speech in the classroom.
6 The principal opinion claims that its test is no more vague than WRTL’s test. See ante, at 21, n. 7. I disagree. WRTL’s test requires yes or no answers to a series of precise and focused questions: Does the ad take a position on a legislative matter? Does it mention the election? Does it expressly say the candidate is wrong for the office? A group of children—indeed, even a group of college students—could answer these questions with great consistency. The principal opinion’s test, by contrast, hinges on assessment of the reasonableness of a determination that something does not constitute advocacy of the election or defeat of a candidate.
7 The same must be said, I think, of the test proposed by the principal opinion. While its coverage is not entirely clear, it would apparently protect even McConnell’s paradigmatic example of the functional equivalent of express advocacy—the so-called “Jane Doe ad,” which “condemned Jane Doe’s record on a particular issue before exhorting viewers to ‘call Jane Doe and tell her what you think,’ ” 540 U. S., at 126–127. Indeed, it at least arguably protects the most “striking” example of a so-called sham issue ad in the McConnell record, the notorious “Yellowtail ad,” which accused Bill Yellowtail of striking his wife and then urged listeners to call him and “[t]ell him to support family values.” Id., at 193–194, n. 78 (internal quotation marks omitted). The claim that §203 on its face does not reach a substantial amount of speech protected under the principal opinion’s test—and that the test is therefore compatible with McConnell—seems to me indefensible. Indeed, the principal opinion’s attempt at distinguishing McConnell is unpersuasive enough, and the change in the law it works is substantial enough, that seven Justices of this Court, having widely divergent views concerning the constitutionality of the restrictions at issue, agree that the opinion effectively overrules McConnell without saying so. See post, at 24–25 (Souter, J., dissenting). This faux judicial restraint is judicial obfuscation.
8 Justice Kennedy’s opinion in McConnell explained why the possibility of corporations’ funding speech out of a PAC does not save the statute from constitutional infirmity. See 540 U. S., at 330–333. McConnell’s rejection of those arguments rested, of course, upon the assumption that for non-PAC genuine issue ads as-applied challenges would be available. See id., at 207; WRTL I, 546 U. S. 410, 411–412 (2006) (per curiam). The discussion today shows that to be mistaken. The dissent asserts, post, at 31, that there is no reason “why substituting the phrase ‘Contact your Senators’ for the phrase ‘Contact Senators Feingold and Kohl’ would have denied WRTL a constitutionally sufficient … alternative.” Surely that is not so. The purpose of the ad was to put political pressure upon Senator Feingold to change his position on the filibuster—not only through the constituents who accepted the invitation to contact him, but also through the very existence of an ad bringing to the public’s attention that he, Senator Feingold, stood athwart the allowance of a vote on judicial nominees. (Unlike the principal opinion, I think that the fair import of the ad in context.)
9 See, e.g., Seminole Tribe of Fla. v. Florida, 517 U. S. 44 (1996) (overruling Pennsylvania v. Union Gas Co., 491 U. S. 1 (1989) ); Adarand Constructors, Inc. v. Peńa, 515 U. S. 200 (1995) (overruling in part Metro Broadcasting, Inc. v. FCC, 497 U. S. 547 (1990) ); United States v. Dixon, 509 U. S. 688 (1993) (overruling Grady v. Corbin, 495 U. S. 508 (1990) ); Payne v. Tennessee, 501 U. S. 808 (1991) (overruling South Carolina v. Gathers, 490 U. S. 805 (1989) , and Booth v. Maryland, 482 U. S. 496 (1987) ); Daniels v. Williams, 474 U. S. 327 (1986) (overruling in part Parratt v. Taylor, 451 U. S. 527 (1981) ); Garcia v. San Antonio Metropolitan Transit Authority, 469 U. S. 528 (1985) (overruling National League of Cities v. Usery, 426 U. S. 833 (1976) ); United States v. Scott, 437 U. S. 82 (1978) (overruling United States v. Jenkins, 420 U. S. 358 (1975) ); National League of Cities, supra, (overruling Maryland v. Wirtz, 392 U. S. 183 (1968) ); Edelman v. Jordan, 415 U. S. 651 (1974) (overruling in part Shapiro v. Thompson, 394 U. S. 618 (1969) ; State Dept. of Health and Rehabilitative Servs. of Fla. v. Zarate, 407 U. S. 918 (1972) ; and Sterrett v. Mothers’ & Children’s Rights Organization, 409 U. S. 809 (1972) ); Miller v. California, 413 U. S. 15 (1973) (overruling Book Named “John Cleland’s Memoirs of a Woman of Pleasure” v. Attorney General of Mass., 383 U. S. 413 (1966) ); Perez v. Campbell, 402 U. S. 637 (1971) (overruling Kesler v. Department of Public Safety of Utah, 369 U. S. 153 (1962) ).