Freedom of Speech and Speech About Political Candidates: The Unintended Consequences of Three Proposals
Most effective speech to the public requires money. If spending more than $1000 on expression is outlawed, then you may not place more than a tiny ad in any major newspaper, buy virtually any television time, put up a billboard, or mail more than a few thousand newsletters. That’s an extremely serious restriction on your ability to express your views.
This, of course, has obvious implications for campaign finance laws, because this means that some such laws may be unconstitutional. But it also has implications for free speech law, since the precedents set in campaign finance cases will affect the rules governing speech restrictions generally. And because campaign finance cases arouse such passions, they risk creating rules that can have unexpected—and possibly dangerous—consequences in many areas far outside campaign finance.
This essay aims to briefly highlight the potential unexpected consequences of three kinds of arguments about campaign finance restrictions, arguments that as it happens correspond to three opinions in the recent Nixon v. Shrink Missouri Government PAC case. And if the analysis here is right, then it may well be that the Nixon majority and even the much-lambasted Buckley v. Valeo actually make a good deal of constitutional sense.
Let me begin with Justice Breyer’s opinion, which I found eloquent enough that I would like to quote it at some length; but let me at the same time quote another similarly eloquent opinion alongside it.
[T]he principal dissent oversimplifies [and] takes a difficult constitutional problem and turns it into a lopsided dispute between political expression and government censorship. Under the cover of this fiction and its accompanying formula, the dissent would make the Court absolute arbiter of a difficult question best left, in the main, to the political branches. . . .
[T]his is a case where constitutionally protected interests lie on both sides of the legal equation.
For that reason there is no place for a strong presumption against constitutionality, of the sort often thought to accompany the words “strict scrutiny.” Nor can we expect that mechanical application of the tests associated with “strict scrutiny”—the tests of “compelling interests” and “least restrictive means”—will properly resolve the difficult constitutional problem that campaign finance statutes pose. Cf. Kovacs v. Cooper, 336 U.S. 77, 96 (1949) (Frankfurter, J., concurring) (objecting, in the First Amendment context, to “oversimplified formulas”) . . . .
I recognize that [earlier cases] used language that could be interpreted to the contrary. . . . But those words cannot be taken literally. . . . In such circumstances—where a law significantly implicates competing constitutionally protected interests in complex ways—the Court has closely scrutinized the statute’s impact on those interests, but refrained from employing a simple test that effectively presumes unconstitutionality. . . .
Rather, it has balanced interests. . . . And in practice that has meant asking whether the statute burdens any one such interest in a manner out of proportion to the statute’s salutary effects upon the others (perhaps, but not necessarily, because of the existence of a clearly superior, less restrictive alternative). Where a legislature has significantly greater institutional expertise . . . [the Court] in practice defers to empirical legislative judgments . . . .
Our judgment is solicited on a conflict of interests of the utmost concern to the well-being of the country. This conflict of interests cannot be resolved by a dogmatic preference for one or the other, nor by a sonorous formula which is in fact only a euphemistic disguise for an unresolved conflict. If adjudication is to be a rational process, we cannot escape a candid examination of the conflicting claims with full recognition that both are supported by weighty title-deeds. . . .
The plain fact [is] that the interest in speech, profoundly important as it is, is no more conclusive in judicial review than other attributes of democracy or than a determination of the people’s representatives that a measure is necessary to assure the safety of government itself. . . .
Just as there are those who regard as invulnerable every measure for which the claim of national survival is invoked, there are those who find in the Constitution a wholly unfettered right of expression. Such literalness treats the words of the Constitution as though they were found on a piece of outworn parchment instead of being words that have called into being a nation with a past to be preserved for the future. The soil in which the Bill of Rights grew was not a soil of arid pedantry. . . .
Nor is the argument . . . adequately [decided] by citing isolated cases. Adjustment of a clash of interests which are at once subtle and fundamental is not likely to reveal entire consistency in a series of instances presenting the clash. It is not too difficult to find what one seeks in the language of decisions reporting the effort to reconcile free speech with the interests with which it conflicts. The case for the defendants requires that their conviction be tested against the entire body of our relevant decisions. . . .
How are competing interests to be assessed? Since they are not subject to quantitative ascertainment, the issue necessarily resolves itself into asking, who is to make the adjustment?—who is to balance the relevant factors and ascertain which interest is in the circumstances to prevail? . . . Primary responsibility for adjusting the interests which compete in the situation before us of necessity belongs to the Congress. . . .
The opinion on the left is Justice Breyer’s; the one on the right was written by Justice Frankfurter—the same Justice whose Kovacs v. Cooper concurrence Breyer approvingly cites—and it was written in Dennis v. United States. Restrictions on Communist advocacy, Justice Frankfurter argued, couldn’t be judged by wooden formulae or subjected to a rigid presumption of unconstitutionality, because they implicated constitutional values (free speech versus democracy) on both sides of the inquiry. Instead, these restrictions required context-sensitive balancing, and ultimately deference to the legislative branch. Justice Breyer similarly argues that restrictions on campaign-related speech can’t be judged by wooden formulae or subjected to a rigid presumption of unconstitutionality, because they implicate constitutional values (free speech versus democracy) on both sides of the inquiry. Instead, the restrictions require context-sensitive balancing, and ultimately deference to the legislative branch.
I think Justice Frankfurter got it wrong in Dennis, and I am happy that the Court has since implicitly retreated from its holding in that case. But there’s much to what he says, just as there’s much to what Justice Breyer says. Their opinions outline a powerful, even if ultimately insufficiently speech-protective, alternative to conventional free speech libertarianism.
But I doubt that many of the liberals who support the Breyer view of campaign speech restrictions would be wild about a revival of First Amendment Frankfurterianism. Justice Frankfurter certainly didn’t limit his balancing to Communist advocacy: He was willing to balance away free speech rights in a wide variety of cases. If Justice Breyer is modeling himself on Justice Frankfurter, that’s not a good sign for those who support strong speech protection.
What about the one limiting principle that Breyer’s opinion seems to suggest—that his deferential balancing applies only when “constitutionally protected interests lie on both sides of the legal equation,” an approach I call the Constitutional Tension Method? This may sound like a helpful constraint, until one recognizes how often judges can identify some constitutionally protected “interest” supporting an abridgement of a constitutionally protected right.
Justice Frankfurter’s own jurisprudence provides many such examples. In Dennis, he justified restricting Communist speech by pointing to the constitutional interests in “democracy” and “[t]he constitutional power to act upon [the] basic principle” of “[t]he right of a government to maintain its existence,” which is “the most pervasive aspect of sovereignty.” In Bridges v. California, he would have upheld a court’s punishing a newspaper for commenting on pending litigation, on the grounds that in such a case “the claims on behalf of freedom of speech and of the press encounter claims on behalf of liberties no less precious”—the right to a fair trial. Likewise, in Terminiello v. City of Chicago, he endorsed Justice Jackson’s opinion, which would have allowed punishment of speech that could lead to mob violence, even when the speaker isn’t intentionally trying to incite such violence; without such a power, the opinion argued, constitutional liberties, including the freedom of speech itself, would suffer.
Other judges and commentators have used the Constitutional Tension Method to argue for yet other speech restrictions. In rough chronological order, here are a few examples:
· The Sedition Act: Justice Iredell defended the Act on the grounds that seditious speech jeopardized the constitutionally established regime of republican government and even the freedom of the press itself.
· Antiwar speech: The World War I-era Supreme Court justified restrictions on antiwar speech on the grounds that such speech undermined the constitutionally secured war power.
· Advocacy of illegal conduct: Gitlow v. New York and other contemporaneous decisions argued that advocacy of illegal conduct was punishable because such speech was contrary to the constitutional framework of democratic decisionmaking.
· Bitter criticism of religion: Throughout the 1940s and early 1950s, Justice Jackson argued that the government may restrict religious advocacy that harshly criticizes others’ religions, because such speech is in tension with the constitutional value of religious freedom.
· Racist and sexist advocacy: Modern advocates of restrictions on bigoted advocacy argue that these restrictions are permissible because they further the constitutionally established value of equality.
And we can easily imagine still other possible examples of what might happen if the Constitutional Tension Method is adopted. Speech aimed at pressuring newspapers into dropping a columnist whose work the speaker finds offensive could be punished on the theory that such pressure itself jeopardizes the constitutional value of free speech. Speech praising the killing of police officers or justifying the actions of looters could be punished on the theory that it undermines the constitutionally protected values of life and property.
The Court has generally rejected these arguments, and for good reason. First, these examples of “constitutional tension” do not actually involve any real conflict between specific constitutional provisions. Our government is built along certain constitutionally specified lines, has certain constitutionally specified powers, and labors under certain constitutionally specified disabilities. But even when speech somehow undermines one of the “values” embodied in this structure (whether it be democracy, the war power, the rule of law, the power of the judiciary, religious tolerance, or racial or sexual equality), the speech does not in fact violate any of these constitutional provisions.
Antiwar advocacy may make it harder to wage war, but it doesn’t actually contradict Congress’s war power. People’s spending a lot of money to further their agenda may be in some way undemocratic, but it isn’t actually prohibited by any of the rules set forth in the Constitution. There is thus no need to choose between violating one constitutional provision or the other—courts may insist that the government obey the Free Speech Clause without actually contradicting any other constitutional provision.
Second, accepting the constitutional tension argument in one area seriously risks undermining free speech protection in all the other areas to which I’ve pointed. Perhaps Justice Breyer (unlike Justice Frankfurter) would be willing to limit his application of the constitutional tension method to election-related speech, though I’m not sure about that. But in any event, if the method is adopted by the Court, it will become a precedent for many other cases.
After all, if free speech may be trumped by “constitutionally protected interests” even when the speech—advocacy of the election or defeat of a candidate—lies at the heart of First Amendment protection, why shouldn’t the same happen in a variety of other areas? If strict scrutiny is inapplicable here, why shouldn’t it become equally inapplicable elsewhere? Even those who support restrictions on campaign-related speech might be properly hesitant to endorse a First Amendment exception as potentially broad as this one.
Justice Stevens’s Shrink Missouri opinion was written to “make one simple point. Money is property; it is not speech.” Therefore, Justice Stevens argued, the use of money in election campaigns to pay “hired laborers,” “mercenaries,” or “gladiators” should be protected only under the rules applicable to other “property rights,” and should not be given “the same protection as the right to say what one pleases.”
Now Justice Stevens’s “simple point” is literally true—money is not speech—but it doesn’t show much by itself. After all, expenditure limits don’t just bar the use of money; they single out the use of money to speak. A law restricting people from flying places to give speeches would be a speech restriction, not because “flying is speech” but because giving a speech is speech and burdening such speech (you may not fly in order to do it) is a speech restriction. Likewise for restrictions on spending money for speech.
Justice Stevens would reject this standard First Amendment analysis, and hold that a restriction on using money to speak is not properly seen as a speech restriction. This position, though, would have some striking consequences. To begin with, it’s hard to see how his logic would be limited to spending money on campaign-related speech. When a newspaper hires reporters or pays editorialists, for instance, it’s also using “[m]oney . . . to pay hired laborers to perform [speech-related] tasks.” When a publisher buys the rights to a book, it is also engaging in a sort of “‘speech by proxy’” and in a sense “hir[ing a] gladiator” rather than “say[ing] what [it by itself] pleases.” When a movie director uses actors to make a movie, he is “hir[ing] mercenaries to work in [his speech product].” Thus, under Justice Stevens’s approach, the government would acquire broad power to control newspapers, book publishers, directors, and anyone else who hires others to participate in creating a speech product, rather than doing all the speaking himself.
Nor is it clear how Justice Stevens’s framework could be limited just to speech, rather than applying to a wide variety of other constitutional rights. Say that the fact that “money is property, it is not speech” means the government has a fairly free hand in restricting your using money to speak. The fact that “money is property, it is not abortion, contraception, education, assistance of counsel, or free exercise of religion” would then equally enable the government to restrict your using money to get abortions, buy contraceptives, send your kids to private schools, hire a lawyer, or build a church.
After all, the doctor paid to perform an abortion, the lawyer paid to represent you, and the contractor paid to build a church are as much “hired laborers,” “mercenaries,” and “gladiators” as is a newspaper that is paid money to run a political ad. Under Justice Stevens’s analysis, only conduct that relies on “inspir[ing] volunteers” would be protected using the traditional constitutional rules; any conduct that involves spending money would not.
These may seem like radical hypotheticals—but Justice Stevens’s is a radical proposal. If it’s taken seriously, and not limited in an arbitrary way to the situation in which it’s suggested, then it would indeed extend to these examples. And if it’s inappropriate for these examples, why is it sound for free speech?
Justice Stevens might reply that he’s not rejecting all constitutional protection for the spending of money: Rather, he’s suggesting only that spending money would need to be judged under the constitutional rules protecting property rights generally, and not the special constitutional rules applicable to free speech (or to reproductive rights, parental rights, the right to counsel, the free exercise of religion, and the like). Justice Stevens’s opinion, for instance, suggests that one’s property rights might possibly protect some sort of right to spend one’s money on independent campaign expenditures.
But this is only a suggestion, with no elaboration or explanation. Justice Stevens never states exactly what protection his proposal would provide. I suspect that it won’t provide that much, given that the constitutional right to use one’s property as one sees fit hasn’t gotten particularly strong protection from the Court—and certainly not from Justice Stevens.
Perhaps under Justice Stevens’s approach parents’ property rights (for example) would still be construed broadly enough to let parents use their money to educate their children. Or perhaps not. Given Justice Stevens’s past abortion opinions, I take it that he would view the property rights of pregnant women and contributors to Planned Parenthood as broad enough to let them use their money to fund abortions; but no-one knows under what principle. It is for this mystery that Justice Stevens suggests we trade existing constitutional protections.
Justice Thomas, joined by Justice Scalia, dissented in Shrink Missouri, and called on the Court to strike down limits on campaign contributions. Such limits, he argued, should be “subject[ed] to strict scrutiny” (a point on which Buckley agreed), and should fail this scrutiny because the limits have not been proven to be narrowly tailored to the interest in preventing corruption. And the speakers’ ability to communicate through alternative means, such as independent expenditures, shouldn’t matter. “The First Amendment protects [individuals’] right not only to advocate their cause but also to select what they believe to be the most effective means for so doing.” “[T]he Constitution leaves it entirely up to citizens and candidates to determine who shall speak, the means they will use, and the amount of speech sufficient to inform and persuade.”
But according to a long string of Supreme Court cases, the Constitution does not generally allow individuals “to select what they believe to be the most effective means for [speaking],” or “leave it entirely up to [speakers] to determine . . . the means they will use.” The government may stop me from speaking using soundtrucks, residential picketing, billboards, the burning of a draft card, or a variety of other means, and it may do so even if I want to speak on core political topics. Content-neutral speech and conduct restrictions that leave ample alternative channels for speaking are judged under a far lower standard than strict scrutiny, even when they deny speakers the ability to use what the speakers think are the most effective means of communication.
Restrictions on contributions to candidates that are justified by a concern with quid pro quo corruption are similar to content-neutral speech restrictions. They apply the same principle that underlies the broader restraints on any valuable gifts to officeholders, restraints that are likewise justified by the concern that a gift may be a covert bribe.
On their face, contribution restrictions bar all contributions of money to political campaigns, regardless of the content of the speech that the contribution is fostering, or even of whether the contribution is being used for speech. And they are justified without reference to the communicative impact of speech: The concern that a contribution might be a tacit bribe doesn’t turn on the communicative impact of the speech for which the contribution is likely to be spent. In this respect, restrictions on campaign contributions justified by the danger of subtle bribery differ from limits on what candidates spend out of their own funds and from limits on total spending by a candidate campaign: These latter limits are not justifiable as attempts to prevent quid pro quo corruption; rather, they are “designed to ensure that the political speech of the wealthy not drown out the speech of others” and are thus indeed “concerned with the communicative impact of the regulated speech.”
Now it’s true that some limits on campaign contributions may in fact be motivated by the desire to limit the communicative impact of well-funded speech, even if the limits are ostensibly aimed at preventing quid pro quo corruption. Worse yet, some supporters of such limits may even be motivated by the desire to suppress those ideas that wealthy contributors are particularly likely to promote.
But while this possibility of content-based legislative motivation is a serious concern, the Court has generally held that it will not inquire into such matters, because it’s just too hard to determine why legislators really implemented a content-neutral restriction. This is the holding of United States v. O’Brien, where the Court refused to strike down the ban on destroying draft cards, despite the plausible claim that some of its supporters voted for it precisely because they wanted to suppress the message of defiance communicated by public draft card burning. So long as the law is facially justifiable without reference to the communicative impact of the prohibited behavior, it is treated as content-neutral.
Likewise, it’s true that limits on campaign contributions have the foreseeable effect of disproportionately restricting speech of a particular content—speech advocating the election of candidates. But the same is true of all content-neutral restrictions. Bans on soundtrucks in an environment where soundtrucks are used largely for political campaigns have a similar effect. Bans on residential picketing have a disproportionate effect on those causes that disproportionately use residential picketing. Content-neutral injunctions disproportionately affect certain speakers. The Court has concluded, however, that these disparate impacts on various kinds of speech do not turn a facially content-neutral restriction—one that’s justified without reference to the communicative impact of the speech—into a content-based one.
So if this analysis is correct, the proper standard of review of campaign contribution restrictions must be different from that used by Justice Thomas’s dissent. The Court may not rely on the argument that the speaker generally gets to choose his means of communication: While this principle generally applies when the law restricts the content of speech (as in Cohen v. California, one of the cases on which Justice Thomas relies) and thus interferes with the speaker’s choice of content, it generally doesn’t apply when a content-neutral law restricts the manner of speech and thus interferes with the speaker’s choice of manner. Nor may the Court demand that the law be the least restrictive alternative, unless it overrules its precedents holding that the least restrictive alternative inquiry is inapplicable to content-neutral restrictions.
True, even the scrutiny applicable to content-neutral restraints imposes two important demands on the law. First, content-neutral restrictions are subject to lower scrutiny only if they leave open ample alternative channels. This is why independent expenditure restrictions would be unconstitutional even if they were treated as content-neutral: Limiting people to spending only $1000 to spread their messages leaves them with virtually no effective alternative means of reaching their intended audience.
One can likewise argue that campaign contribution limits also fail to leave open ample alternatives. To begin with, one can argue that they fail to leave open such alternatives for would-be contributors who want to say to the public “vote for candidate Joe Schmoe.” The alternative of speaking via independent expenditures, the argument would go, is grossly ineffective compared to contributing money to Schmoe’s campaign, where it would be pooled with other funds and would become part of Schmoe’s coherent “vote for Schmoe” strategy. And one can also argue that campaign contribution limits fail to leave open ample alternatives for the candidates themselves: By diminishing the funds available to candidates, they cut back on the amount of speaking the candidate can do.
These are serious problems, and Justice Thomas’s opinion points them out. Nonetheless, under the existing law of content-neutral restrictions, it’s not clear that these problems are serious enough to make the restriction unconstitutional. Especially given the ability of many people to pool their independent expenditures—the expenditures would only have to be independent of the candidate, not of each other—a regime that allowed unlimited expenditures but only modest contributions would still leave people with considerable opportunities to speak.
“That more people may be more easily and cheaply reached by sound trucks, perhaps borrowed without cost from some zealous supporter,” the Court held in Kovacs v. Cooper, “is not enough to call forth constitutional protection for what those charged with the public welfare reasonably think is a nuisance when easy means of publicity are open.” “That the city’s limitations on volume [of music in Central Park] may reduce to some degree the potential audience for respondent’s speech,” the Court held in Ward v. Rock Against Racism, “is of no consequence, for there has been no showing that the remaining avenues of communication are inadequate.”
Likewise, that more people may be more easily and cheaply reached using large campaign contributions, or that the shift to independent expenditures may reduce to some degree the potential audience for a contributor’s speech, may not be enough to call forth constitutional protection for what may be a potential source of a considerable amount of hidden bribery. A campaign contribution limit may be invalidated if it significantly frustrates contributors or (more likely) candidates’ ability to express their views, and perhaps $250 limits (or even $1000 limits, given today’s costs) might indeed have such a drastic effect. But if the effect is modest enough that the candidate and the contributors are seen as having ample alternative means of communicating, then such a modest burden may be constitutional even if it in some measure interferes with speakers’ “select[ion of] what they believe to be the most effective means for [advocating their cause].”
The second important rule applicable to content-neutral restrictions is that they may not be overinclusive: “Government may not regulate expression in such a manner that a substantial portion of the burden on speech does not serve to advance its goals.” Justice Thomas’s dissent likewise rested on this argument: “I cannot fathom,” Justice Thomas, wrote, “how a $251 contribution could pose a substantial risk of securing a political quid pro quo.” If this argument is correct, and I think it might be, then the contribution restriction is overinclusive, since a substantial portion of it—the limit on restrictions from $251 to whatever total does begin to pose a substantial bribery risk—does not serve to advance the goal of preventing quid pro quo corruption.
But even if the argument is correct, it does not support overruling Buckley altogether; at some point, the corruption risk does become great enough that the law stops being overinclusive. Moreover, even $251 contributions may be used as a sort of implicit bribe if they are bundled together, as contributions often are. One hundred $300 contributions from members of some group, raised at the group’s fundraiser, can be a substantial amount in a race for an office that represents fewer than 100,000 people (the sort of office for which the $250 limit was in effect in Shrink Missouri).
How might Justice Thomas’s dissent respond to this? Well, it argues, contribution limits are overinclusive not just because some contributions are so small that they are ex ante incapable of being bribes. Rather, they are overinclusive because looking at matters ex post, we can see that only a small minority of contributions were actually bribes. “[A] blunderbuss approach which prohibits mostly innocent speech cannot be held a means narrowly and precisely directed to the governmental interest in the small minority of contributions that are not innocent.”
This argument raises an important question, one that goes to the heart of the overinclusiveness inquiry. Say that a law restricts certain kinds of speech or expressive conduct on the grounds that each statement has the potential to cause a certain harm, and that it’s impossible to tell up front whether any particular statement will cause such harm. But say also that in retrospect it’s clear that most such statements end up not causing the harm. How should this law fare under a First Amendment “overinclusiveness” inquiry?
Sometimes, the Court has taken the approach that Justice Thomas suggests. In Schneider v. New Jersey, for instance, the government defended a leafleting ban on the grounds that it was needed to prevent littering; though many leaflets wouldn’t be thrown on the ground, each leaflet had the potential to be littered. Nonetheless, the Court struck down the law, and later cases characterized this result as reflecting the principle that handbilling bans are unconstitutional because they “would suppress a great quantity of speech that does not cause the evils that it seeks to eliminate.” (Schneider itself didn’t explicitly discuss this point.)
On the other hand, in most such cases the Court has upheld prophylactic content-neutral restrictions. In Regan v. Time, Inc., for instance, the Court upheld a ban on color reproductions of U.S. currency and on same-size (even if black-and-white) reproductions, on the grounds that the ban was justified by the interest in preventing counterfeiting. Of course, the overwhelming majority of such reproductions weren’t used for counterfeiting, but the Court stressed the possibility that such reproductions “increase a counterfeiter’s access to . . . negatives and plates and enable him to more easily use them for counterfeiting purposes under the guise of a legitimate project . . . . That the limitations may apply to some photographs that are themselves of no use to counterfeiters does not invalidate the legislation.”
Likewise, United States v. O’Brien held that a ban on destroying draft cards was justified by the government’s interests in enabling people to prove their registration status, facilitating communication with draft boards, and reminding the registrant of his duties, even though in practice many of the draft cards would never be used for this purpose, and thus many destructions of draft cards wouldn’t really implicate the government’s interests. Heffron v. ISKCON similarly held that a state fair’s ban on soliciting and leafleting by wandering speakers was justified by the fair’s interest in avoiding congestion, even though in practice much of the restricted speech would not have caused congestion.
These results make considerable sense, because the prophylactic restrictions in those cases did not “burden substantially more speech than is necessary to further” the government interest. The whole burden, including the burden on the speech that might ultimately prove not to implicate the interest, was necessary, since there was no way of telling up front which speech would have implicated the interest and which wouldn’t have.
Under this dominant approach, in which “[t]he requirement of narrow tailoring is satisfied so long as the regulation promotes a substantial government interest that would be achieved less effectively absent the regulation,” a restriction on campaign contributions should probably be seen as narrowly tailored. If the restriction were removed, and only “[b]ribery laws” and “disclosure laws” remained—which is what Justice Thomas suggests—the government interest in preventing de facto bribes would be “achieved less effectively.”
Now perhaps Justice Thomas is right that the government shouldn’t be able to achieve its interests with maximum effectiveness, because “the First Amendment does not permit the State to sacrifice speech for efficiency.” But such a test, which would predominantly rest on precedents borrowed from the law of content-based speech restrictions, would require a substantial change to the existing jurisprudence of content-neutral speech restraints.
Likewise for the reintroduction of a least restrictive alternatives inquiry, or for adapting the principle that speakers have a broad right to choose the content of their speech into the principle that speakers have an equally broad right to choose the manner of their speech. People considering these proposals—especially conservatives, who have traditionally endorsed a great deal of legislative flexibility as to content-neutral speech restrictions—should thus think seriously about such proposals’ implications outside the campaign speech area.
The alternatives to Buckley v. Valeo offered by both the left and the right should give pause even to those who might at first find them appealing. Liberals who support campaign speech restrictions should balk at either Justice Breyer’s or Justice Stevens’s broad First Amendment revisionism. Conservatives who oppose such restrictions should be hesitant to endorse an approach such as Justice Thomas’s that might potentially disable the government from enacting many sensible content-neutral speech restraints. And when all is said and done, Buckley might prove to be not quite the travesty that many on both the left and the right sometimes suggest it might be.
* Professor of Law, UCLA School of Law (firstname.lastname@example.org). This essay builds heavily on the ideas of scholars whose experience with campaign speech restrictions vastly exceeds my own; I can’t easily cite all of their works, or trace exactly where those works influenced this one, but I can at least mention a small subset: Lillian R. BeVier, Money and Politics: A Perspective on the First Amendment and Campaign Finance Reform, 73 Calif. L. Rev. 1045 (1985); Sanford Levinson, Regulating Campaign Activity: The New Road to Contradiction?, 83 Mich. L. Rev. 939 (1985); Daniel Hays Lowenstein, A Patternless Mosaic: Campaign Finance and the First Amendment After Austin, 21 Cap. U. L. Rev. 381 (1992); L.A. Powe, Jr., Mass Speech and the Newer First Amendment, 1982 Sup. Ct. Rev. 243; Bradley A. Smith, Faulty Assumptions and Undemocratic Consequences of Campaign Finance Reform, 105 Yale L.J. (1996). Also, many thanks to Dan Lowenstein for his specific suggestions on a draft of this article.
This essay is a revised version of remarks delivered at the Federalist Society Nineteenth Annual Student Symposium on “Law and the Political Process” at Harvard Law School, March 3-4, 2000.
 120 S. Ct. 897 (2000).
 424 U.S. 1 (1976).
 This is not necessarily to say that the regime established by the Federal Election Campaign Act as it exists post-Buckley is necessarily wise as a policy matter, only that Buckley correctly identified the constitutional constraints on any campaign finance speech and contribution restrictions. Cf. Richard L. Hasen, Shrink Missouri, Campaign Finance, and “The Thing That Wouldn’t Leave”, 17 Const. Comm. ___, Part III.C (forthcoming 2000) (arguing that the current “loophole-ridden system is nonsensical,” and ultimately concluding that the government should have more power to restrict speech and contributions than Buckley allows); Shrink Mo., 120 S. Ct. at 916 (condemning “the existing distortion of speech caused by the half-way house we created in Buckley,” and seemingly leaning towards the view that the government should have less power to restrict speech and contributions than Buckley allows).
 Shrink Mo., 120 S. Ct. at 910-12 (Breyer, J., concurring) (some paragraph breaks added) (parallel citations omitted).
 341 U.S. 494, 517 (1951) (some paragraph breaks added).
 As I’ll explain in Part III, I actually don’t disagree much with Justice Breyer’s bottom line on contribution limits, the issue that was specifically at stake in Shrink Missouri. Justice Breyer, though, made clear that his analysis also applies to restrictions on independent expenditures, and there I think his framework would reach results that are quite unsound. See Shrink Mo., 120 S. Ct. at 911 (stressing the significance of the government interest in “democratiz[ing] the influence that money itself may bring to bear upon the electoral process,” an interest that has long been used as an argument for restrictions on independent expenditures); id. at 912 (defending the notion that government may sometimes restrict the speech of some “in order to enhance the relative voice of others,” a notion that has likewise long been used to defend restrictions on independent expenditures); id. at 913 (arguing that Buckley should be reinterpreted to “mak[e] less absolute the contribution/expenditure line”); id. at 913-14 (suggesting that if Buckley “denies the political branches sufficient leeway to enact comprehensive solutions to the problems posed by campaign finance,” then Buckley—presumably referring to Buckley’s protection of independent expenditures—“would [have to be] reconsider[ed]”).
 See, e.g., Brandenburg v. Ohio, 395 U.S. 444 (1969) (setting forth a First Amendment test for incitement that was significantly more speech-protective than that used in Dennis); Yates v. United States, 354 U.S. 298 (1957) (substantially limiting Dennis as a matter of statutory construction); Laurence H. Tribe, American Constitutional Law 846-48 (2d ed. 1988) (summarizing these developments).
 See, e.g., Beauharnais v. Illinois, 343 U.S. 250 (1952) (Frankfurter, J., for the majority); Terminiello v. City of Chicago, 337 U.S. 1, 8 (1949) (Frankfurter, J., dissenting); Winters v. New York, 333 U.S. 507, 520 (1948) (Frankfurter, J., dissenting); Bridges v. California, 314 U.S. 252, 279 (1941) (Frankfurter, J., dissenting).
 Later in the opinion Breyer reiterates this focus on “constitutionally protected interests . . . on both sides,” Shrink Mo., 120 S. Ct. at 911, by arguing that the purpose of contribution restrictions is “protect[ing] the integrity of the electoral process—the means through which a free society democratically translates political speech into concrete governmental action,” “democratiz[ing] . . . the electoral process,” and “encouraging the public participation and open discussion that the First Amendment itself presupposes.” Id.; see also First Nat’l Bank of Boston v. Bellotti, 435 U.S. 765, 803-04 (1978) (White, J., dissenting) (making a similar argument); Julian N. Eule, Promoting Speaker Diversity: Austin and Metro Broadcasting, 1990 Sup. Ct. Rev. 105, 130 (likewise).
 See Alex Kozinski & Eugene Volokh, A Penumbra Too Far, 106 Harv. L. Rev. 1639, 1653-56 (1993); Eugene Volokh, Freedom of Speech and the Constitutional Tension Method, 1996 U. Chi. Roundtable 223 (discussing and criticizing the Constitutional Tension Method in more detail). Much of the discussion in this section is based on the analysis in those articles.
 Dennis, 341 U.S. at 519.
 “Free speech is not so absolute or irrational a conception as to imply paralysis of the means for effective protection of all the freedoms secured by the Bill of Rights. In the cases before us, the claims on behalf of freedom of speech and of the press encounter claims on behalf of liberties no less precious.” Bridges v. California, 314 U.S. 252, 282 (1941) (Frankfurter, J., dissenting).
 337 U.S. 1 (1949).
 See Terminiello, 337 U.S. at 36-37 (Jackson, J., dissenting); id. at 12 (Frankfurter, J., dissenting) (endorsing the Jackson dissent).
The necessity [of punishing libels in a republic is], I conceive[,] greater [than in a monarchy] because in a republic more is dependent on the good opinion of the people for its support, as they are, directly or indirectly, the origin of all authority, which of course must receive its bias from them. Take away from a republic the confidence of the people, and the whole fabric crumbles into dust.
Case of Fries, 9 F. Cas. 826, 839 (C.C.D. Pa. 1799) (No. 5,126) (Iredell, J., riding circuit); see also id. at 838 (“[T]o censure the licentiousness is to maintain the liberty of the press.”); Respublica v. Dennie, 4 Yeates 267, 270 (Penn. 1805) (arguing that a diatribe criticizing democracy was constitutionally unprotected because it created disaffection and thus undermined the constitutional structure).
 Gilbert v. Minnesota, 254 U.S. 325, 332 (1920) (“the curious spectacle [is] presented of the Constitution of the United States being invoked to justify [speech subversive of constitutional values], and by a strange perversion of its precepts [being] adduced against itself”); Schaefer v. United States, 251 U.S. 466, 477 (1920) (stressing that the Constitution “empowered Congress to declare war and war is waged with armies,” and marveling at the “curious spectacle . . . presented” of “[t]hat great ordinance of government and orderly liberty[ being] invoked to justify the activities of anarchy or of the enemies of the United States, and by a strange perversion of its precepts [being] adduced against itself”); see also United States v. Macintosh, 283 U.S. 605, 622 (1931) (arguing that the Constitution itself creates Congress’ war power, and therefore “[t]o the end that war may not result in defeat, freedom of speech may, by act of Congress, be curtailed or denied so that the morale of the people and the spirit of the army may not be broken by seditious utterances”).
 268 U.S. 652 (1925).
 See, e.g., United States ex rel. Milwaukee Social Democratic Pub. Co. v Burleson, 255 U.S. 407, 414 (1921) (“Freedom of the press may protect criticism and agitation for modification or repeal of laws, but it does not extend to protection of him who counsels and encourages the violation of the law as it exists. The Constitution was adopted to preserve our government, not to serve as a protecting screen for those who while claiming its privileges seek to destroy it.”); Gitlow, 268 U.S. at 666 (arguing that statements which “necessarily imply the use of force and violence” are “in their essential nature . . . inherently unlawful in a constitutional government of law and order” and that “the punishment of those who publish articles which tend to destroy organized society [is] essential to the security of freedom and the stability of the state”).
 See, e.g., Kunz v. New York, 340 U.S. 290, 302 (1951) (Jackson, J., dissenting):
Is official action the only source of interference with religious freedom? Does the Jew, for example, have the benefit of these freedoms when, lawfully going about, he and his children are pointed out as “Christ-killers” to gatherings on public property by a religious sectarian sponsored by a police bodyguard? We should weigh the value of insulting speech against its potentiality for harm. Is the Court, when declaring Kunz has the right he asserts, serving the great end for which the First Amendment stands?
See also Douglas v City of Jeannette, 319 U.S. 157, 177, 180, 181-82 (1943) (Jackson, J., dissenting) (arguing that Jehovah’s Witnesses who proselytized door-to-door, often stridently condemning Catholicism, interfered with the householders’ “religious liberty” and their “exercise of their own faith in peace”); American Communications Ass’n v Douds, 339 U.S. 382, 444 (1950) (Jackson, J., concurring in part and dissenting in part) (“I have pointed out [in Douglas] that men cannot enjoy their right to personal freedom if fanatical masses, whatever their mission, can strangle individual thoughts and invade personal privacy.”).
 See, e.g., Catherine A. MacKinnon, Only Words 71 (1993) (“[T]he upheaval that produced the Reconstruction Amendments . . . move[d] the ground under the expressive freedom, setting new limits and mandating new extensions.”); Charles R. Lawrence III, If He Hollers Let Him Go: Regulating Racist Speech on Campus, in Mari J. Matsuda et al., Words That Wound: Critical Race Theory, Assaultive Speech, and the First Amendment 53, 61 (1993) (“[When] individual racist acts [are viewed] as part of a totality . . . white supremacists’ conduct or speech is forbidden by the equal protection clause.”); Mari J. Matsuda, Public Response to Racist Speech: Considering the Victim’s Story, in Mari J. Matsuda et al., Words That Wound: Critical Race Theory, Assaultive Speech, and the First Amendment 17, 24-25 (1993) (arguing that racist speech may be restricted because it causes harm of “constitutional dimensions”); Richard Delgado, Campus Antiracism Rules: Constitutional Narratives in Collision, 85 Nw. U. L. Rev. 343, 346 (1991) (suggesting that speech might be restricted “to protect core values emanating from the [T]hirteenth and [F]ourteenth [A]mendments”); Mary Ellen Gale, Reimagining the First Amendment: Racist Speech and Equal Liberty, 65 St. John’s L. Rev. 119, 162 (1991) (arguing that a proposed “equality-based theory of the first amendment [which] permits government regulation of some categories of racist speech . . . is strengthened and sharpened by the [F]ourteenth [A]mendment’s protection of equality”); Brian Owsley, Racist Speech and “Reasonable People”: A Proposal for a Tort Remedy, 24 Colum. Hum. Rts. L. Rev. 323, 324 (1993) (“[T]he Fourteenth Amendment must be used to counteract the existing disproportionate emphasis on the First Amendment.”).
 See, e.g., Jill Stewart, Free This Man; Can Black Conservatives Speak Their Minds in America? Ask KABC Talk-Show Host Larry Elder, the Target of a Black Nationalist Group in L.A., New Times (L.A.), July 3, 1997 (describing boycott of sponsors of black conservative talk show host Larry Elder’s radio show, aimed at getting the radio station to take him off the air); James Warren, Andy Rooney Suspended, But Denies Racist Comment, Chi. Trib., Feb. 9, 1990, § 1, at 3 (describing public pressure that caused CBS’s suspension of 60 Minutes commentator Andy Rooney for allegedly making a racist comment); Jerry Berger, Kennedy Decries Reagan Civil Rights Policies, United Press Int’l, Jan. 18, 1988, available in LEXIS, News Library, UPI File (describing public pressure that caused CBS’s firing of Jimmy “The Greek” Snyder on similar grounds); Youth for Justice, Tonight’s Menu (flyer listing various San Francisco business owners and others who contributed to the California Civil Rights Initiative, saying that “[t]hey’ve left a bad taste in our mouths with their dirty donations to CCRI,” and implicitly but pretty clearly calling for a boycott of at least one of the businesses, a restaurant) (on file with author).
 See, e.g., Thomas D. Elias, TV and Radio Stations Should Be Stripped of Their Licenses If They Aren’t More Responsible in Covering Civil Unrest, L.A. Daily J., Jan. 26, 1993, at 6 (analogizing “irresponsible” coverage of the L.A. riots to “shouting ‘fire’ in a crowded theater”); see also Susan Carpenter McMillan, Both Pro-Life and Pro-Choice Bear Responsibility, L.A. Times, Jan. 5, 1995, at B7 (arguing that “[t]he inexcusable tolerance of the Los Angeles riots and constant threats of ‘no justice, no peace’ may have had a link to the beating of Reginald Denny”); David Crump, Camouflaged Incitement, 29 Ga. L. Rev. 1, 76-78 (1994) (suggesting that songs like Cop Killer should be unprotected because they may lead to attacks on police officers); Chuck Philips, North Steamed at Ice T; He Wants Time Warner to Face Sedition Charges over Rap Song, L.A. Times, July 2, 1992, at D1 (describing Lt. Col. Oliver North’s suggestions to the same effect).
 See Volokh, supra note 9, at 235-37; cf. id. at 236 n.80 (describing some unusual, and rightly controversial, contexts where the Court has seemingly tentatively accepted some aspects of the constitutional tension argument).
 Justice Breyer’s recent decisions suggest that he is willing to uphold a fairly wide variety of speech restrictions, such as restrictions on sexually themed speech, picketing, religious speech, and attorney advertising. See United States v. Playboy Entertainment Group, Inc., 120 S. Ct. 1878, 1898 (2000) (Breyer, J., dissenting); Schenck v. Pro-Choice Network, 519 U.S. 357, 395 (1997) (Breyer, J., concurring in part and dissenting in part); Rosenberger v. Rector & Visitors of University of Va., 515 U.S. 819, 863 (1995) (Souter, J., dissenting, joined by Breyer, J., and others); Florida Bar v. Went for It, Inc., 515 U.S. 618 (1995) (O’Connor, J., for the majority, joined by Breyer, J., who provided the fifth vote); see also Denver Area Educ. Telecomm. Consortium, Inc. v. FCC, 518 U.S. 727 (1996) (Breyer, J., for the plurality) (taking a position that was more speech-protective than that of some Justices, but less so than that of others); 44 Liquormart, Inc. v. Rhode Island, 517 U.S. 484, 528 (1996) (O’Connor, J., concurring in the judgment, joined by Breyer, J., and others) (voting to strike down the speech restriction, but using a theory that was considerably less speech-protective than the plurality’s). See generally Eugene Volokh, How the Justices Voted in Free Speech Cases, 1994-2000, 48 UCLA L. Rev. ___ (forthcoming 2001) (concluding that Justice Breyer takes the narrowest view of free speech among all the Justices on the Court today).
 See, e.g., Eu v. San Francisco County Democratic Cent. Comm., 489 U.S. 214, 223 (1989); Brown v. Hartlage, 456 U.S. 45, 53 (1982); Monitor Patriot Co. v. Roy, 401 U.S. 265, 272 (1971); Garrison v. Louisiana, 379 U.S. 64, 74-75 (1964).
 Shrink Mo., 120 S. Ct. at 910 (Stevens, J., concurring). Justice Stevens suggests one caveat to this principle: The principle wouldn’t apply, “of course, [when] the prohibition entirely forecloses a channel of communication, such as the use of paid petition circulators.” Id. at n.* (citing Meyer v. Grant, 486 U.S. 414 (1988), a Justice Stevens opinion that held that a ban on the use of paid petition circulators violated the First Amendment). Why “of course”? If the First Amendment simply doesn’t apply because “[m]oney is property; it is not speech,” then how does money become any less property and any more speech when the “prohibition [on the spending of money] entirely forecloses a channel of communication”?
What’s more, why is “the use of paid petition circulators” treated as an entire “channel of communication”? It seems more logical to say that the “channel of communication” in Meyer was the use of petition circulators generally, and the ban on paid petition circulators—no less than a ban on paid advertisements that cost more than $1000—is merely a ban on “us[ing] one’s own money to hire gladiators,” id. at 910 (Stevens, J., concurring).
 See, e.g., Simon & Schuster, Inc. v. Members of the N.Y. State Crime Victims Bd., 502 U.S. 105, 115-16 (1991); Meyer v. Grant, 486 U.S. 414, 422-23 (1988); New York Times Co. v. Sullivan, 376 U.S. 254, 266 (1964); Smith v. California, 361 U.S. 147, 150 (1959). Cf. Shrink Mo., 120 S. Ct. at 911 (Breyer, J., concurring) (“a decision to contribute money to a campaign is a matter of First Amendment concern—not because money is speech (it is not); but because it enables speech”); this would be a fortiori true of a decision to independently spend money on behalf of a candidate.
 The notion that the Free Speech Clause protects speech on which money is spent—speech published using “hired laborers”—hardly originated with the campaign finance cases. See, e.g., New York Times Co. v. Sullivan, 376 U.S. 254, 265-66 (1964) (holding that the fact that a newspaper “was paid for publishing [a political] advertisement is as immaterial” for free speech purposes “as is the fact that newspapers and books are sold”).
 Id. at 910 (Stevens, J., concurring).
 Id. (Stevens, J., concurring). But see Simon & Schuster, Inc. v. Members of the N.Y. State Crime Victims Bd., 502 U.S. 105, 115-16 (1991) (restrictions on publishers’ ability to pay certain authors require strict scrutiny under the First Amendment).
 Shrink Mo., 120 S. Ct. at 910 (Stevens, J., concurring).
 Note that just as people have argued that it’s unfair that wealthy people have more opportunities to speak about candidates, so others have argued that it’s unfair that wealthy people have more speech opportunities generally, have access to better education for their children, or have access to better criminal lawyers.
 Of course, the religious freedom example may be explained on the grounds that the Free Exercise Clause bars discrimination against religion, and that a ban on the use of money to build churches thus violates this antidiscrimination principle. But the Free Speech Clause likewise bars content-based discrimination against speech, the very sort of discrimination that takes place in expenditure restrictions.
 Shrink Mo., 120 S. Ct. at 910 (Stevens, J., concurring).
 See id.
 Justice Stevens cites his private-property-protective concurrence in the judgment in Moore v. City of East Cleveland, 431 U.S. 494, 513 (1977), but that opinion got only Justice Stevens’s vote. Moreover, Justice Stevens’s own recent Takings Clause votes have generally taken a fairly narrow view of the private property owner’s rights. See, e.g., Phillips v. Washington Legal Foundation, 524 U.S. 156, 179 (1998) (Breyer, J., dissenting, joined by Stevens, J., and others); Dolan v. City of Tigard, 512 U.S. 374, 396 (1994) (Stevens, J., dissenting); Lucas v. South Carolina Coastal Council, 505 U.S. 1003 (1992) (Stevens, J., dissenting); Nollan v. California Coastal Comm’n, 483 U.S. 825, 866 (1987) (Stevens, J., dissenting); Keystone Bituminous Coal Ass’n v. DeBenedictis, 480 U.S. 470 (1987) (Stevens, J., for the majority).
 Shrink Mo., 120 S. Ct. at 916.
 See Buckley v. Valeo, 424 U.S. 1, 25 (1976) (requiring “the closest scrutiny”).
 See Shrink Mo., 120 S. Ct. at 924.
 Id. at 920 n.5 (citing Meyer v. Grant, 486 U.S. 414, 424 (1988)); see also id. at 919 n.3 (citing Cohen v. California, 403 U.S. 15, 25-26 (1971), for a similar point).
 Id. at 922.
 See Kovacs v. Cooper, 336 U.S. 77 (1949); see also Ward v. Rock Against Racism, 491 U.S. 781 (1989) (upholding limit on volume of amplified music); Grayned v. City of Rockford, 408 U.S. 104 (1972) (upholding limit on volume of speech near schools).
 See Frisby v. Schultz, 487 U.S. 474 (1988).
 See Regan v. Time, Inc., 468 U.S. 641, 656 (1984) (suggesting that “size and height limitations on outdoor signs” may be constitutional); Baldwin v. Redwood City, 540 F.2d 1360 (9th Cir. 1976) (so holding); Temple Baptist Church, Inc. v. City of Albuquerque, 98 N.M. 138, 146 (1982) (likewise).
 See United States v. O’Brien, 391 U.S. 367 (1968).
 This would not be true, in my view, of restrictions justified by an interest in equalizing the amount of speech that people can engage in; this interest is directly related to the desire to diminish the communicative impact of some people’s speech and increase the relative communicative impact of other people’s. See infra text accompanying note 50.
 See 2 U.S.C. § 31-2 (limiting gifts to Senators to $250); Committee on Standards of Official Conduct, U.S. House of Representatives, Rules of the U.S House of Representatives on Gifts and Travel, available at http://www.house.gov/ethics/Gifts_and_Travel_Chapter.htm (limiting gifts to Representatives to $50, with a limit of at most $100 from a single source in a calendar year).
 Cf. Vannatta v. Keisling, 931 P.2d 770, 776 (Or. 1997) (“The money may never be used to promote a form of expression by the candidate; instead, it may (for example) be used to pay campaign staff or to meet other needs not tied to a particular message.”).
 See Turner Broadcasting Sys. v. FCC, 512 U.S. 622, 643, 658 (1994); United States v. Eichman, 496 U.S. 310, 317-18 (1990); Texas v. Johnson, 491 U.S. 397, 411-12 (1989).
 Turner, 512 U.S. at 658 (citing Buckley v. Valeo, 424 U.S. 1, 17 (1976)).
 See, e.g., Levinson, supra note *, at 945; Powe, supra note *.
 391 U.S. 367, 382-84 (1968).
 The restrictions do not restrict only this speech; many candidates use contributions to make speeches that are aimed at linking the candidate to a particular issue or event but do not explicit ask for people’s votes (e.g., a speech on Memorial Day praising the nation’s veterans and intentionally omitting an explicit “vote for me”).
 See, e.g., Madsen v. Women’s Health Ctr., Inc., 512 U.S. 753, 763-64 (1994). Justice Thomas, though, might be unimpressed by this particular analogy; he joined Justice Scalia’s dissenting opinion in Madsen, which argued that the risk of “targeted suppression of particular ideas” did justify strict scrutiny even of injunctions that are content-neutral on their face. Id. at 792-93 (Scalia, J., dissenting).
 See id. at 763-64. Buckley expressly rejected the argument that the contribution restrictions in the Federal Election Campaign Act should be judged under the standards applicable to content-neutral restrictions. The Act, the Court reasoned, was—when viewed in its entirety—a content-justified attempt to equalize people’s voices and not just a content-neutral attempt to restrict quid pro quo corruption. See Buckley v. Valeo, 424 U.S. 1, 16-17 (1976). But it seems to me that laws that are, unlike FECA, limited to restricting contributions and are justified only as attempts to prevent corruption must be treated as content-neutral under established First Amendment principles. Upholding such laws on those grounds would be more consistent with existing First Amendment law than subjecting them to strict scrutiny and upholding them despite that.
 403 U.S. 15 (1971).
 Cf. Shrink Mo., 120 S. Ct. at 926-27 (Thomas, J., dissenting).
 See Ward v. Rock Against Racism, 491 U.S. 781, 798 (1989); Clark v. Community for Creative Non-Violence, 468 U.S. 288, 299 (1984).
 O’Brien didn’t include the ample alternative channels inquiry in its test for expressive conduct restrictions, but in that case ample alternative channels were obviously present—nothing stopped the demonstrators from burning photocopies of draft cards. Clark v. CCNV, on the other hand, stressed that the test for expressive conduct restrictions was pretty much the same as the test for content-neutral time, place, and manner restrictions, see Clark v. CCNV, 468 U.S. at 298, and that test has long contained an ample alternative channels inquiry, see, e.g., Heffron v. ISKCON, 452 U.S. 640, 648 (1981); Virginia State Bd. of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U.S. 748, 770 (1976) (dictum). See also Clark, 428 U.S. at 295 (stressing the availability of alternative ways for the CCNV to communicate).
 I think they shouldn’t be treated as content-neutral, because they are facially content-based: They distinguish between spending for speech that advocates the election of a candidate and spending for other speech. On the other hand, some recent cases suggest that even facially content-based laws are properly treated as content-neutral if they are justified without reference to the communicative impact of the speech. See, e.g., Boos v. Barry, 485 U.S. 312 (1988) (plurality); City of Renton v. Playtime Theatres, 475 U.S. 41 (1986). Some could argue that limits on candidate-related spending in fact constitute the rare case where a facially content-based restriction is in fact justified with reference to a concern—the risk that the expenditure will be an implicit bribe—that is only distantly linked to the communicative impact of the speech.
 Cf. City of Ladue v. Gilleo, 512 U.S. 43, 57 (1994) (striking down a speech restriction because it significantly inhibited a speaker from communicating to his intended audience); Buckley v. Valeo, 424 U.S. 1, 18 n.17 (1976) (stressing that spending limits, unlike bans on soundtrucks, dramatically limit the speaker’s ability to communicate).
 See Shrink Mo., 120 S. Ct. at 920 (Thomas, J., dissenting).
 336 U.S. 77, 88-89 (1949); see also Members of the City Council v. Taxpayers for Vincent, 466 U.S. 789, 812 (1984) (making the same point as to the practice of posting political flyers on government-owned utility poles).
 491 U.S. 781, 802 (1989).
 Compare Shrink Mo., 120 S. Ct. at 922 (Thomas, J., dissenting) (arguing that the restrictions in Shrink Mo. did have this effect) with id. at 908-09 (majority opinion) (arguing the contrary).
 Id. at 920 n.5 (Thomas, J., dissenting).
 Ward, 491 U.S. at 799.
 Shrink Mo., 120 S. Ct. at 924 (Thomas, J., dissenting) (internal quotation marks and brackets omitted).
 See id. at 902.
 Id. at 926; see also Colorado Republican Campaign Comm. v. FEC, 116 S. Ct. 2309, 2329 (1996) (Thomas, J., dissenting) (internal quotation marks and citations omitted).
 308 U.S. 147 (1939).
 Ward v. Rock Against Racism, 491 U.S. 781, 799-800 n.7 (1989); see also Turner Broadcasting Sys. v. FCC, 512 U.S. 622, 682 (1994) (O’Connor, J., dissenting).
 468 U.S. 641 (1984).
 Id. at 657.
 391 U.S. 367 (1968).
 See id. at 378-79.
 452 U.S. 640 (1981).
 Ward, 491 U.S. at 799 (emphasis added); see also Turner Broadcasting Sys. v. FCC, 520 U.S. 180, 186 (1997); Turner Broadcasting Sys. v. FCC, 512 U.S. 622, 682 (1994).
 Ward, 491 U.S. at 799 (internal quotation marks and ellipsis omitted).
 See Shrink Mo., 120 S. Ct. at 926 (Thomas, J., dissenting).
 Id. at 927 (quoting Riley v. National Fed’n of the Blind, Inc., 487 U.S. 781, 795 (1988) (a case involving content-based speech compulsions)).