An Indiana statute, challenged in this suit as an infringement of free speech, requires that political advertising that "expressly advocat[es] the election or defeat of a clearly identified candidate" contain "a disclaimer that appears and is presented in a clear and conspicuous manner to give the reader or observer adequate notice of the identity of persons who paid for ... the communication," Ind.Code §§ 3-9-32.5(b)(1), (d), and makes violation a misdemeanor. § 3-14-1-3. "Disclaimer" is a misnomer; the correct word would be "disclosure" — but as we'll see, that word has been appropriated to describe a reporting requirement.

The district court dismissed the suit on jurisdictional grounds that we concluded were unsound, 317 F.3d 719, 721-23 (7th Cir.2003), but we decided that we should not attempt to decide the merits of the plaintiffs' constitutional challenge until we obtained an authoritative interpretation of the statute from the Indiana Supreme Court. The state had argued that despite using the word "person" to denote who was subject to it the statute was limited to candidates, campaign committees, and the committee's agents. We said that if the statute was as narrow as the state claimed it was — a claim no court of Indiana had passed on — it was a straightforward antifraud statute unlikely to present serious constitutional problems. For on the state's interpretation, the statute merely forbids the candidate and his organization to create the impression that independent voices support him or oppose his opponent, when in fact the voices are those of the candidate himself, playing ventriloquist. Cf. McIntyre v. Ohio Elections Comm'n, 514 U.S. 334, 351, 354, 115 S. Ct. 1511, 131 L. Ed. 2d 426 (1995). But if instead, as the plaintiffs argued, the statute reached all persons, then it was a blanket prohibition of anonymous campaign-related speech (unless the speech is costless, for it is only the identity of someone who pays or contributes to paying, either directly or by soliciting payment, for political advertising that is required to be disclosed), and thus might discourage political speech by exposing persons who want to express themselves for or against a particular candidate to the risk of retaliation, a risk from which anonymity would shield them.

Although the McIntyre decision held that government may not forbid the distribution of anonymous campaign literature, id. at 357, 115 S. Ct. 1511; see also Talley v. California, 362 U.S. 60, 64-65, 80 S. Ct. 536, 4 L. Ed. 2d 559 (1960), several subsequent decisions upheld statutes similar to the Indiana statute interpreted to reach all persons. Federal Election Comm'n v. Public Citizen, 268 F.3d 1283, 1287-91 (11th Cir.2001) (per curiam); Gable v. Patton, 142 F.3d 940, 944-45 (6th Cir.1998); Kentucky Right to Life, Inc. v. Terry, 108 F.3d 637, 646-48 (6th Cir.1997). Those cases point out that the statute struck down in McIntyre applied to issue referenda as well as to candidate elections and that only issue referenda were before the Court, a difference on which McIntyre had relied to distinguish Buckley v. Valeo, 424 U.S. 1, 80, 96 S. Ct. 612, 46 L. Ed. 2d 659 (1976) (per curiam), which had upheld a provision of the federal campaign finance law that was similar to these state statutes. "The Federal Election Campaign Act of 1971, at issue in Buckley, regulates only candidate elections, not referenda or other issue-based ballot measures; and we construed `independent expenditures' [in Buckley] to mean only those expenditures that `expressly advocate the election or defeat of a clearly identified candidate.'" McIntyre v. Ohio Elections Comm'n, supra, 514 U.S. at 356, 115 S. Ct. 1511. The opinions that distinguish McIntyre also point out that Ohio had defended its statute only on the basis that knowing the author of a document helps one to evaluate its truthfulness, whereas a weightier ground is that "disclosure protects the integrity of the electoral process by ensuring that the words of an independent group are not mistakenly understood as having come from the mouth of a candidate." Federal Election Comm'n v. Public Citizen, supra, 268 F.3d at 1288; see also Buckley v. Valeo, supra, 424 U.S. at 66-67, 96 S. Ct. 612; Gable v. Patton, supra, 142 F.3d at 944-45; Kentucky Right to Life, Inc. v. Terry, supra, 108 F.3d at 646-48; Seymour v. Elections Enforcement Comm'n, 255 Conn. 78, 762 A.2d 880, 886-87 (2000). It also deters corruption by identifying large contributors who may be seeking a quid pro quo and — a related point — it provides information helpful to the enforcement of the provisions of election campaign law, both also being purposes that had been emphasized in Buckley, 424 U.S. at 66-68, 96 S. Ct. 612. See generally Malcolm A. Heinicke, Note, "A Political Reformer's Guide to McIntyre and Source Disclosure Laws for Political Advertising," 8 Stan. L. & Pub. Policy Rev. 133 (1997).

Although the Indiana statute is inapplicable to issue referenda (the only type of political campaign that McIntyre had considered), we realized when we first heard the appeal in this case that if the Indiana statute reached political advertising wholly independent of the candidate or his campaign organization, a serious constitutional question would be created. Interest groups contest issue referenda just as candidates and parties contest elections of officials, and so the public interest in knowing the source of an anonymous contribution to the debate is as great in the one case as in the other, though it may be small in both if the contributor is an obscure individual. The Court in McIntyre said that "insofar as the interest in informing the electorate means nothing more than the provision of additional information that may either buttress or undermine the argument in a document, we think the identity of the speaker is no different from other components of the document's content that the author is free to include or exclude," and added that in the case of "a private citizen who is not known to the recipient, the name and address of the author add little, if anything, to the reader's ability to evaluate the document's message." 514 U.S. at 348-49, 115 S. Ct. 1511. Yet this too is an observation that seems apt to campaigns to elect officials, as well as to issue referenda, though perhaps less so than in the latter case.

Our doubts about the constitutionality of the Indiana statute if interpreted more broadly than the state thought it should be interpreted impelled us to certify to the Indiana Supreme Court, pursuant to 7th Cir. R. 52 and Ind.Code § 33-2-4-1, the following question:

Is the term "persons" in Ind.Code §§ 3-9-3-2.5(b)(1), (d) limited to candidates, authorized political committees or subcommittees of candidates, and the agents of such committees or subcommittees, or does it have a broader scope, and, if so, how much broader?

317 F.3d at 725. The Indiana Supreme Court accepted the certification and ruled that the statute was not so limited; that "person" means "person" — any person. It thus rejected the state's narrowing construction. 792 N.E.2d 22 (Ind.2003).

The litigation in our court then resumed. With the major constitutional challenge to the Bipartisan Campaign Reform Act (popularly known as the McCain-Feingold Act) pending in the Supreme Court, we decided to hold off on further consideration of the appeal until the case was decided, because the Act contains a provision rather similar to the Indiana statute challenged in this case. When the Supreme Court rendered its decision, McConnell v. Federal Election Comm'n, 540 U.S. 93, 124 S. Ct. 619, 157 L. Ed. 2d 491 (2003), we directed the parties to submit memoranda discussing the bearing of that decision on the appeal; and they have now done so.

The constitutional issue now ripe for resolution is difficult because it entails a balancing of imponderables. On the one hand, forbidding anonymous political advertising reduces the amount of political advertising because some would-be advertisers are unwilling to reveal their identity. On the other hand, the quality of the political advertising that continues to be produced and disseminated under such a regime is enhanced because the advertising contains additional information useful to the consumer. The avidity with which candidates for public office seek endorsements is evidence (as if any were needed) that the identity of a candidate's supporters — and opponents — is information that the voting public values highly. In areas of inquiry where logic or exact observation is unavailing, a speaker's credibility often depends crucially on who he is. As Aristotle said, "persuasion is achieved by the speaker's personal character when the speech is so spoken as to make us think him credible. We believe good men more fully and more readily than others: this is true generally whatever the question is, and absolutely true where exact certainty is impossible and opinions are divided." Aristotle, Rhetoric, in 2 The Complete Works of Aristotle 2152, 2155 (Jonathan Barnes ed.1984). "Where exact certainty is impossible and opinions are divided" is a pretty good description of politics.

Can we get help in answering the thorny question presented by this appeal from the case law, and in particular from the Supreme Court's recent and very lengthy opinions in the McConnell case? The provision of the Bipartisan Campaign Reform Act that is analogous to the Indiana statute regulates "electioneering communications," which are advertisements broadcast within 60 days of a general election or 30 days of a primary that refer to a candidate for federal office. 2 U.S.C. § 434(f)(3)(A)(i). Individuals who spend more than $10,000 producing such communications, or contribute at least $1,000 to an organization that produces them, must report (in the case of the contributions it is the recipient who must report) their identities to the Federal Election Commission. 2 U.S.C. §§ 434(f)(1)-(2). Without attempting to narrow the class of covered "individuals," the Supreme Court upheld this provision on the ground that it served "important state interests ... [in] providing the electorate with information, deterring actual corruption and avoiding any appearance thereof, and gathering the data necessary to enforce more substantive electioneering restrictions." 540 U.S. at ___, 124 S.Ct. at 690. Like the Indiana statute, the provision of the Bipartisan Campaign Reform Act that the Court upheld requires identifying any person who contributes to the making of the ad, even if the person is not a candidate or a member of the candidate's campaign staff.

True, what is required is disclosure to an agency rather than disclosure in the political ad itself, though, as is apparent from the Court's reference to "providing the electorate with information," the identity of the contributor is available to the public rather than secreted by the FEC. 2 U.S.C. §§ 434(a)(11)-(12), (d)(2). That may not seem a big difference from the standpoint of protecting the advertiser from retaliation, but the Court had earlier indicated that having to identify itself to the entire audience for the ad has as a practical matter a greater inhibiting effect than just a reporting requirement does because it broadcasts the advertiser's name to the entire electoral community. Buckley v. American Constitutional Law Foundation, Inc., 525 U.S. 182, 197-200, 119 S. Ct. 636, 142 L. Ed. 2d 599 (1999). The reaction may not be "retaliation" in any strong sense, but there is a weak sense as well; there may be a degree of social ostracism, some dirty looks, a few snide comments, and such, and we and other courts have long recognized that mild forms of retaliation can be effective in deterring the exercise of free speech. Coady v. Steil, 187 F.3d 727, 734-35 (7th Cir.1999); Pieczynski v. Duffy, 875 F.2d 1331, 1335-36 (7th Cir.1989); Bart v. Telford, 677 F.2d 622 (7th Cir.1982); Keenan v. Tejeda, 290 F.3d 252, 259-60 (5th Cir. 2002); Suppan v. Dadonna, 203 F.3d 228, 234-35 (3d Cir.2000); Allen v. Scribner, 812 F.2d 426, 434 n. 17 (9th Cir.1987).

The Court in McIntyre thought "the intrusion" on freedom of political advocacy brought about by a reporting requirement was "a far cry from compelled self-identification on all election-related writings," 514 U.S. at 355, 115 S. Ct. 1511, which is what we have here — and what the Bipartisan Campaign Reform Act does not have; it does not even require identifying the specific ads financed by the reporting contributor. McConnell v. Federal Election Comm'n, supra, 540 U.S. at ___-___, 124 S.Ct. at 693-94; see 2 U.S.C. §§ 434(f)(1)-(2). But of course the very thing that makes reporting less inhibiting than notice in the ad itself — fewer people are likely to see the report than the notice — makes reporting a less effective method of conveying information that by hypothesis the voting public values. It's as if cigarette companies, instead of having to disclose the hazards of smoking in their ads, had only to file a disclosure statement with the Food and Drug Administration.

The only reference to McIntyre by the majority in McConnell appears in a footnote that distinguishes "genuine issue ads" from "regulation of campaign speech" and assumes that restrictions constitutionally applicable to the latter, such as the restrictions both in the Bipartisan Campaign Reform Act and in the Indiana statute, might not be applicable to the former; McIntyre is cited noncommittally as having invalidated a "statute banning the distribution of anonymous campaign literature." 540 U.S. at ___ n. 88, 124 S.Ct. at 696 n. 88. Remember that McIntyre had only been about issue referenda, where there are no candidates and so, it might be thought, "campaign literature" is more likely to consist of "genuine issue ads." Thus the Court may have so far narrowed McIntyre (one of the dissenting opinions said the Court had overruled it, 540 U.S. at ___ - ___, 124 S.Ct. at 735-36) that it no longer overlaps the Indiana statute.

An alternative interpretation, however, is that because the Bipartisan Campaign Reform Act and therefore the McConnell decision are about campaign financing, the decision is inapplicable to people who pay for political ads themselves, since they are not engaged in fund-raising. On that reading, McIntyre, which was such a case, is unaffected by McConnell, and so the Indiana statute, which is also about requiring self-financiers to identity themselves, is condemned by McIntyre's holding. But campaign financing and fund-raising are not synonyms, as the argument assumes; and the Bipartisan Campaign Reform Act is not just about fund-raising — the relevant provision that the Court upheld applies equally to self-financed and other-financed ads. The disclosure statement must be filed by "every person who makes a disbursement for the direct costs of producing and airing electioneering communications in an aggregate amount in excess of $10,000 during any calendar year," whether he produces the ad himself or gives the money to someone else to produce it, 2 U.S.C. § 434(f)(1); and sections 434(a)(6)(B) and (E) expressly impose a requirement of reporting campaign disbursements by a candidate from his personal funds. And the first governmental interest that the Court recited in upholding the provision — that it would provide the electorate with information — is applicable to self-financed ads. An ad might seem disinterested, but if the voting public knew who had paid for it — maybe it was an interest group that the candidate was known to have done favors for — the existence of an interest might be revealed. To draw the constitutional line between self-and other-financed campaign ads would be to deliver a gratuitous benefit to wealthy candidates and wealthy supporters of candidates.

But what must give us considerable pause, in light of the distinction the Supreme Court has drawn between "disclosure" (reporting one's identity to a public agency) and "disclaimer" (placing that identity in the ad itself), is the fact that the Indiana statute requires the latter and not merely the former. Buckley v. American Constitutional Law Foundation, Inc., supra, 525 U.S. at 197-200, 119 S. Ct. 636, invalidated a state law that required people who circulated petitions for issue referenda (actually initiatives, but the only and irrelevant difference is that a referendum asks the people to vote on a law proposed by the legislature, while in an initiative the proposal has not been before the legislature) to wear identification badges. But the requirement was inapplicable to elections of candidates. Federal Election Comm'n v. Public Citizen, supra, 268 F.3d at 1287-91, interpreting a provision of the previous federal campaign finance law, upheld a requirement of a disclaimer in a candidate election. But all that had to be disclaimed (for once, the word was apt) was that the advertiser was independent of the candidate — yet the court assumed that a separate requirement, that the identity of the advertiser be disclosed in the ad, was also valid. Id. at 1290. There is a similar assumption in McConnell. See 540 U.S. at ___, 124 S.Ct. at 710. A statute quite like the Indiana statute was invalidated in Citizens for Responsible Government State Political Action Comm. v. Davidson, 236 F.3d 1174, 1198-1200 and n. 10 (10th Cir.2000) — and upheld in Gable v. Patton, supra, 142 F.3d at 944-45, and Kentucky Right to Life, Inc. v. Terry, supra, 108 F.3d at 646-48. Several cases, signally McIntyre itself, expressly or implicitly contrast the fragility of the small independent participant in political campaigns with large corporations or other organizations. McIntyre v. Ohio Elections Comm'n, supra, 514 U.S. at 348-49, 351-54, 115 S. Ct. 1511; Buckley v. American Constitutional Law Foundation, Inc., supra, 525 U.S. at 197-200; Citizens Against Rent Control/Coalition for Fair Housing v. City of Berkeley, 454 U.S. 290, 292-94 and n. 4, 298-99, 308 n. 4, 102 S. Ct. 434, 70 L. Ed. 2d 492 (1981); Gable v. Patton, supra, 142 F.3d at 944-45; Doe v. Mortham, 708 So. 2d 929, 934-35 (Fla.1998); cf. First Nat'l Bank of Boston v. Bellotti, 435 U.S. 765, 777-78 and n. 13, 788-90, 792 n. 32, 98 S. Ct. 1407, 55 L. Ed. 2d 707 (1978); Seymour v. Elections Enforcement Comm'n, supra, 762 A.2d at 888.

The Indiana Supreme Court, recognizing this last point, did a bit of judicial legerdemain, expanding the statutory exemption for mailings of up to 100 pieces of "mail" that are "substantially similar," Ind.Code § 3-9-3-2.5(a)(9), "to include any form of delivery of any written material, including personal delivery or use of some service other than use of the United States Postal Service," with the result that "Indiana's law permits some individual pamphleteering and applies only to candidate elections." Majors v. Abell, supra, 792 N.E.2d at 27 n. 11, 28. The statutory exemption as expanded by judicial interpretation protects the most vulnerable independent contributors to political advocacy. And as we said earlier, to require only the reporting of the advertiser's name to a public agency, while it would as a practical matter allay some of the anxieties of potential advertisers, would at the same time reduce the amount of information possessed by voters. Both sides of the First Amendment balance would be depressed. We cannot say that the net effect of invalidating the Indiana statute would be to promote political speech.

As an original matter it could be objected that speech and the press would no longer be free if the government could insist that every speaker and every writer add to his message information that the government deems useful to the intended audience for the message, and that it is arbitrary for the government to single out the identity of the writer or speaker and decree that that information, though no other that potential voters might value as much or more, must be disclosed. But the Supreme Court crossed that Rubicon in McConnell. Reluctant, without clearer guidance from the Court, to interfere with state experimentation in the baffling and conflicted field of campaign finance law without guidance from authoritative precedent, we hold that the Indiana statute is constitutional.

The decision of the district court is modified to place the dismissal of the suit on the merits, and as so modified is

AFFIRMED.

EASTERBROOK, Circuit Judge, dubitante.

Four decisions of the Supreme Court hold or strongly imply that the ability to speak anonymously — and thus with less concern for repercussions — is part of the "freedom of speech" protected by the first amendment against governmental interference. Talley v. California, 362 U.S. 60, 80 S. Ct. 536, 4 L. Ed. 2d 559 (1960); McIntyre v. Ohio Elections Commission, 514 U.S. 334, 115 S. Ct. 1511, 131 L. Ed. 2d 426 (1995); Buckley v. American Constitutional Law Foundation, 525 U.S. 182, 199-200, 119 S. Ct. 636, 142 L. Ed. 2d 599 (1999); Watchtower Bible & Tract Society of New York, Inc. v. Stratton, 536 U.S. 150, 166-67, 122 S. Ct. 2080, 153 L. Ed. 2d 205 (2002). See Jonathan Turley, Registering Publius: The Supreme Court and the Right to Anonymity, 2001-02 Cato Sup.Ct. Rev. 57. Although the scope of protected speech has been held to differ across subject matter, the ability to denounce public officials by name and call for their ouster is the core of the Constitution's protection. See New York Times Co. v. Sullivan, 376 U.S. 254, 84 S. Ct. 710, 11 L. Ed. 2d 686 (1964).

There have been times and places in the United States when opposing elected political officials risked both wealth and health. Decisions such as O'Hare Truck Service, Inc. v. Northlake, 518 U.S. 712, 116 S. Ct. 2353, 135 L. Ed. 2d 874 (1996), and Rutan v. Republican Party of Illinois, 497 U.S. 62, 110 S. Ct. 2729, 111 L. Ed. 2d 52 (1990), hold that government is not entitled to retaliate, but what people do in fact often differs from what they should do, and litigation after the fact is no cure-all. It is no cure at all for retaliation by private actors, who are free to penalize those whose political views they do not accept, provided only that they do not cross the line into violence — and that line has not always been observed; think of persons whose candidates were the "wrong" race. Anonymity thus may be especially valuable when opposing entrenched actors. Disclosure also makes it easier to see who has not done his bit for the incumbents, so that arms may be twisted and pockets tapped. Labor law deems it improper for employers to nose out union adherents' names; judges and members of the NLRB perceive that knowledge may facilitate retaliation and that fear of this outcome will stifle speech. Yet although union organizers may operate in secret, and everyone may vote in secret (our adoption of the Australian ballot came from awareness that disclosure could affect political support), political advocates must disclose their identities. Today the court holds that a state may require persons engaged in core political speech to identify themselves so that the officeholders and their allies can pinpoint their critics. How can this be?

According to my colleagues, the answer lies in the fact that McConnell v. Federal Election Commission, 540 U.S. 93, ___-___, 124 S. Ct. 619, 689-94, 157 L. Ed. 2d 491 (2003), rejected a constitutional challenge to § 201 of the Bipartisan Campaign Reform Act of 2002, which amended § 304 of the Federal Election Campaign Act, 2 U.S.C. § 434. Section 304 as amended requires any person who makes disbursements exceeding $10,000 in any year for speech in federal campaigns, or who donates $1,000 or more to another person or group engaged in advocacy, to disclose his identity to the Federal Election Commission. Indiana's law differs — it starts from a lower threshold (100 sheets of paper) and requires disclosure to the public in the electioneering literature rather than to an agency, see Ind.Code § 3-9-3-2.5 — but once it is settled that speakers must reveal their identities directly to the political establishment, five Justices may think that everything else is mere detail.

Still, the Justices' failure to discuss McIntyre, or even to cite Talley, American Constitutional Law Foundation, or Watchtower, makes it impossible for courts at our level to make an informed decision — for the Supreme Court has not told us what principle to apply. Does McConnell apply to all electioneering? All speakers? To primary communications (as opposed to notices sent to agencies)? The Supreme Court wrote that § 304 is valid because it is (in the view of five Justices) a wise balance among competing interests. Yet the function of the first amendment is to put the regulation of speech off limits to government even if regulation is deemed wise. See American Booksellers Ass'n v. Hudnut, 771 F.2d 323 (7th Cir.1985), affirmed, 475 U.S. 1001, 106 S. Ct. 1172, 89 L. Ed. 2d 291 (1986). See also John Hart Ely, Flag Desecration: A Case Study in the Roles of Categorization and Balancing in First Amendment Analysis, 88 Harv. L.Rev. 1482 (1975); Geoffrey R. Stone, Restrictions on Speech Because of its Content: The Strange Case of Subject-Matter Restrictions, 46 U. Chi. L.Rev. 81 (1978). For the judiciary to say that a law is valid to the extent that it is good is to operate as a council of revision and to deny the power of a written constitution to constrain contemporary legislation supported by the social class from which judges are drawn. And when, as in McConnell, the judgment is supported by a one-vote margin, any Justice's conclusion that a particular extension is unwise will reverse the constitutional outcome. How can legislators or the judges of other courts determine what is apt to tip the balance?

Footnote 88 to the lead opinion in McConnell, 540 U.S. at ___-___ n. 88, 124 S.Ct. at 696-97 n. 88, hints that one or more members of the majority may believe that the validity of the federal statute depends on its entire complement of rules. This footnote — the only place in which a majority opinion discusses McIntyre (though not when dealing with § 304!) — says that "BCRA's fidelity to those imperatives" sets it apart from the law held invalid in McIntyre. This treats the statute as a unit. It is difficult to say that the Indiana legislation at issue today displays "fidelity" to the "imperatives" of curtailing public corruption while allowing room for expression. As far as I can see, it has nothing to do with the risk of subtle bribery, and it attaches no weight to the risks borne by supporters of unpopular candidates. The majority in McConnell emphasized that the disclosure to the agency did not include the content of the advertisement. 540 U.S. at ___ - ___, 124 S.Ct. at 693-94. In Indiana the disclosure is affixed to the speech; the association is unavoidable; does this make a difference? My colleagues think not; I am not so sure.

Doubtless "a speaker's credibility often depends crucially on who he is." 361 F.3d at 352. But how does this support obligatory disclosure? Speakers who prefer concealment in order to reduce their personal risks, and who accept the discount that readers attach to advocacy from unnamed sources, do not impose burdens on strangers. What then is the justification for regulation? "People are intelligent enough to evaluate the source of an anonymous writing. They can see it is anonymous. They know it is anonymous. They can evaluate its anonymity along with its message, as long as they are permitted, as they must be, to read that message." McIntyre, 514 U.S. at 348 n. 11, 115 S. Ct. 1511. Arguments that speech may be regulated to protect the audience from misunderstanding should fare poorly and outside of electioneering have fared poorly. See, e.g., Virginia State Board of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U.S. 748, 769-70, 96 S. Ct. 1817, 48 L. Ed. 2d 346 (1976). Anyway, we must consider the possibility that anonymity promotes a focus on the strength of the argument rather than the identity of the speaker; this is a reason why Madison, Hamilton, and Jay chose to publish The Federalist anonymously. Instead of having to persuade New Yorkers that his roots in Virginia should be overlooked, Madison could present the arguments and let the reader evaluate them on merit.

Trade associations and other interest groups will have little difficulty complying with Indiana's law. Factions that hope to secure political favors enjoy legal counsel who specialize in election matters. Professionals in the field not only will assure compliance but also will exploit the inevitable loopholes. The identity of these interest groups is no mystery; many operate from marble-clad buildings and deploy full-time lobbyists. Statutes such as Indiana's have their real bite when flushing small groups, political clubs, or solitary speakers into the limelight, or reducing them to silence. Indiana's statute, which requires disclosure from the first dollar of speech, bears especially heavily on political outsiders. Indiana has essentially forbidden all spontaneous political speech, perhaps all electioneering by individuals and small groups. Before favoring or opposing any candidate, a would-be speaker must navigate a thicket of rules.

These laws and regulations are written in language that only specialists can fathom. For example, Indiana requires a "disclaimer" of identity; yet, as my colleagues observe, the state uses this word to mean the opposite of its normal connotation. In everyday language, a disclaimer is a repudiation or denial of responsibility. In Indiana's election code, however, that word denotes a statement accepting responsibility or authorship — a pro claimer (or just a "disclosure") rather than a dis claimer. Getting through this kind of double-talk requires help. Even a lawyer might not be enough: answering the question that we had certified, the Supreme Court of Indiana held that both the state's executive branch and the federal district judge had misunderstood the law's coverage. Majors v. Abell, 792 N.E.2d 22 (Ind.2003). (George Orwell, who coined the term "Newspeak" for evasive governmental expression, used a pseudonym to conceal his own identity. Anonymity did not reduce the power of his work or justify mandatory disclosure. Orwell might call Indiana's use of language doubleplusungood.)

Often the Supreme Court says that even a small fee or tax, or a short delay in obtaining a free license (as in Watchtower), is an unacceptable burden on speech. Cf. United States v. Playboy Entertainment Group, Inc., 529 U.S. 803, 120 S. Ct. 1878, 146 L. Ed. 2d 865 (2000). Yet in McConnell the Court was sanguine about the delays, and non-trivial legal expenses, entailed in complying with complex rules for campaign speech. These outlays come on top of the costs that must be borne by persons who back the wrong horse and incur the enmity of elected officials — for the winners now are entitled to learn all of their vocal opponents' identities. Maybe these effects can be justified with respect to electioneering at the national level by deep-pocket interest groups — though I think that the Justices have been too ready to equate political support to bribery, see Ronald A. Cass, Money, Power, and Politics: Governance Models and Campaign Finance Regulation, 6 Sup.Ct. Econ. Rev. 1 (1998) — but for local elections the equation is impossible to sustain.

Indiana does not contend that requiring disclosure by plaintiffs Carol Antun, Perry Metzger, and Bruce Martin — who want to use their own resources to speak on behalf of candidates of the Libertarian Party (and oppose incumbents, for libertarians do not occupy any major office in Indiana) — is essential to avert a material risk of underground favor-trading or bribery. Nor does the state try to justify mandatory disclosure by any truly independent speaker. Instead Indiana contends that it is entitled to regulate all electioneering by every speaker in order to avoid drawing lines. Given McConnell, I cannot be confident that my colleagues are wrong in thinking that five Justices will go along. But I also do not understand how that position can be reconciled with established principles of constitutional law.