The decision, in Doe v. Reed, was just handed down yesterday, and is based on the signers' First Amendment rights to speak anonymously; the decision basically keeps in place the temporary restraining order handed down July 29, but is accompanied with a detailed opinion containing the court's reasoning.

That reasoning strikes me as unpersuasive, for the same reasons I mentioned in late July; I don't think that secrecy of signatures is constitutionally mandated by the First Amendment, just as I don't think that a secret ballot is constitutionally mandated by the First Amendment. True, the anonymous speech precedents bar the government from requiring that people sign their political statements. But political statements are just speech. Signing an initiative, referendum, or recall petition is a legally operative act — it helps achieve a particular result not just because of its persuasiveness, but because it is given legal effect by the state election law.

The government is surely entitled to require that people who want their signature to have such a legally operative effect must disclose their identities to the government. And I see no reason why the government might not then disclose those identities to the public, who after all are in charge of the government. To do that is to inform the people about who is taking legally operative steps to change the state's laws (or the state's elected representatives, in the case of a recall).

Informing the public about this might well deter such legally operative acts, though of course leaving people free simply to engage in persuasive speech, which can indeed generally be done anonymously rather than in legally significant signing of petitions. But I don't think that deterrence is unconstitutional, especially since the legal significance of the signature is there only because state law creates it. Even overt government condemnation of certain speakers is not a First Amendment violation, though such condemnation might deter speakers. The same is even more true, I think, of a simple release of over 100,000 names, though again I can't deny that there would be a deterrent effect.

I agree that there are plausible plausible arguments that voter signatures shouldn't be publicly released by the government. Just as we have a secret ballot for the ultimate votes, we could have at least a quasi-secret signature system for the signing of referendum, initiative, recall, and candidate nomination petitions. Both an election and a threshold signature requirement to put something on an election are generally aimed at accurately measuring public opinion. Such an accurate measurement is much more likely if the measurement is undistorted by people's fear of being attacked, fired, ostracized, or even annoyed by those who disagree with them.

To be sure, unlike with a secret ballot, a petition signature would not be fully secret -— for instance, the government would know what you signed, though it doesn't know how you voted, and it's possible that the signatures would be briefly visible as other people are signing the petitions (though that could be minimized, for instance if there's just one signature per page, and each page is concealed after it's signed). But there are good reasons why we might choose to make it as close to a secret ballot as possible.

Still, the judgment about how secret signatures or even ballots should be is a judgment that should be made legislatively (or by voter initiative). The First Amendment and First Amendment caselaw does not preclude either option, and the court's opinion here doesn't persuade me to the contrary.