> I have an intuition that personal privacy is important in maintaining democracy.
I don't think this is true. People need social pressure of some kind to cooperate. It could be positive pressures like getting positive feedback from peers, or negative feedback in the form of shame from peers.
Authenticity is important here, to ensure that one's peers can provide accurate feedback (both positive and negative). Anonymity breaks this model, allowing people to pretend to have one viewpoint
in order to reap the positive rewards, while simultaneously allowing them to dodge the negative feedback for antisocial behaviors.
You needn't look further than the behavior of pseudonymous commenters on the Internet to see this play out. So many "normal" people are capable of being complete assholes online because they're able to hide those actions from those who can impose social consequences.
I mean, if I found out a close friend was the kind of person who demonstrates the kind of casual cruelty that online trolls exude, I might stop being their friend.
That's not too say that the benefits of privacy are all bad. Vulnerable people certainly benefit greatly from privacy, especially in adversarial environments... But perhaps the answer in less adversarial environments (i.e. direct/indirect democracies) more often isn't "more privacy", but "less judgement"?
The other framing for this question involves trust. Anonymity undermines trust, but trust is _required_ for coordinated action. From this perspective isn't pretty obvious that democracies and privacy pursue opposing goals.
Put differently, when it comes to voting, I don't think it's a good idea to incentivize lying--or worse still, actively promote it as a positive. If making votes public causes negative externalities, perhaps there are solutions other than "let people lie about how they vote"?