As expected, our protocol (Protocol 9) makes some traditional performance for privacy tradeoffs. We analyze the differences between ours and existing protocols, primarily that of Protocol 7, below. In all of our comparisons we assume that the fraction of authorities needed to reconstruct a Shamir shared secret is p; that is, if a secret is shared among T authorities, pT of them are required to reconstruct the secret and (1-p)T of them are needed to prevent reconstruction.