Yeah, no one actually does this. The point about cryptography was me being glib, but the math does exist. The relevant subfield is "homomorphic encryption" an

Yeah, no one actually does this. The point about cryptography was me being glib, but the math does exist. The relevant subfield is "homomorphic encryption" and I remember seeing a paper over the last few years (probably linked here on HN if I had to guess) claiming to have an election-relevant algorithm that was supposed to be workable in practice. I can't find it right now though.