Abstract
Should ENS default Snapshot proposals to shielded voting? Snapshot already supports it natively via Shutter. Votes are encrypted during the voting window and decrypted once the proposal closes. Everything after the close stays public, exactly as today.
Motivation
- Produces sincere preference revelation. Voters express what they actually believe instead of what’s popular or expected.
- Neutralizes strategic voting. Voters can’t game a mechanism whose state they can’t see. With transparent tallies, large holders wait until the end to vote, sniping just enough to flip outcomes. Shielding removes the informational advantage of voting late. Rossello (2024) documents this empirically across 75 DAOs and names it “sniping blockholders.”
- Decreases politics, coercion and social pressure. Voters suffer less lobbying for choices that aren’t visible until the decision is locked in.
- Audit trail without live manipulation. Delayed reveal preserves the accountability that makes on-chain governance valuable, neutralizing the obvious counterargument.
What changes, what doesn’t
While the vote is open, you can still see who has voted, but not what they voted for or how the tally is going. The moment the vote closes, everything gets decrypted: addresses, weights, and choices all become public, just like today. Forum discussion and delegate rationales work the same way they always have.
Open questions
How should we go about implementing this? Pilot it on one class of proposals first, or flip the default for the whole space?
Personally I can only see good outcomes, so I’d prefer to have all votes being shielded. But I’d love to hear concerns or motives to not do this.
Rossello, R. (2024). Blockholders and Strategic Voting in DAOs’ Governance.