Updating blockchain-based protocols remains a significant challenge. If the community does not come to an agreement, a hard-fork can occur, splitting the blockchain’s community. Previous protocols have provided mechanisms to establish community consensus through the protocol itself, but these protocols either facilitate substantial, infrequent updates, or they allow more frequent but only minor changes. This work offers a mechanism that allows clients to vote by locking tokens, making the clients’ tokens temporarily unavailable in exchange for their vote. This design introduces an economic cost to voting, allowing us to measure both breadth and depth of support. Since there is an economic cost to voting, we wish to make non-contentious issues cheap to pass, but still allow the community to establish agreement on larger, more disputatious proposals. We achieve this property by a ping-pong governance model. An issue is tentatively accepted when it achieves enough votes within a fixed period. The proposal then enters a review period, where the opponents must gather enough votes to veto it. The supporters then have their own opportunity to overrule the veto. This process continues with new voting rounds until one side is unable to exceed the needed threshold, settling the issue. Our simulations show that this model allows the community to come to agreement quickly on popular changes, but still come to resolution when the community is more divided. Finally, we define the ideal properties of a blockchain governance protocol, and evaluate different governance protocols under these criteria.