A proposal in Starknet's governance forum would let validators vote with their delegators' staked $STRK by default, shifting governance power toward the operators users choose to stake with unless holders opt out.
The plan, SNIP-45, is scheduled for a governance vote from Nov. 27 to Dec. 4, 2026. It has not been approved. Any delegator could keep their own vote or redirect it to any address in one transaction.
The proposal targets a largely inactive voting base on Starknet, an Ethereum Layer 2 network. According to the draft specification, an average of just 0.5% of staked $STRK participated across the three votes and three polls it analyzed. Delegators, rather than validators staking their own tokens, hold 91% of staked $STRK, it says.
Know when your
coins move
The trade-off is the potential for more staked $STRK to be voted, in exchange for concentrating default voting power in validators. The draft explicitly warns that large operators, including exchanges and liquid-staking providers, could gain significant weight. The change would not affect holders who are not staking.
A New Default, Not a Staking Overhaul
Today, delegators must vote themselves for their staked $STRK to count, the draft says. Under SNIP-45, that voting power would instead go to their pool's validator, or a separate voting address the validator names. An explicit choice by a delegator would override the default. Validators could also decline default delegations, leaving those holders with their own votes.
A proposed onchain registry would record those preferences without holding funds. The specification calls for an immutable contract with no owner, upgrade or pause powers; each address could change only its own settings. Staking contracts, rewards and exit windows would remain unchanged.
Voting power from staked $STRK would still be calculated offchain using balances and preferences recorded at each proposal's snapshot block. Changes made afterward would not affect that proposal. The Starknet Foundation would publish the resulting stake-derived voting-power list, calculation script and inputs for both binding governance votes and nonbinding polls.
The proposal's back-test illustrates how more stake could be represented without more individuals voting. It adds pool members' stake to validators that actually voted in two earlier decisions; the results are retrospective estimates, not forecasts. If a validator does not vote, the voting power delegated to it by default would remain unused.
SNIP-45's Turnout Back-Test
Share of staked $STRK voting, as reported in the proposal. 'Turnout today' means turnout in the original decision. The back-test adds pool members' stake to validators that voted; it is not a forecast. Source: SNIP-45 governance-forum post.
The Foundation is scheduled to deploy the registry and publish its address before the vote. If SNIP-45 passes, the new default would apply only to proposals whose snapshot block falls at least seven days after the result, giving delegators and validators time to register their preferences.


