Galaxy Research has proposed an adjustment to the voting system that determines the future inflation rate of Solana.
We just introduced a new Solana proposal called Multiple Election Stake-Weight Aggregation (MESA) to reduce SOL inflation: a more market-based approach to agreeing on the rate of future SOL emissions.https://t.co/mcVdkRiM8y
— Galaxy Research (@glxyresearch) April 17, 2025
The initiative is named MESA (Multiple Election Stake-Weight Aggregation).
Instead of the traditional yes/no responses to determine the inflation level, the system will allow validators to vote on multiple deflation metrics and calculate a weighted average of the received options as the result.
The firm proposed MESA after the previous proposal, SIMD-228 failed to reach consensus on specific parameters. However, it did demonstrate unanimity regarding the need to reduce inflation.
SIMD-228 suggested a shift from a fixed schedule to a dynamic model: if the percentage of tokens in staking falls below the 33% threshold, SOL emissions would increase to incentivize participants. With a high volume of locked tokens, the metric would decrease, reflecting a reduced need to “overpay” for network security.
MESA maintains a fixed inflation rate of 1.5%, offering several deflation options for voting.
For example, if 5% voted for 15% deflation, 50% for 30%, and 45% for 33%, the new level would be 30.6%. The goal is to achieve a terminal inflation rate of 1.5%.
The aim of MESA is to offer a more market-driven approach, allowing validators to express preferences across a spectrum of deflation, while maintaining emission predictability.
Under the current mechanism, inflation starts at 8% per year and decreases by 15% annually until it reaches 1.5%. The current level is 4.6%.
According to Solana Compass, 64.7% of the total supply (387 million SOL) is locked in staking.
Back in February 19, Blockworks researcher Carlos Gonzalez Campo reported that annual SOL inflation rose by 30.5% following the implementation of a new fee distribution model on the platform.