Introducing Multiple Election Stake-Weight Aggregation (MESA) Vote for Reducing SOL Inflation

[Galaxy-Research]

Multiple Election Stake-Weight Aggregation (MESA) Vote for Reducing Inflation

Authors: Galaxy Research
Created: April 17, 2025

#Summary

This proposal details a market-driven process to deciding the SOL emissions curve. It is not to be confused with a market-driven curve as detailed in SIMD-228. Rather, it suggests a fixed, time-dependent disinflationary curve that is enforced each epoch and reduced by an annual percentage, as is currently implemented, but recommends a market-driven approach to deciding which annual deflation rate will be adopted in the future, rather than relying on individual proposals. This proposal suggests an innovative approach to deciding the future SOL emissions curve by allowing validators to vote on multiple deflation rates and settling on the aggregate as the outcome, which we call Multiple Election Stake-Weight Aggregation (MESA).

#Motivation
SIMD-228 was an important proposal with wide participation that sparked significant Solana community engagement. While there was plenty of debate, the community appeared to agree that SOL inflation and security overpayment should be reduced. What we learned through 228’s vote was less so that the community disagreed with the intention of the proposal and more so that limitations in the governance process impeded the community’s ability to converge on universal agreement around the parameters of such a highly impactful change to the network. In such a proposal where all network participants are impacted differently, and may seek a range of possible outcomes on the affirmative side, proposals with only YES / NO / ABSTAIN may not be the most efficient way to reach consensus.

A market-driven approach to governance proposed below allows the community to act on the shared belief that inflation is too high while maintaining predictability, preserving self-interest, and eliminating the need to repeatedly take the idea to single-outcome vote until a universally acceptable number is proposed. Instead of throwing darts until the community is happy with an individual proposal, it is more efficient to simply ask each person what they want and settle on the aggregate.

#Proposal

This proposal suggests maintaining the fixed, time-dependent disinflationary curve with the set 1.5% terminal rate that is already implemented (to limit input assumptions on its impact and maintain relatively higher levels of predictability) and sets forth multiple outcomes that steepen the rate of deflation (currently 15%) from which an average is aggregated. For example:

This hypothetical vote can have ABSTAIN, NO, and YES choices, where there is a spectrum of predetermined options that live under YES (e.g., YES-UNCHANGED, YES-17.5%, YES-20%, and so on). Assuming the quorum and minimum YES threshold (when all YES sub-votes are aggregated) are reached, a vote weighted average of the percentages chosen under YES determines how steep the curve becomes. In an example where 5% of the YES vote chooses no change to the curve (so it stays at 15%,) 50% of the vote elects to change the rate of deflation to 30%, and 45% of the vote chooses to change to 33%, the inflation curve would deflate at a new rate of 30.6%, or (.05 * .15) + (.5 * .3) + (.45 * .33).

This approach is much more market-driven than single-outcome votes, effectively allowing each individual validator to vote for the new inflation schedule that best aligns with how they or their delegators’ view, assume, or prefer the future, while ultimately maintaining the predictability of a fixed curve. Paired with split voting, we think this market driven approach to protocol governance is democratic and progressive, offering a new method for protocol governance.

To facilitate this, the vote would have the typical ABSTAIN and NO accounts with multiple YES accounts bucketed by the proposed percentage:

• ABSTA1Nsimdxxx….
• Nosimdxxx….
• YES15simdxxx….
• YES175simdxxx….
• YES20simdxxx….
• YES225simdxxx….
• YES25simdxxx….
• YES275simdxxx….
• YES30simdxxx….

The chart below highlights how quickly SOL’s inflation curve could reach the 1.5% terminal rate based on a spectrum of updated deflation rates (measured in epochs). For example, if the deflation rate is left unchanged at 15% the terminal rate will be reached around epoch 2,135 (1,362 epochs after epoch 772, or ~7.4 epoch years).

#Recommended Next Steps and Things to Consider

• Determining the spectrum of YES votes: is it a barbell of unchanged at 15% and some established number > 15% (allowing validators to break up their votes to reach some average in between)? Or should a gradual step up in options be presented (e.g. 15%, 17.5%, 20%, and so on)?
• Determining quorum and participation constraints: should SIMD-228’s 33% quorum and 2/3 supermajority apply? In this case, a 2/3 supermajority of votes would be across the YES spectrum (Y = (Y / Y+N) > 2/3 with Y = total votes for any of the potential deflation rates).
• Is a weighted average of YES election values appropriate or would some other math be more fair (average is the simplest)?

In making this proposal, Galaxy Research seeks to suggest a genuinely alternative process to achieving what we believe is the community’s broad goal, and not necessarily proscribe any particular inflation rate outcome. We welcome all input, feedback, or criticism, and look forward to engaging with the Solana community on the future of the protocol.

• Galaxy Research
• [email protected]
• x.com/glxyresearch
• galaxy.com/research

An affiliate of Galaxy Research, Galaxy Strategic Opportunities LLC (the “Galaxy Staking Affiliate”), provides staking and validation services in respect of SOL. While the foregoing proposal speaks only to the details of a market-driven process to deciding the SOL emissions curve and does not recommend any particular outcome, the Galaxy Staking Affiliate may benefit from such a vote depending upon its outcome. Please refer to this link for additional important Galaxy Research disclaimers.

[MaxResnick]

Thanks for taking the time to write this up. I like the idea of allowing validators to express a richer set of preferences than just yes/no on the proposal. I am concerned that the average aggregation is not incentive compatible and would lead to a lot of headaches in the voting process.

Suppose I believe the best policy is 25% a year. How should I vote to make the resulting policy as close to 25% as possible? Ideally everyone would truthfully express their policy preferences and we would aggregate them. But with the average aggregation rule the best thing to do is try to forecast where the final outcome will be and set the most extreme policy in whichever direction you want to pull the policy from there.

Choosing the Median policy would be a truthful aggregation rule which would simplify the strategy of voting for validators (the best thing you can do is report what you want the policy to be).

I continue to believe that a dynamic market based approach to issuance is better for the health of the network than a static rate. Obviously, the simplicity of the static curve is desirable, but there are tradeoffs. A dynamic issuance curve is more secure in bad times and less costly in good times than a static curve. I have faith that the Solana community is Intelligent enough to understand a dynamic inflation policy and do not need to be coddled by "left curve" proposals which are demonstrably worse for the network than their "right curve" counterparts. In no other part of the stack would we say something like: "turbine is too complicated for block propagation we should do a 'left curve' version". But somehow when it comes to the core economics of the network this seems to be an acceptable line of thinking.

In particular the nice thing about this sort of proposal is that it will result in a much less polarizing vote than the one we had for 228. Yes/No voting forces people into separate camps which creates unproductive conversations. People inevitably focus on more on "winning" than on helping Solana win.

[aeyakovenko]

If we end up with median, why not just add a way for validators to adjust the rate via median stake weight continuously.

[Teddy-James]

MESA is a smart step toward more flexible and inclusive governance. By letting validators vote across a range of deflation rates and using a weighted average, it avoids the pitfalls of binary decisions. This approach to consensus reflects the importance of fair representation in decision-making, similar to the broader discussions on ensuring that every voice counts.