Solana Inflation Cut Vote Explained: What SGP01/02/03 Mean for SOL Stakers

The Solana network is holding a critical governance vote this week, with three proposals on the ballot at the same time. Unlike most coverage that focuses only on "should inflation go down," what this round of voting actually decides is: how fast staking yields decline, how transaction fees are split and burned, and which set of standards the governance rules themselves are counted by.

The Current State of Solana's Inflation Mechanism

Solana's current actual inflation rate is 3.82%, not the 8% often cited by outsiders. The 8% was the starting point of the 2020 inflation schedule, which has since declined at 15% per year, eventually settling at 1.5% and stopping there.

There's an easily overlooked detail here: Solana's inflation schedule assumes 180 epochs per year (each roughly two days). But for much of 2021 to 2023, epochs actually ran longer. One developer compiled this deviation into a public dashboard, which showed the network's actual progress was about 276 days behind the existing schedule — in other words, three-quarters of a year's worth of disinflation that should have occurred never actually happened. This deviation also went through no vote.

What Each of the Three Proposals Does

There are three proposals in this round, each counted independently, and they can all pass, partially pass, or all fail.

SGP01: Governance Charter

This is a governing charter document that sets out how the other two proposals are counted and recognized. It does not directly change economic parameters itself.

SGP02: Doubling the Inflation Decay Rate

This proposal does just one thing: raise the annual inflation decay rate from 15% to 30%. The inflation floor remains 1.5%, and the starting point is unchanged; the only thing that changes is the time needed to reach the floor — from 5.7 years down to 2.8 years. The proposal was written by two engineers at Solana infrastructure company Helius.

SGP03: Fee Structure Restructuring and Burn

Solana currently charges a fixed base fee per transaction. This proposal splits it into two parts:

• A fixed inclusion fee (2500 lamports) paid to the validator that produces the block;
• A second fee component that floats based on the compute units requested by the transaction, and is entirely burned.

The Real Impact on SOL Staker Yields

This is the most contested part of this round of voting. According to the SGP02 proposal document's own estimates (based on a 68% staking participation rate assumption), if the proposal passes, the nominal staking yield would decline along the following path:

• Current: 5.84%
• Year 1: 4.34%
• Year 2: 3.00%
• Year 3: 2.25%

But the key point is that even if the proposal fails, yields will still decline, just more slowly. Under the current schedule, the corresponding path is 4.93%, 4.17%, 3.52%, with both paths ultimately converging to 2.26%. The reason is that the 1.5% inflation floor is not within the scope of this vote.

It should be noted that this model does not account for validator commissions, MEV, or block rewards, so it is a lower bound rather than a forecast. This also explains why one public company disclosed an actual net staking yield of 6.14%, higher than the 5.84% given by the model.

Knock-on Effects at the Validator Level

The proposal document modeled 738 validators. Currently 422 are profitable, 26 are breaking even, and 290 are operating at a loss — these 290 loss-making validators existed before this vote.

If the decay rate doubles:

• In Year 1, 2 additional validators turn unprofitable;
• By Year 2, a cumulative 13;
• By Year 3, a cumulative ~30.

Under the current schedule, reaching the same 30 unprofitable validators would take 6 years. In other words, the same outcome, just about four years earlier.

The break-even threshold for validators to cover costs is currently around 274,000 SOL in delegated stake; over the long term, under both schedules that threshold rises to 698,000 SOL — the only difference is that the proposal reaches it within 2.8 years, while the current schedule takes 5.7 years.

The Validator Proxy Voting Mechanism and the Voting Deadline

This is the most easily overlooked mechanism in this round of voting, yet it may directly determine the outcome.

If you stake natively through a validator, that validator will vote on your behalf by default. You don't need to do anything, and you may not even be notified. The validator's position is your position, unless you actively override it.

According to Solana governance documentation, native stakers can override the operator's position for an individual stake account before the validator votes, after the validator votes, or if the validator abstains, and:

• No need to unstake;
• No loss of staking rewards;
• Simply connect your wallet to the governance dashboard to vote yourself.

Awareness of this mechanism is extremely low. One staker posted on a forum saying that after searching for a long time he could not confirm whether native stakers could influence the vote, and at one point thought only validators had voting rights; 13 days later, he updated the same post saying he had just learned he could override his validator's vote without unstaking, and felt relieved — because he had delegated to a small validator that opposed the proposal, and he didn't want to abandon them just to express a differing opinion.

The voting deadline is the end of Epoch 1024, which Solana developers give as approximately Thursday 15:30 UTC.

Voting Progress and Subsequent Execution Risk

As of Tuesday morning, the governance dashboard data:

• SGP02 (inflation decay): 77.33 million SOL in favor, 9.7 million SOL against, 7.89 million SOL abstaining. Total system stake is 433.49 million SOL, with 21.91% participation. The quorum requirement is one-third, i.e. 144.5 million SOL, leaving a current shortfall of about 49.5 million SOL. Support among those who have voted is 88.8%.
• SGP01 (governance charter): 79.9 million SOL in favor, about 18.84 million SOL against, with 18.43% participation.

Notably, opposition to the governance rules themselves is minimal, with the real disagreement concentrated on the economic parameters — the volume of votes against the financial terms is about 500 times the volume of votes against the rules. At the same time, participation in the governance charter is actually lower than in the proposals it governs.

Two Conflicting Sets of Governance Rules

The rules shown on the governance dashboard are "one-third participation + two-thirds approval"; yet the policy document in the governance code repository states that there is no quorum requirement, and that approval only needs to exceed opposition by 2:1. The two rule sets cannot both hold, and as of now, the final confirmed code only records vote counts and does not resolve this conflict.

Historical Reference

The SIMD228 proposal in March 2025 held a similar vote: 910 validators participated, and 74% of staked supply voted. Calculated on total staked supply, in favor was 43.6%, against 27.4%, abstaining 3.3%, and the proposal failed. But if calculated as "in favor as a share of valid votes (for + against)," the approval ratio was 61.4%, just 5 percentage points short of the 66.67% threshold. The same proposal, measured with a different denominator, can see the gap narrow from 23 percentage points to 5 — which number you cite determines which version of the story you tell.

What's worth noting is that the main force that rejected SIMD228 that year was precisely the group of validators holding under 500,000 SOL, more than 60% of whom voted against. This proposal pushes their break-even threshold up to 698,000 SOL. The same group, the same arithmetic logic — but this time, they don't need to show up to win, they just need to stay silent.

Real-World Constraints at the Execution Level

Even if the proposal passes, no tokens will be burned on Thursday, and Friday's issuance won't change immediately. A passed proposal is only a directional mandate, and the code to implement it does not yet exist. The proposal assumes a lag of about 4.5 months before activation; if SIMD0550 is not merged and implemented before mid-January next year, then the result of this vote will be nothing more than an expression of sentiment.

From a mechanism design perspective, Solana has built a governance system that makes validators' economic interests clearly visible before you hand over your voting rights, and this structure itself works fine. The problem is that almost no one looks at it.