ETH’s job in Ethereum

Ethereum runs a general-purpose execution environment for smart contracts. ETH is the asset the protocol uses to (1) meter scarce computation and state access, and (2) collateralize consensus security through staking. Those two roles matter more than any “money” framing because they determine who must hold ETH, when they must spend it, and what happens to that spend.

On L1, every transaction fee is denominated in ETH and expressed via gas. Under the post-London fee market, what you pay is split into a protocol-set base fee and a user-set priority fee (tip), as described in Ethereum’s gas and fees documentation. The base fee is burned. The tip is paid to validators as the inclusion incentive.

On the security side, validators lock ETH and earn rewards for correct participation, with penalties and slashing for faults under the protocol’s rewards and penalties rules. This is not a side feature. It is the issuance sink. ETH’s issuance policy is basically “pay enough to keep honest stake online,” then let fee revenue and burns push net issuance around that baseline.

History that changed ETH’s economics

August 5, 2021 is when ETH’s “burn narrative” became mechanically real. The London network upgrade activated EIP-1559 on mainnet, changing fees from a single auction price to base fee + tip, with base fee burning.

September 15, 2022 is when the security budget regime changed. The Merge moved Ethereum from proof-of-work to proof-of-stake.

April 12, 2023 is when staked ETH became meaningfully liquid again. The Shanghai/Capella upgrade enabled withdrawals.

March 13, 2024 is when Ethereum introduced blobs as a dedicated data-availability lane for rollups via Dencun, centered on EIP-4844. This matters for tokenomics because it changes what users buy (execution gas vs blob gas) and how much fee pressure remains on classic L1 blockspace.

May 7, 2025 is when Pectra shipped, including EIP-7251 (raising max effective balance), and EIP-7691 (higher blob throughput). These are not cosmetic. They alter validator operations and blob fee-market capacity, which feed directly into issuance-versus-burn outcomes.

Supply, issuance, and genesis distribution

ETH does not have a fixed cap. Supply is tracked on-chain and reduced by burns. For a burn-forward contrast, see our BNB tokenomics review.

The first-order analytical mistake is treating “burn” as a monetary policy target. It is not. EIP-1559 itself states the system cannot assert whether ETH will be inflationary or deflationary because demand for blockspace is not controllable, and that the change reduces direct control over long-term quantity.

Post-Merge, issuance is primarily a function of total active stake. Ethereum’s PoS reward math makes that concrete. The base reward depends on protocol constants and on the total staked ETH (sum of active validator balances), and the docs note that more validators increases overall issuance roughly with sqrt(N) while decreasing per-validator rewards roughly with 1/sqrt(N).

Before the Merge, issuance also came from proof-of-work mining rewards. Ethereum’s own post-Merge issuance write-up summarizes this shift as moving from dual-source issuance to execution-layer issuance going to zero after the Merge.

Genesis distribution is unusually important context for ETH because the asset was bootstrapped via an early sale and organization allocations, not a “pure mining” launch. The Ethereum whitepaper specifies a sale amount of 60,102,216 ETH and two additional allocations each equal to 0.099x of the amount sold, plus an original plan for ongoing miner issuance.

The whitepaper also proposed perpetual ongoing issuance to miners equal to 0.26x the amount sold per year. That historical detail matters less as a current parameter and more as evidence of intent. Ethereum’s monetary policy has always been designed to support security and development, not to preserve a hard-capped scarcity narrative.

Fees, burns, and the real fiscal flow

The burn mechanism is simple in concept and easy to overinterpret in valuation. Each block has a base fee. It is computed from prior blocks and adjusts based on congestion, with a maximum change of 12.5% per block and explicit burning of the base fee when a block is produced.

Users can add a priority fee, which is the part that actually compensates the block producer for including a transaction. Total fee is base fee + tip: base fee is protocol-set and burned, and the tip is what makes a transaction attractive to validators.

This creates a fiscal split:

Burn: a demand-dependent sink that scales with L1 fee pressure and is structurally volatile.

Validator revenue: a mix of newly issued ETH and recycled ETH paid as tips and MEV-related income.

If you want a burn-skeptic read that stays honest, keep one thought in your head: burns do not fund security. Burns reduce supply. Security is funded by validator rewards and fee revenue that is not burned.

Blobs extend this logic to rollup data availability. EIP-4844 introduces blob gas with its own targeting rule “similar to EIP-1559,” and specifies that the blob fee is deducted and burned, with no refund even if the transaction fails.

Capacity changes then feed back into burn dynamics. Pectra’s EIP-7691 raises blob throughput to a target of 6 blobs per block and a max of 9, and updates blob gas parameters used by clients. More capacity usually means lower unit fees when demand is unchanged. That is great for users. It also weakens any expectation that blob-related burns will be a persistent scarcity engine.

The burn is best understood as an automatic, variable “protocol buyback.” It is not guaranteed. It is not smoothed. It is not designed to be stable across market cycles. That is a feature for UX and fee predictability. It is a bug for anyone trying to underwrite long-run ETH value on burn alone.

Staking economics: issuance curve, consolidation, and liquidity gates

Post-Merge, issuance is the explicit security spend. Ethereum’s supply documentation summarizes the shift as execution-layer issuance going to 0 after the Merge, with rewards flowing from the consensus layer, while the burn continues as an opposing force.

The pre-Merge world had a clearer mental model: miners got block rewards, fees went to miners, and issuance was visibly inflationary. Ethereum’s supply page uses an approximate pre-Merge issuance magnitude of ~13,000 ETH/day for miners, which it contrasts with much lower PoS-era issuance.

In PoS, issuance is elastic with respect to total stake. The base reward formula depends on total active stake and constants including base_reward_factor = 64 and base_rewards_per_epoch = 4, and total issuance grows as sqrt(N) as the validator set grows.

Pectra’s EIP-7251 changes validator accounting constraints in a way that is easy to miss if you only watch APR dashboards. The EIP keeps the minimum activation balance at 32 ETH while raising MAX_EFFECTIVE_BALANCE to 2048 ETH, enabling consolidation of multiple 32 ETH validators and allowing compounding within a single validator.

That is an operational efficiency upgrade. It does not automatically decentralize stake. It can reduce network overhead. It can also make it easier for large operators to manage massive positions with fewer keys and less infrastructure, which is neutral at best for concentration risk.

Liquidity gates matter because they affect how quickly staked ETH can respond to price shocks and incentives. Withdrawals became possible after the Shanghai/Capella upgrade on April 12, 2023.

Ethereum also rate-limits exits to maintain stability. Its issuance documentation describes caps on how many validators can exit per epoch, with a baseline of four validators per epoch and an adaptive increase as validator count grows.

From a burn-skeptic perspective, staking is where ETH’s “economic floor” actually lives. If fees go quiet, burns fade. Security spend does not disappear. It shifts back toward net issuance, and ETH becomes inflationary again. That is not a moral judgment. It is simply the design’s equilibrium behavior.

Governance and parameter control

ETH is not a governance token in the on-chain voting sense. Ethereum’s governance is primarily off-chain. Protocol changes flow through discussion, the EIP process, core developer coordination, client implementation, and then voluntary adoption by node operators. If you want the framework we use to evaluate these tradeoffs, see our tokenomics methodology page.

The EIP process is formalized. Ethereum.org’s EIP overview points readers to EIP-1 and notes that proposals are discussed publicly before drafts are submitted.

EIP-1 is the canonical process document describing what EIPs are and how they are proposed and curated.

For tokenomics analysis, the key implication is uncomfortable but important: ETH’s monetary parameters are ultimately a product of social consensus. They are not protected by a hard cap. They are protected by norms, incentives, and the coordination costs of changing a live system. That is a real form of constraint. It is not the same as “cannot change.”

Even the Ethereum Foundation’s upgrade announcements emphasize opt-in. Upgrades require explicit adoption by node operators, and client developers are not the ultimate deciders of whether an upgrade is adopted.

This is where “burn as guarantee” breaks down. Burn can be a strong contributor to net scarcity during high-fee periods. It does not remove governance discretion over issuance mechanics in the long run.

Risk analysis

ETH’s tokenomics work best when you accept the design goal: balance security and usability, and let supply outcomes float. They strain when the market prices ETH like a deterministic scarcity asset, while the protocol is engineered as a variable-rate security budget with an opportunistic burn. For a high-throughput L1 comparison, see our Solana tokenomics review.

Top 3 risks

  1. Security budget compression if fee + MEV revenue falls faster than staking participation adjusts. Trigger: sustained periods of low L1 fee pressure, including lower execution fees and lower blob fees as throughput rises. Mechanism: base-fee burns fall because they are demand-driven, while validator compensation still must clear the hurdle rate needed to keep sufficient stake online, pushing net issuance upward or increasing reliance on tips/MEV. Who bears it: ETH holders (via higher net issuance and weaker “scarcity” support), and Ethereum users (if governance responds by shifting fee policy or issuance to protect security). Indicators: net issuance trending positive over long windows, declining base-fee burn per day, declining priority fees per block, and declining staked ETH participation rate relative to supply.

  2. Staking concentration and correlated operational risk. Trigger: increasing consolidation of validators and continued dominance of a small set of operators, amplified by operational efficiency improvements. Mechanism: fewer entities control more effective stake, raising tail risks from correlated downtime, governance influence, and MEV supply-chain dependence, even if the protocol is formally permissionless. Who bears it: users (liveness and censorship risk), and smaller stakers (relative influence and potential for adverse selection in delegation markets). Indicators: rising share of stake controlled by top operators, increased incidence of large-scale correlated misses, and client diversity deterioration (a related systemic risk often highlighted in upgrade communications).

  3. Monetary policy model risk from social-layer changeability. Trigger: contentious periods where major stakeholders advocate parameter changes to issuance, burn, or fee mechanics. Mechanism: Ethereum’s off-chain governance and upgrade opt-in makes parameters adaptable, which is good engineering, but reduces certainty for long-horizon “fixed policy” valuation models. Who bears it: long-duration holders and applications depending on stable fee semantics. Indicators: increased volume of monetary-policy discussion, repeated proposals targeting issuance/burn parameters, and contentious upgrade coordination signals.

If you want ongoing tracking of these kinds of indicators, we publish related analysis in our research reports.

Dominant risk: Security budget compression driven by successful scaling.

The dominant risk is not “ETH becomes inflationary again.” That can be perfectly fine. The dominant risk is that the market’s mental model lags the protocol’s actual cashflow physics, and that mismatch feeds governance pressure in the worst possible direction.

Ethereum’s fee burn is mechanically linked to demand for L1 blockspace, not to Ethereum’s “economic activity” in the broader sense. If activity migrates to L2s and blob capacity rises, the system can succeed at scaling while reducing the fee intensity that produces burns. Dencun explicitly introduced blobs to reduce L2 costs by moving data into a dedicated lane. Pectra then increased blob throughput again via EIP-7691.

At the same time, issuance exists to pay for security. It does not shut off when burns are low. It adjusts with total stake via the PoS reward curve. The Merge removed execution-layer issuance, but it did not remove the requirement to compensate validators.

Put those together and you get a structurally pro-cyclical outcome:

When demand is hot, burns rise, net issuance can fall, and ETH looks “scarce.”

When demand cools or moves to cheaper lanes, burns fall, net issuance rises, and the scarcity optics weaken.

That is fine if everyone prices ETH as a security-collateral asset with variable net issuance. It becomes a problem if large segments of the market price ETH as a persistent deflation trade and then react politically when the protocol behaves as designed. Ethereum’s governance process is intentionally social and opt-in. That makes the system resilient. It also means monetary-policy debates can become cyclical when narratives outrun fee reality.

Practically, the “burn skeptic” stance is not anti-burn. It is anti-underwriting. Burns are valuable when they are the natural byproduct of genuine fee demand. They are weak when they are treated as a guaranteed yield to passive holders.

If you are designing a protocol that depends on ETH fee dynamics or staking incentives, treat ETH’s net issuance as a regime variable, not a constant. For teams that need an external review, a short tokenomics engagement can be useful to stress-test assumptions against protocol-level mechanics and upgrade-driven parameter drift.



This article is part of our Tokenomics Deep Dive series.