> For the complete documentation index, see [llms.txt](https://income.treasuries.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://income.treasuries.io/approaches/transaction-fee-approaches.md).

# Transaction fee approaches

Comparing the different transaction fee approaches that a network could use for their fee model

Every transaction submitted to a Web3 network uses resources to process, verify and store information. Transaction fees commonly help pay the node operators who provide these resources. They also discourage transaction spam and help manage demand when capacity is limited. However, if a network funds its operations through a network coin tax, the purpose of transaction fees changes: rather than generating income, they should primarily protect the network while keeping productive usage as affordable as possible.

## Approaches overview

There are four main transaction fee approaches to consider:

* **Fixed fee** - A set charge for a transaction, regardless of the value of the assets being transferred.
* **Percentage fee** - A charge calculated as a percentage of the value being transferred.
* **Finality fee** - An optional additional charge to prioritise a transaction, potentially reducing the time a user waits for confirmation. This does not necessarily change the network’s underlying finality rules.
* **Demand fee** - A charge that adjusts with network utilisation, increasing when demand approaches or exceeds available capacity and decreasing when capacity is available.

These approaches can be combined. For example, a network could use a low fixed fee under normal conditions and temporarily adjust it during congestion.

## Network coin tax consideration

A network coin tax can provide income for node operators and the network treasury independently of transaction fees. If that income adequately funds network operations, transaction fees no longer need to carry the responsibility of generating revenue.

This separates two objectives: **funding the network** and **protecting its capacity**. The network coin tax addresses the first, while transaction fees can focus on the second.

The goal then becomes keeping fees as low as possible without compromising reliability. More productive usage should be welcomed, rather than treated as an opportunity to extract additional income from users.

## Preventing transaction spam

Funding operations through a network coin tax does not automatically make unrestricted free transactions practical. Transactions still consume finite processing, bandwidth and storage resources. Without a cost or another effective restriction, attackers could flood the network with transactions and prevent legitimate users from accessing it. This is a form of denial-of-service attack.

Transaction fees make this behaviour costly. They do not guarantee protection against every attack, but they provide an important economic deterrent alongside technical safeguards.

A future identity solution could offer another approach. If the network could reliably distinguish unique users and prevent attackers from cheaply creating unlimited identities, it might enforce fair usage limits without charging for every transaction. Such a solution would also need to preserve privacy and accessibility: identity alone would not prevent spam without effective usage controls.

Until a suitable alternative exists, a minimum transaction fee remains a pragmatic safeguard.

## Necessity of fixed fees

A fixed minimum fee ensures that every transaction carries a cost, including transactions that transfer little or no value. This matters because a percentage-only fee would allow attackers to submit large numbers of tiny transactions at almost no expense, despite those transactions consuming network resources.

Within this fee-based approach, a fixed minimum is therefore essential to closing that loophole. Its level should reflect the network’s capacity, resource costs and resistance to spam instead of being a target for transaction-fee revenue. The processing and storage cost of a typical transaction can provide a useful starting point, although cost alone does not determine the appropriate deterrent.

Fixed fees nevertheless impose a greater proportional burden on small transactions, which can particularly affect poorer users. They should therefore be reduced wherever safely possible. Scalable infrastructure helps by allowing the network to support more activity without making ordinary transactions expensive.

## Percentage fees

Percentage fees are fairer in one important respect: users pay the same proportion of the value they transfer. Unlike a fixed fee, the charge does not become proportionally smaller as transaction value increases.

However, transaction value is not the same as a user’s wealth, and percentage fees introduce an undesirable incentive. Users making larger transfers may avoid transacting or move their activity to cheaper networks. This discourages liquidity and usage that the network would otherwise want to attract.

If a network coin tax already funds operations, there is less justification for imposing this additional charge. The objective should be to minimise the cost of productive activity, not maximise revenue from it.

A low, base-cost fixed fee is therefore the more pragmatic starting point: retain enough cost to discourage spam, while avoiding a charge that grows with transaction value. A suitable identity-based alternative could eventually help reduce or remove that minimum fee.

## Finality fee

Finality fees allow users to pay for priority. While this can help allocate scarce capacity, it creates an inherent fairness problem: wealthier users can more easily afford faster service, while poorer users may face longer and less predictable waits.

Paid priority can also encourage competition over transaction ordering, including opportunities for front-running, making the network more complicated to use safely.

Where sharding and horizontal scaling allow capacity to grow with demand, the preferable objective is timely processing for everyone. Sharding distributes work across different parts of the network, while horizontal scaling adds capacity through additional participating infrastructure.

These techniques do not eliminate every bottleneck, but they reduce the justification for making paid priority a permanent feature. Expanding capacity is generally more desirable than requiring users to bid against one another for access.

## Demand fee

Demand fees increase the cost of transacting when the network becomes congested. As a routine response to growing adoption, this is undesirable: successful usage should lead to additional capacity rather than persistently higher prices. Sharded, horizontally scalable networks should aim to accommodate that growth.

However, capacity cannot always expand immediately. Even a scalable network may face a sudden demand spike or spam attack before additional node operators come online.

A temporary, flat increase in the transaction fee could help during these periods. Applied to comparable transactions rather than calculated from their value, it would make sustained spam more expensive and encourage non-urgent activity to wait. If the additional income supports capacity expansion, it could also help incentivise operators to bring resources online sooner.

When utilisation is low, the fee could fall below the normal base-cost benchmark, provided the network coin tax covers the funding gap and adequate spam protection remains. Lower fees would encourage productive use of spare capacity.

The aim should be a utilisation level that supports efficient operation while leaving room for growth and unexpected activity, not continuous operation at maximum capacity. Demand-based adjustments should protect this balance, rather than become another revenue-maximising mechanism. Because higher fees also affect legitimate users, increases should be limited and reversed as capacity becomes available.

## Summary

Where a network coin tax adequately funds operations, a low fixed fee is the most pragmatic starting point. It creates a cost for spam without charging users according to the value they transfer.

The fee should be set as low as safely possible to maximise productive usage. Percentage fees and paid priority are less desirable when revenue is already handled separately and infrastructure can scale with demand.

Demand-based adjustments may still be useful: lowering fees when spare capacity is available and temporarily raising them when sudden congestion threatens reliable access. The guiding principle is to fund the network separately, expand capacity as usage grows, and charge only what is necessary to protect legitimate activity.
