Series 2 — Prescription · Paper 2
Dispute Resolution: A System That Actually Works
Every payment system that handles real money between real people will eventually face a dispute. A charge that wasn’t recognised. A product that didn’t arrive. A service that fell short of what was promised. These are not edge cases — they are a normal and inevitable feature of commercial life. The question is not whether disputes will occur. It is whether the system handling them is designed to resolve them fairly, efficiently, and in a way that preserves the trust of everyone involved.
The chargeback mechanism — the card networks’ answer to this question — asks one narrow question: should the money go back to the customer? It routes disputes through the customer’s bank rather than back to the merchant who might resolve them in minutes. It has no incentive structure that encourages early resolution. And it produces outcomes that satisfy neither party while generating cost for both. It was designed as a financial reversal instrument, not a dispute resolution process — and fifty years of evidence confirm it is the wrong instrument for the job.
Building a genuine alternative to the card networks requires building a genuine alternative to the chargeback. Not a better version of the same mechanism. A different mechanism, designed from first principles, for the job it actually needs to do.
This paper describes that mechanism in detail — including the parts that are uncomfortable to specify, such as what happens when parties behave unreasonably, and how the system discourages abuse without compromising the trust of the payers whose participation makes everything else possible.
The design principles
Before describing the mechanism, it is worth being explicit about the principles that shaped it. These are not aspirations — they are constraints that every design decision was tested against.
Payer trust is the foundation. Without it, nothing else works. A payer who fears that raising a dispute will be costly, complicated, or held against them will not raise disputes — and will lose confidence in the scheme. Every element of the dispute process must be accessible, transparent, and safe for payers to use.
Resolution should happen at the lowest possible level. Most disputes are misunderstandings. A charge that wasn’t recognised. A delivery that was later than expected. A service that fell short of expectations but not of the merchant’s legal obligations. These disputes do not need adjudication. They need a conversation. The mechanism should facilitate that conversation before it does anything else.
Incentives should point toward early resolution. The mechanism cannot simply make resolution available — it must make escalation genuinely costly for the party who was wrong to escalate. Without that, the escalation path will be used by parties who have no reasonable case but nothing to lose by trying.
Penalties must be asymmetric. Merchants and payers have different relationships with the scheme, different commercial positions, and different levels of sophistication. A penalty structure that treats them identically would be neither fair nor effective. Financial penalties apply to merchants. Reputational consequences apply to payers. Both are real. Neither is disproportionate.
The neutral arbiter must be structurally neutral. An arbiter with a financial stake in the outcome cannot be trusted to adjudicate fairly. The Scheme Authority — whose revenue comes from membership fees rather than transaction flow — is the only party in the scheme with no financial interest in how any individual dispute is resolved. Its neutrality is structural, not merely stated.
The three stages
The dispute mechanism has three stages. Each stage exists to resolve disputes that the previous stage could not. Movement between stages is governed by defined rules, defined timeframes, and defined consequences. There is no ambiguity about what triggers escalation, who makes the decision at each stage, or what the consequences of that decision are.
Stage 1 — Proposal and response
Only the payer can initiate a dispute. This is not a procedural detail — it is the structural guarantee that the entire mechanism is built around. A merchant cannot open a dispute against a payer. The scheme cannot open a dispute on behalf of either party. The right to raise a dispute belongs exclusively to the payer, is always free to exercise, and cannot be restricted, delayed, or penalised at the point of initiation under any circumstances.
Stage 1 is a strictly sequential two-move process. The payer moves first by raising the dispute. The merchant moves second by responding with a proposed resolution. Only once the merchant has made their position explicit can the payer decide what to do next. This sequencing is enforced by the scheme — the payer cannot escalate to Stage 2 until the merchant has responded.
Move 1 — The merchant proposes a resolution
When a payer raises a dispute, the merchant is notified and a response window opens. Within that window, the merchant must submit a proposed resolution. This can be a full refund, a partial refund with a stated reason, or a rejection with supporting evidence. What it cannot be is silence.
If the merchant does not respond within their window, the dispute is automatically resolved in the payer’s favour — a full refund is issued and a financial penalty applied to the merchant. Non-engagement is not a neutral act. It is treated as a concession, with the consequences of the wrong decision applied immediately. No PISP adjudication is required and no SA involvement is needed.
Move 2 — The payer accepts or escalates
Once the merchant submits their proposal, the payer’s response window opens. The payer has two options: accept the proposal, in which case the dispute closes on the agreed terms with no penalty to either party; or escalate to Stage 2, in which case the merchant’s proposal becomes part of the evidence record that the PISP uses to adjudicate.
There is no reject-and-return option. The outcome at Stage 1 is binary. This is deliberate — a negotiation loop would create delay without accountability and would give bad-faith merchants a mechanism to stall. The merchant has one opportunity to make their position explicit. The payer has one opportunity to accept or reject it.
If the payer does not respond within their window, the dispute is automatically dismissed with no penalty to either party. A payer who raises a dispute and then disengages does not leave the merchant in perpetual limbo.
The expected outcome at Stage 1 is acceptance. A merchant who has behaved honestly and offers a fair resolution has every commercial incentive to do so — because escalation to Stage 2 carries financial risk if the PISP finds against them. A payer who receives a fair proposal has every incentive to accept it — because escalation to Stage 2 adds a reputational note if the PISP finds against them. Stage 1 is designed to make the honest outcome the rational one for both parties.
Stage 2 — PISP adjudication
If direct resolution fails, the dispute escalates to Stage 2. The merchant’s PISP adjudicates, using the transaction record, the scheme rules, and the submissions of both parties — including the merchant’s Stage 1 proposal, which travels as evidence and is central to the adjudication.
The PISP can reach exactly one of three verdicts. There is no fourth option.
Reject — the merchant’s position was correct. The payer’s claim is not upheld. No refund is issued.
Uphold partial — the merchant’s Stage 1 partial refund offer was the correct amount. The refund is issued at exactly the amount the merchant proposed — not more, not less. This verdict is only available when the merchant’s Stage 1 proposal was a partial refund. If the merchant rejected the claim entirely at Stage 1, this verdict is not available.
Uphold full — the merchant was wrong. A full refund of the original transaction amount is issued.
The constraint on partial amounts is deliberate and important. The PISP adjudicator cannot award an arbitrary refund amount of their own determination. The only partial amount available is the one the merchant themselves proposed at Stage 1. This is not a limitation on the adjudicator’s judgment — it is a design choice that removes subjective calculation from the process and ensures that two different adjudicators reviewing identical cases will reach identical outcomes. Consistency is not a courtesy in dispute resolution. It is what makes the process trustworthy.
The merchant’s PISP is the adjudicator at this stage for a specific reason: it holds the transaction record, it has the commercial relationship with the merchant, and it has agreed, as a condition of scheme membership, to adjudicate disputes fairly and in accordance with scheme rules. It is not a neutral party in the philosophical sense — it has a relationship with the merchant — but it is a party with strong incentives to adjudicate correctly. A PISP that consistently rules in favour of its merchants regardless of merit will accumulate a pattern of Stage 3 escalations and SA scrutiny. That is a powerful incentive to get Stage 2 right.
Both parties have a defined window to submit their case. The PISP has a defined window to reach a verdict. Neither party can extend these windows unilaterally.
Stage 3 — Scheme Authority arbitration
If either party disputes the Stage 2 decision, they can escalate to Stage 3 — SA arbitration. The SA reviews the case independently, considering the transaction record, the Stage 2 decision, and the submissions of both parties. Its decision is final and binding on all parties, including the PISP.
The SA’s authority to make binding decisions is not a courtesy. It is a condition of scheme membership. Every PISP, by joining the scheme, has agreed to enforce SA decisions against its own customers — including decisions that go against them. The mechanism for this enforcement is the scheme’s message signing infrastructure: every instruction is attributable, every decision is recorded, and defiance of an SA ruling is visible and consequential. A PISP that fails to enforce an SA decision faces scheme suspension. That is an existential consequence. It does not happen.
Stage 3 is designed to be rare. Its existence is the guarantee that Stages 1 and 2 have teeth — parties at those stages know that if they behave unreasonably, the SA will eventually see it and rule accordingly. But the SA’s caseload should be small, because the incentive structure at Stages 1 and 2 resolves the overwhelming majority of disputes before they reach it.
The penalty matrix
The following table sets out every permutation of escalation and outcome, and the consequence for each party. The principle throughout is consistent: you are only penalised if you made the wrong decision and that decision was confirmed wrong by the next stage.
Stage 1 outcomes
| Scenario | Merchant consequence | Payer consequence |
|---|---|---|
| Merchant doesn’t respond within window | Dispute automatically resolved in payer’s favour — full refund issued, financial penalty applied | None |
| Merchant proposes, payer accepts | Dispute closed on agreed terms | None |
| Merchant proposes, payer escalates to Stage 2 | None yet — proposal carried forward as evidence | None yet |
| Merchant proposes, payer doesn’t respond within window | Dispute automatically dismissed | None |
Stage 2 outcomes
The payer escalates after rejecting the merchant’s Stage 1 proposal
| Stage 2 verdict | Merchant consequence | Payer consequence |
|---|---|---|
| Reject — merchant was right | None | Reputational note recorded |
| Uphold partial — merchant’s Stage 1 offer confirmed as correct amount | None — offer was fair | Reputational note — unnecessary escalation |
| Uphold full — full refund awarded | Financial penalty (Level 1) + full refund | None |
Stage 3 outcomes — payer escalates Stage 2 rejection or partial uphold
Payer disagrees with Stage 2 verdict and escalates to SA
| Stage 3 outcome | Merchant consequence | Payer consequence |
|---|---|---|
| SA upholds — full refund | Financial penalty (Level 2) | Stage 2 reputational note expunged |
| SA upholds partial — same amount as Stage 2 partial | None | Stage 2 reputational note stands |
| SA rejects — merchant wins | None | Significant reputational note recorded |
Stage 3 outcomes — merchant escalates Stage 2 full uphold
Merchant disagrees with Stage 2 full uphold verdict and escalates to SA
| Stage 3 outcome | Merchant consequence | Payer consequence |
|---|---|---|
| SA overturns — merchant wins | None | Reputational note recorded |
| SA upholds — full refund stands | Financial penalty (Level 2) | None |
The financial penalties at Level 1 and Level 2 are set by the Scheme Authority and reviewed periodically. They are calibrated to be meaningful relative to typical transaction values in the scheme — sufficient to make frivolous escalation commercially irrational, without being punitive enough to deter legitimate escalation by smaller merchants.
The reputational consequence for payers
The asymmetry between merchant and payer consequences is deliberate and important. A financial penalty applied to a consumer who loses a dispute would erode the trust that makes the scheme viable. A reputational consequence that affects only how the scheme treats future disputes from that payer is proportionate, transparent, and entirely consistent with the principle that raising a dispute is always free and frictionless.
The reputational system works as follows:
A payer’s dispute record is visible only to themselves and to their PISP. It is not visible to merchants, not shared across the scheme, and not used for any purpose other than dispute management. A payer can view their own record at any time.
The record tracks lost escalations — disputes the payer escalated and lost — within a rolling twelve-month window. Records expire after twelve months. A payer who had a poor dispute record two years ago faces no consequence today.
When a payer has accumulated lost escalations within the window, the consequence is operational rather than financial: additional information is requested before a dispute proceeds at Stage 1. Not a barrier — the dispute still proceeds — but a signal that the system has noted a pattern. This information requirement scales with the number of lost escalations, and resets when the record clears.
In cases of persistent and clearly bad faith behaviour — a pattern that the PISP identifies through the record — the PISP may review its relationship with the payer. This is a commercial decision made by the PISP, not a scheme-level penalty. The scheme provides the signal. The PISP makes the judgment.
Why the incentives work
The penalty matrix is designed so that at every decision point, the rational choice is to resolve the dispute honestly and at the lowest possible stage.
For the merchant: engaging honestly at Stage 1 costs nothing. A merchant whose partial offer is confirmed correct at Stage 2 pays no penalty — their offer was fair and the payer was wrong to escalate. A merchant whose claim is fully upheld at Stage 2 pays a Level 1 financial penalty. Escalating to Stage 3 and losing costs a Level 2 penalty — significantly higher. The rational choice at every point is to offer what is genuinely fair at Stage 1 and accept the outcome if it reaches Stage 2.
For the payer: raising a dispute costs nothing. Winning at any stage costs nothing. Escalating to Stage 3 and losing adds a significant mark to a record that is visible to their PISP and affects how future disputes are processed. A payer who raises legitimate disputes and wins faces no consequence at any stage, ever. A payer who repeatedly escalates disputes they then lose accumulates a record that reflects that pattern accurately.
For the PISP: adjudicating fairly at Stage 2 protects its scheme standing. A PISP whose Stage 2 decisions are consistently overturned at Stage 3 faces SA scrutiny — a pattern of poor adjudication is visible in the scheme’s dispute data. That scrutiny is a powerful incentive to get Stage 2 right, which in turn creates pressure on merchants to resolve disputes honestly at Stage 1.
The incentives cascade correctly at every level. No party benefits from escalation unless they were right to escalate. Every party bears the consequence of escalating when they were wrong.
What this means for trust
Trust in a payment system is built over time and lost in moments. The dispute mechanism described here is designed to build it steadily and protect it carefully.
For payers: raising a dispute is always free, always frictionless, and never penalised if the dispute is legitimate. The payer who has a genuine grievance faces no barrier and no risk. The mechanism is unambiguously on their side at the point of raising a dispute — and only holds them accountable if they subsequently escalate a case they then lose.
For merchants: the mechanism is demanding but fair. Engage honestly, resolve quickly, and the cost is zero. Stonewall, escalate frivolously, or refuse to accept a legitimate outcome, and the costs accumulate quickly. The mechanism is not designed to be easy for merchants. It is designed to make honest behaviour the obvious commercial choice.
For the scheme: a dispute mechanism that resolves disputes fairly and visibly is the foundation of the legitimacy that makes the scheme worth joining. A scheme where disputes disappear into a black box — or where outcomes are determined by the relative power of the parties rather than the merits of the case — is a scheme that merchants and payers will eventually abandon. Transparency and fairness are not values layered on top of the commercial model. They are the commercial model.
The question this answers
The most common objection to any new payment scheme is some version of: “cards have chargeback protection — what do you have instead?”
The answer is not “we have something similar.” The answer is: we have something better, designed for the job that chargebacks were always being asked to do but never equipped to handle.
A staged process that starts with direct resolution and only escalates when necessary. An incentive structure that makes honest, early resolution the rational choice for every party at every stage. A neutral arbiter whose structural position — funded by membership fees, with no stake in transaction outcomes — guarantees independence. Consequences that are asymmetric by design, protecting payer trust while holding merchants accountable. And enforcement that is built into the scheme’s infrastructure, not dependent on goodwill.
This is not a theoretical framework. It is a designed system with specific rules, specific consequences, and specific guarantees. It is also, unlike the chargeback mechanism it replaces, a system that can be inspected, challenged, and held to account — because the rules are published, the decisions are recorded, and the consequences are defined in advance rather than determined by the most powerful party in the room.
Thomas Larsen is a cloud platform architect and engineering leader with twenty years’ experience building open infrastructure for public-sector and defence organisations. He is currently working on an open payment scheme for the UK market.