The cost of the appeal

There is a comfortable fiction embedded in the phrase you have the right to appeal. The fiction is that a right, once granted in words, is thereby made available in fact — that naming a remedy is the same as delivering it. It is not. Between the offer of an appeal and the use of one lies a distance measured in hours, forms, expertise, patience, and often money, and that distance is not fixed by nature. Somebody chose how far it would be. A contest mechanism is a designed object, and like every designed object it distributes effort. The question that decides whether a right to appeal is real is not whether the process exists. It is who was made to pay to walk through it.

Contestability is not free. This is the plain fact that most accounts of accountability skip past, because it is inconvenient to a tidy story in which fairness is a matter of intentions. Every mechanism for contesting a decision imposes costs — the time to assemble evidence, the effort to understand an unfamiliar process, the expertise to frame a grievance in the terms the process will accept, the money for whatever the process requires, and the quieter tax of the emotional toll of pressing a claim against an institution that would prefer you stopped. These costs are real whether or not anyone books them. And because they are real, the choice of where to place them is not administrative housekeeping. It is a substantive allocation of power, made once, at the design stage, and then enforced silently on everyone who arrives afterward.

Who pays to be heard

Consider the two sides of any appeal and the asymmetry between what it costs each of them. On one side stands the affected individual, who must reconstruct what happened, locate the relevant facts, learn a procedure built by people who do this daily while she does it once, wait through whatever delays the process contains, and absorb the risk that all of it comes to nothing. On the other side stands the institution, for whom maintaining the original denial costs almost nothing. The decision is already made. The default is already in its favor. Every day that passes without a successful contest is a day the institution keeps what the decision gave it, at no marginal expense. To deny is to do nothing; to contest a denial is to do a great deal.

That asymmetry is the whole game. When the burden of contestability falls on the party with the least capacity to bear it, and the burden of maintaining the status quo falls on the party with the most, the process can be scrupulously fair in its rules and still be rigged in its economics. Nobody has to lie. Nobody has to deny a single valid appeal. The design simply ensures that the cost of raising one exceeds what most people can spend — and the sorting happens before any appeal is ever heard, in the private arithmetic of people deciding it is not worth it. An institution can build a genuinely impartial tribunal and place it at the top of a hill it knows most claimants cannot climb, and then point, accurately, to the tribunal.

Rational abandonment

What follows from this asymmetry is not a wave of denied appeals. It is something quieter and harder to see: appeals that are never brought. A person weighing whether to contest a decision does a rough sum, often without naming it as one. On one side, the expected value of winning — the remedy, discounted by the odds of getting it. On the other, the cost of trying — the hours, the learning, the waiting, the strain. When the second exceeds the first, the rational move is to walk away, and most people, being rational, do. The grievance was valid. The person simply could not afford to prove it. This is not resignation and it is not apathy. It is arithmetic, and the institution set the numbers.

Call it rational abandonment, and notice that it is engineered by cost structure rather than by any explicit refusal. A system that produces it can claim, without contradiction, that it offers appeals to everyone and denies them to no one. Both statements are true. What the system does instead is make the exercise of the right cost more than the right is worth to the person holding it, so that the right is offered in form and withdrawn in economics. The appeals process becomes a monument — impressive, well-documented, technically open, and used by almost no one it was ostensibly built for. The people who do use it are, predictably, the ones who can absorb the cost: the well-resourced, the represented, the institutionally fluent. For everyone else the right exists the way a door exists in a wall thirty feet up.

A right you cannot afford to exercise is not a right the institution granted you; it is one it can point to while you cannot reach it.

This is why the economics of the appeal, and not its existence, is the honest measure of contestability. The access-to-justice literature has understood for a long time that a remedy priced beyond reach is functionally no remedy, and that the cost of contesting is itself a mechanism of exclusion — one that operates most efficiently precisely because it never has to say no. It connects directly to the prior questions this family keeps returning to: who pays for the account nobody keeps, and what it means when the appeal simply goes in a circle. An appeal that returns the same answer wastes the claimant's cost for nothing. An appeal too expensive to bring never spends it at all. Both leave the decision standing; they differ only in whether the person was made to pay before losing or gave up before paying.

Making the right affordable

A right to contest is only as real as it is affordable to exercise. That is the whole of the argument, and it points at a corrective that is unglamorous and concrete: if contestability is priced too high for most people to use, the way to make it real is to lower its price to the affected party. Not to abolish the cost — to move it, and to shrink the part of it that falls on the person least able to carry it. The single most expensive part of most appeals is the first one: reconstructing what happened. Before you can argue that a decision was wrong, you must establish what the decision was, what it rested on, and what rules governed it — and that reconstruction is hardest for exactly the party who was not in the room and does not hold the records.

This is where a ready-made decision record changes the arithmetic. A Decision Receipt that carries the evidence actually consulted, the rules that were active, and enough state to see how the decision followed from its inputs removes the reconstruction cost from the shoulders of the person least equipped to bear it. The claimant no longer has to prove what happened before she can begin to argue about it; she starts with the account already in hand and spends her limited resources on the argument that matters. That is not a favor to the aggrieved. It is a re-pricing of the appeal — a deliberate shift of its most punishing cost off the weakest party and onto the party that made the decision and is best placed to document it.

There is an honest limit here, and it belongs in the argument rather than outside it. Appeals cannot be made costless, and they should not be. A contest that costs nothing to bring invites contests brought for their own sake, and a system swamped by frivolous appeals serves no one, least of all the next person with a real grievance waiting behind them. The aim is not zero cost. It is proportionate cost — enough friction to deter the frivolous, not so much that it excludes the genuine — and, above all, cost borne fairly rather than dumped by default on whoever happens to be weakest. A well-designed appeal asks the claimant to make her case. A badly designed one asks her to first rebuild, at her own expense, the record the institution already holds. The difference between them is not whether contesting costs something. It is who was made to pay, and whether the price was set to be met or set to be refused.

— Dispatches · Summit Cognitive