An Appeal Process Can Work and Still Be Built Not to Be Used

An appeal can reverse a denial and leave the rule that produced it untouched. The insurers that run Medicare Advantage plans have appeals processes that work. From 2014 through 2016, when a patient or a provider appealed a denial, the insurers overturned 75 percent of their own decisions, about 216,000 a year. Patients and providers appealed 1 percent of denials, so fewer than one denial in a hundred was ever reversed that way.1

When the inspector general went back to a sample of prior authorization denials from one week in June 2019, 13 percent were for care that met Medicare’s coverage rules. One common cause was the insurers’ own clinical criteria, which asked for steps Medicare doesn’t, such as an X-ray before an MRI would be approved. The inspector general couldn’t say the criteria broke a rule, because CMS’s guidance wasn’t detailed enough to settle whether insurers could deny care on criteria that went beyond Medicare’s. Every denial in the sample that met Medicare’s rules and was appealed was reversed. Counting the few the insurers caught on their own, reversals came to 3 percent of the denials, against the 13 percent that met the rules.2 Each reversal corrected one decision. The criteria were not a party to any appeal. Limiting them took a federal rule, which CMS issued in April 2023.

An institution can be built to hear every appeal and still be arranged so that hearing them is the most expensive thing it does. It can have the whole machinery of correction: people who act through institutions they depend on,3 an appeal that re-runs one decision along with the rarer mechanism that puts the rule itself on trial,4 and a boundary naming which promises a fallible institution has to give up when they conflict.5 Correction can still fail in an institution that has all of it, and the Medicare Advantage figures show why.

Whether correction happens depends on an institution’s surroundings, its ecology, as much as on its design. An institution lives inside an economy, of funding, labor, competition and career risk, and that economy prices every act of correction. A design can be corrigible on paper and starve in practice.

Appeals are cheap, rule trials are not

The appeal works at the level of the decision. It asks whether one determination was reached correctly. Its protection is also its ceiling, since the determination can be reopened but the criteria can’t. Meanwhile the stamina it demands screens most people out before anyone decides anything.

The rule trial works at the level of the category. It asks whether the criteria are defective, whether the exception count has crossed the line where isolated mistakes have to be read as a pattern. One-off settlements let an institution avoid it. Mercy stands in for correction, the pressure releases, and the rule survives.

The regime question works at the level of the arrangement. It asks whether the institution can be made to yield at all, whether the people it governs hold enough material power to stop the machinery when the machinery is wrong. Most talk of accountability never reaches it, because regime correction happens through a balance of forces, and no channel delivers it.

Each level sets the price of the one below. An insurer’s ecology rewards appeals, which cost little and prove responsiveness, and punishes rule trials, which would cost the loss ratio. So it’ll process appeals until everyone involved retires. The people in charge already know: letters arrive, dashboards flag and caseworkers file the near-misses, and the institution keeps the rule because the harm is the operating margin. An appeal can’t solve a business model.

Diagram

The three-tier correction economy

View source

%% title: The three-tier correction economy
%% caption: Each level sets the operating price of the one below. Cheap individual appeals release pressure while the profitable category rule stays in place.
flowchart TD
  classDef system stroke-width:1.2px;
  classDef gate stroke-width:1.6px;
  classDef binding stroke-width:3px,font-weight:bold;
  classDef failure stroke-width:2.2px,stroke-dasharray:6 3;
  classDef repair stroke-width:2.4px,font-weight:bold;

  L1["Level 1: Decision Appeal<br/>Cheap • Preserves rule & loss ratio • Individual relief"]:::system
  L2["Level 2: Category Trial<br/>Puts criteria on trial • Threatens revenue • Highly resisted"]:::gate
  L3["Level 3: Regime Correction<br/>Balance of material forces • Strike, refusal, political restructuring"]:::binding

  L1 -. "venting pressure protects" .-> L2
  L2 -. "structural survival bounds" .-> L3
  L3 ==> L2
  L2 ==> L1

Who insulation should protect, and from whom

“Insulation” usually protects whoever is saying it, so direction matters more than the word. People need insulation from institutional and market coercion. Non-domination has to be material rather than reputational. A person who can be ruined by a frozen account can’t contest anything, whatever the channels promise, and an escape hatch for one person amounts to class position.6 For people, insulation means decommodified survival, healthcare, housing and standing that don’t evaporate when an institution takes the dispute personally.

Public-purpose institutions need insulation from the actors who punish them for serving those purposes. The regulator is disciplined by the industry it regulates. The hospital that treats the uninsured is punished by its payer mix, and the newsroom that runs the story loses the advertisers. Nobody in any of these cases needs a bribe, because the cost of correcting is known in advance and falls on whoever corrects.

And institutions must not be insulated from the people they govern, who need standing, contestability, exit and the right to put the manual on trial. An institution that holds rights over the people who depend on it owes them a return path.7 An institution insulated from correction becomes a well-fortified hazard.

Diagram

Three directions of structural insulation

View source

%% title: Three directions of structural insulation
%% caption: Insulation protects people from coercion and public missions from market predation, but must never insulate institutions from the people they govern.
flowchart LR
  classDef actor stroke-width:1.6px;
  classDef system stroke-width:1.2px;
  classDef binding stroke-width:3px,font-weight:bold;
  classDef failure stroke-width:2.2px,stroke-dasharray:6 3;
  classDef repair stroke-width:2.4px,font-weight:bold;

  subgraph CorrectInsulation ["Legitimate Insulation"]
    direction TB
    P([Governed People]):::actor -- "insulated from" --> M["Market / Institutional coercion<br/>Decommodified health & housing"]:::repair
    I["Public Institutions"]:::system -- "insulated from" --> C["Capital discipline & capture<br/>Protected from punitive markets"]:::repair
  end

  subgraph IllegitimateInsulation ["Illegitimate Insulation (Hazard)"]
    direction TB
    H["Governing Institution"]:::system -- "insulated FROM" --> G([Governed Citizens]):::actor
    G -. "denied return path" .-> F([Insulated Domination]):::failure
  end

What the ecology is made of

Correction eats staff time, invites scrutiny, delays metrics and creates precedent. An economy that prices it above its alternatives selects for institutions that perform correction and practice avoidance. Hirschman named the two basic responses to a failing institution fifty years ago, exit and voice, and the question was always what makes voice survivable.8

Exit has to exist. You Were Free to Choose took apart the defense that rests on paper options. Exit that costs someone their healthcare, their visa or their custody arrangement is a price tag on voice, and everyone can read it.

Withdrawal has to be possible. A Great Metric Can Mean Someone Is Quietly Fixing It is about the people an institution depends on and doesn’t count. The capacity to stop a shift, a docket or a service is regime-level power in its most ordinary form, and institutions answer complaints from the people who have it.

Survival cannot be held hostage. The Death of Slack named spare capacity as the material precondition of correction. An institution running at 104 percent has channels and no ability to act on what they hear. Slack does for a bureaucracy what decommodification does for a person, giving it the physical basis for saying no.

The economy must not select against care. Stable on Paper described institutions that survive by moving variability onto the people with the least capacity to hold it. The corrigible institution carries costs the ruthless one externalizes. “Just build good institutions” was never enough. In a competitive field the good ones lose to the cheap ones unless the field is arranged so that externalized harm gets priced.

Domination isn’t a matter of size

Universal Care Without One Big Program argued that domination enters through scale, so keeping care local and un-owned would keep the domination out. That mistook one carrier of domination for the thing itself. A ministry can be unreachable, and so can a co-op whose members are subordinate to its mistakes, a family, a clinic board, or a neighborhood association with one landlord on it. Scale changes what correction costs, since it’s cheaper to walk to the district office than to sue a ministry. But a price difference is not a difference in kind. A local arrangement that punishes correction is domination with better walking distance.

Anarchist Jainism drew the line harder, treating authority as unjustified hierarchy. The two positions reconcile one level up, where hierarchy is presumptively dangerous rather than illegitimate by definition. The burden of justification travels with the authority instead of being paid once at the founding. Authority stays valid while the return path works and becomes void when it doesn’t.

Correction can’t be institutionally suicidal

An anarchist might object that standing up institutions sanctifies them. Their standing here is conditional by design, and authority held without a return path is a hazard with paperwork. An institutionalist might fear permanent procedural sabotage. Procedure does carry real weight, the difference between correction and coup, but a return path inside a punishing economy is decoration. The socialist objection is correct that procedural safeguards can’t overpower capital discipline, so the insulation has to be material.

Power must be corrigible, and institutions must be insulated from pressures that systematically punish correction.

Half of that is the liberal inheritance, appeals and rule trials and standing. The other half is the materialist one, decommodification, funded missions and priced harm, and neither works alone. Build powerful institutions whose authority yields to reality. Give the people they govern the material power to make them yield. Arrange the surrounding economy so that doing the right thing isn’t institutionally suicidal.

The Medicare Advantage insurers built a channel that works at the level of the decision, where three appeals in four won. The criteria sat a level up, where no appeal reached them, and with 1 percent of denials appealed, the price of correcting is high enough that almost nobody pays it.