The Power to Correct the Rules Is Handed Out Too

Two health systems offer identical coverage under the same law, with the same entitlements, forms and audit standards. In the first, paid staff review unusual cases and have a discretionary budget for coordination. A record the patient shows to be wrong gets corrected without a fight. Appeals resolve in weeks. Patterns of failure reach someone with the authority to change the rule that produced them. In the second, the same unusual case produces twenty phone calls the patient places herself and specialists she schedules herself. Her records stay wrong until she corrects them at every visit, her appeals resolve by expiring, and two unusually conscientious employees make up the difference off the books. On paper, the benefits match the first system’s exactly.

Everything the coverage requires of the people receiving it is different.1

Something close to this comparison ran across the states when Medicaid restarted annual renewals in 2023. Eligibility differs from state to state, but every state was under the same federal rule to try renewing people from information it already held before sending them a form. By the end of the unwinding, Arizona, North Carolina and Rhode Island had renewed 90 percent or more of the people who kept coverage that way. Pennsylvania and Texas had done so for 11 percent or less, and renewed the rest through forms the enrollees had to return.2

The distributive triad

First-order distributive politics asks who gets what. It’s a real question, though not the whole one, because a benefit isn’t an abstraction. Receiving it takes time, translation, documentation, travel, composure, and somebody’s afternoon. Using it takes appeal rights that survive contact with a process. And when it breaks, someone decides whether the breakdown counts as normal.

So an institution distributes three things: its resources, the adaptation burdens its categories create, and corrective authority. Adaptation burden means who has to reconcile the rule with the lives it does not fit. Corrective authority means who gets to change the rule when the reconciling keeps failing. Two institutions identical on the first distribution can be opposites on the other two, and no comparison of the first distribution will show the difference.

Diagram

Same coverage: who makes the calls, who can change the rule

View source

%% title: Same coverage: who makes the calls, who can change the rule
%% caption: Both systems give the patient identical coverage under the same law. In the first, paid staff take on her unusual case and repeated failures reach someone who can change the rule; in the second, she places twenty calls herself and her appeals expire with the rule untouched.
flowchart TD
  classDef system stroke-width:1.2px;
  classDef repair stroke-width:2.4px,font-weight:bold;
  classDef failure stroke-width:2.2px,stroke-dasharray:6 3;
  classDef binding stroke-width:3px,font-weight:bold;

  subgraph SystemA ["System A: staff absorb the unusual case"]
    direction TB
    A1["1. What she gets<br/>Coverage set by law"]:::system
    A2["2. Who does the work<br/>Paid staff review her case<br/>on a coordination budget"]:::repair
    A3["3. Who can fix the rule<br/>Appeal resolves in weeks;<br/>repeated failure<br/>changes the rule"]:::binding
    A1 --> A2 --> A3
  end

  subgraph SystemB ["System B: the patient absorbs it"]
    direction TB
    B1["1. What she gets<br/>Identical coverage, same law"]:::system
    B2["2. Who does the work<br/>She places 20 calls,<br/>books each specialist"]:::failure
    B3["3. Who can fix the rule<br/>Her appeal expires after<br/>six months; rule unchanged"]:::failure
    B1 --> B2 --> B3
  end
  SystemA ~~~ SystemB

Domination, specified

When does an uneven second distribution become domination rather than mismanagement? Two variables settle the question: how much of the institution’s mismatch a person has to absorb, and how much power they hold over the condition producing it. Load the first and empty the second and you get a configuration that turns up from every direction. It shows up in the nurse who improvises around the procedure she can’t change, and in the claimant who has to prove eligibility repeatedly to a database she can’t correct. The disabled employee lives in it too, negotiating daily with an environment designed without her and unreachable through any channel she has been given.

Republican theories of domination locate unfreedom in being subject to uncontrolled power. Such power doesn’t have to be used harshly, because the injury is having your circumstances set by someone else’s will at all. The point goes further than that. Uncontrolled power includes the power to stay simple while making someone else become flexible. Its holder can create costs through its own categories, schedules and errors, and leave whoever can’t refuse to soak them up. The rigidity works as a command, and the burden placed on others is its enforcement.

Redistribution leaves the other two in place

Serious politics usually stops here. Redistribution rewrites the first distribution, meaning who owns, who funds and who is entitled, but doesn’t on its own rewrite the other two. A welfare state can be run on the old terms of the second and third distributions. So can a cooperative, a union-run firm, a nationalized industry, a public platform. Collective provision that leaves the burdens and the authority where they were has changed who owns the machine and kept who absorbs its output.

Private provision needs no such accident. The calls the patient places herself are costs the insurer didn’t carry. Revising the rule would cost the loss ratio, so leaving the second distribution alone pays the insurer. Ownership matters to the second distribution, then, because it decides who profits from the burden. It does not by itself settle the third, because owning the machine is not standing to revise its rule, and neither half works alone.

The criterion is demanding on purpose. Intent does not make an institution just, and neither does ownership. Its justice depends on the distribution it actually enforces, of benefits, of burdens, and of the standing to end both. A project that expands collective institutions while leaving the second and third distributions alone will reproduce the experience it mobilized against in whatever it builds. Its institutions will keep the rule out of reach, leave the burden with the people below it, and call the arrangement service.3

Hold the coverage fixed and read the rest

The two-systems comparison is also a test, and you can run anything through it. Take a proposal, public or private, centralized or local, professional or automated. Hold the entitlement constant and read the other two distributions: who coordinates, who corrects, who waits, who pays for the waiting, and who gets to redesign the rule when the pattern repeats. A change that improves the first distribution while worsening the second is a transfer dressed as a reform, and the people transferred usually notice before the evaluators do.

Who should own the hospital is a further question, and this one bears on the answer. An owner paid by the burden it leaves unaccounted has to lose that income before any procedure will hold. What the hospital owes whoever it treats is settled either way. It owes them the burden accounted for, the correction within reach, and, written into the design, the authority to stop being the remainder. Still missing is a route by which the people absorbing an institution’s errors can force a review of the rule that produced them.