The Reform Agenda · evidence gap · no position taken

Coverage architecture: what we do not yet know

This is where a policy institute is most tempted to substitute conviction for evidence. The first object is therefore a list of what is missing, and what would have to be read before a recommendation would be worth publishing.

EVIDENCE GAPNo position taken

Why this opens as a gap and not as a position

Coverage architecture — single payer, multi-payer with a public option, regulated competition, the varieties in between — is the most frequently argued subject in health policy and the one where this publication currently has the least to stand on. The documentary layer holds payment rules, coverage transitions and programme integrity in detail. It holds almost nothing on the comparative performance of whole coverage architectures, because that evidence is contested, jurisdiction-specific, and mostly not in the primary-source form the record layer requires.

Publishing a recommendation here would mean doing what this site was built not to do: dressing a preference as an analysis. So the first object is a statement of what would have to be true, and what would have to be read, before a position could honestly be taken.

What is actually established in the record here

Transitions are where harm happens

The documentary layer carries a California population being moved from managed care to fee-for-service with a provider-enrolment deadline attached. Whatever one thinks of either architecture, the evidence on the site concerns the transition.

Payment design drives clinical behaviour

Two conversion factors, an efficiency adjustment, and the treatment of same-day services are recorded in detail. This is established and it is architecture-neutral.

Administrative cost is real but contested

What is documented here is the burden of specific mechanisms — authorisation, appeals, credentialing. Aggregate administrative-cost comparisons between systems are not, on this site, primary-source material.

Coverage does not equal access

Network adequacy, transport, language and specialist availability are documented in the record layer, and each survives every change of architecture.

What would have to be read before a position is taken

  1. Actuarial primary sources, not advocacy modelling, on transition cost in a jurisdiction of comparable scale — including who bears the cost in the years before any steady state.
  2. Physician supply response under a change of payer structure, read from workforce data rather than from survey intention.
  3. What happened to the specific mechanisms this site documents — prior authorisation, denial and appeal — in systems that changed architecture. A single payer that retains utilisation management has not removed the mechanism, only its owner.
  4. Access outcomes in a comparable geography: the Central Valley question is not whether coverage exists but whether a specialist is reachable within a day.
  5. The failure record: jurisdictions that attempted a transition and abandoned it, read from their own official assessments rather than from the accounts of either side.

Each of those is a Layer 1 acquisition task. Until they exist in the record, this page stays an evidence gap — and says so in its own label.

The arguments against declining to take a position

The strongest arguments against

  1. Refusing to take a position on the central question of health policy is an evasion, and a physician-led institute that will draft seven sections on prior authorisation but nothing on coverage is choosing the safe subjects.
  2. The evidence will never be complete, so the standard is unmeetable and functions as a permanent excuse.
  3. Others take positions on considerably less, and the argument gets made without us.

Answers

  1. Taking one on this record would forfeit the only thing that makes the other positions worth reading. The prior-authorization standard is drafted because the record carries the rule, three audits and a bill; this page is not, because it does not.
  2. The five acquisitions above are specific and finite. That is precisely the difference between a standard and an excuse, and it is why they are listed by name rather than described as “more research”.
  3. True, and it is an argument about them. The value of this publication is that a reader can tell which of its statements rest on a document.

What can be said now, without a position

Three propositions the current record does support, and which any architecture has to answer:

A payer’s identity does not settle a physician’s independence

The mechanisms that determine clinical latitude — targets, templates, utilisation review, coding configuration — are documented here operating under public and private payers alike. The lever list applies either way.

Continuity is the measurable good

Interrupted therapy, re-authorisation and coverage transition appear repeatedly in this record as the point where patients are harmed. An architecture should be judged on continuity before it is judged on aggregate cost.

Administrative burden should be attributed, not asserted

The honest unit of analysis is a named mechanism with a measurable cost, which is how the prior-authorization standard is drafted.

This page will become a position, or several competing positions, when the five acquisitions above are in the record. It is listed on the agenda as a gap rather than omitted, because an honest gap is a more useful object than a confident essay.