What Does “Medicare for All” Mean?

“Medicare for All” sounds almost self-explanatory.

Medicare exists. Extend it to everyone. Done.

But the closer I have looked at S. 1506, the current Senate bill carrying that name, the less the phrase resembles a description and the more it resembles a compressed file. There is quite a lot inside those three words.

The bill does not simply enroll everyone in Medicare as it presently exists. It establishes a new national health insurance program with broader benefits, different financing arrangements, different provider-payment mechanisms, a prohibition on most duplicative private insurance, and substantial new authority delegated to the Department of Health and Human Services.

That is closer to national single payer than “Medicare, but younger people can join.”

Even the word “all” needs decompression.

S. 1506 says that every “resident of the United States” is entitled to benefits, then leaves HHS to define residency. The Secretary may extend eligibility further to ensure access for every person in the country, while also being required to prevent people from coming to the United States solely to obtain care. Curiously, the temporary Medicare buy-in during the transition explicitly refers to citizens, nationals and lawful permanent residents; the permanent program does not.

Given current arguments over immigration and who belongs in the United States, that does not strike me as a minor drafting detail. It feels more like Chekhov's gun: seen in Act One and inevitably discharged.

Then there is Medicare itself.

The familiar program has premiums, deductibles, coinsurance, Medicare Advantage, Part D drug plans and Medigap. S. 1506 would instead cover a broad set of hospital, outpatient, prescription, mental-health, reproductive, maternity, dental, vision, hearing, rehabilitation, transportation, home-care and other services, generally without patient cost-sharing. Institutional long-term care remains largely with Medicaid.

Private insurance does not disappear entirely, but insurance duplicating the federal benefit package largely does. Supplemental coverage for things outside the program remains legal.

That distinction got me thinking about the Bell System.

I am old enough to vaguely remember AT&T television commercials encouraging people to make long distance calls. “Long distance” was once an ordinary household category. People knew that calling relatives in another state cost money. Rates varied by distance and time of day. Families sometimes waited until evening to call.

The breakup of AT&T did not eliminate telephone service. Government action altered the structure through which telephone service was provided. Technology and subsequent industry changes did the rest. Today I can make a FaceTime call across an ocean or sit in a WebEx meeting with people scattered across the country without thinking for a moment about “long distance.”

The cost of communication did not disappear. The category did. Perhaps that may be one way to understand the ambition behind Medicare for All. If such a system worked well for long enough, questions that seem entirely normal today might eventually sound as peculiar as asking whether it is cheaper to call after 9 p.m.

Is this doctor in network?
Have I met my deductible?
What happens to my insurance if I change jobs?
Can I afford COBRA?
Will insurance cover this?

A future generation might hear “your father’s medical coverage depended on which company employed him” and find the arrangement stranger than we do. But the Bell comparison cuts both ways. Communication costs did not vanish. They migrated into broadband bills, mobile plans, devices, cloud infrastructure, subscriptions and taxes. What became less visible in one place appeared somewhere else.

Healthcare costs would do the same.

S. 1506 eliminates most premiums, deductibles and copayments for covered services, but doctors, nurses, drugs, hospitals and MRI machines do not become free. Their costs move. And this is where the bill becomes unexpectedly incomplete.

It establishes a Medicare for All Trust Fund and redirects existing federal health spending into it, but it does not contain the comprehensive tax structure necessary to finance the permanent system. There is no final payroll-tax rate, income surcharge or other complete revenue package.

That means one of the most important questions—who actually pays more and who pays less—cannot be answered from S. 1506 itself. The legislation is quite specific about the benefits people receive and much less specific about the tax incidence required to sustain them.

This reminds me, perhaps unfairly, of the old fable about the mice proposing to put a bell on the cat. Everyone agrees that knowing where the cat is would be useful. The awkward question is who actually attaches the bell.

There is another sort of complexity hiding inside the promise of simplification.

After Enron and WorldCom, Sarbanes-Oxley attempted to solve real problems in financial controls and corporate accountability. It also helped create an extensive ecosystem of auditors, compliance teams, consultants, control frameworks, evidence repositories and software built around proving that those controls exist and operate correctly.

Institutional solutions tend to create institutions of their own, regardless of the validity of those solutions.

S. 1506 would genuinely eliminate some health insurance administration. Underwriting, much plan marketing, network construction, COBRA administration and ACA exchange operations could disappear. But claims, coding, fraud control, coverage determinations, formularies, appeals, provider credentialing and reporting do not vanish. Some functions move to the federal system, and new ones appear around regional administration and annual hospital global-budget negotiations.

Administrative complexity has a way of regenerating around whichever boundaries a new system creates.

Which brings us to the Aflac duck.

S. 1506 permits supplemental insurance for things the federal system does not cover. In a successful implementation, that might remain a relatively modest market: income replacement, travel expenses, amenities and other peripheral risks.

In a less successful implementation, every gap becomes a potential product.

Long waits? Perhaps somebody sells a travel benefit.
Services outside federal coverage? There is a supplemental policy.
Lost income while waiting for treatment? Another policy.
Navigation through the new bureaucracy? Somebody will sell that too.

Soon that duck has ducklings.

None of that demonstrates that single payer would fail. It demonstrates that markets respond to institutional boundaries. Thirty years later people may again wonder why the supplemental-benefits ecosystem became so complicated.

The transition itself raises a different class of questions.

I work with cloud infrastructure, and we recently migrated virtual machines from a data center into AWS. After we retired a cluster of old SQL Server and application servers, we discovered that customer support staff were still using the legacy system to process payments.

The dependency had not been unknown.
It was in the project requirements.
It had simply been ignored.
The deprecated cluster came back up.

That experience makes the four-year transition in S. 1506 difficult for me to regard as merely a political timetable. By the fourth calendar year after enactment, the permanent program begins while current Medicare, most Medicaid coverage, CHIP, FEHB, the ACA exchanges and duplicative employer and private coverage terminate. Meanwhile HHS must establish enrollment, payment systems, regional offices, provider agreements, a national formulary, reporting systems, appeals mechanisms and thousands of institutional budgets.

The bill requires HHS to protect patients against disruption. What it does not provide is particularly striking: no readiness gates, no missed-deadline contingency, no rollback mechanism and no explicit fallback if some critical part of the new system is not ready.

There is a whiff of Hernán Cortés scuttling the ships here. Making retreat impossible can force an organization to complete a difficult transformation. Sometimes that is exactly what entrenched systems require. But irreversibility is not the same thing as readiness.

A project plan saying “the old data center will be shut down on January 1” is a deadline. “Zero production traffic for ninety days, all service owners signed off, rollback tested and no unresolved critical dependencies” is an exit criterion. The bill has a great deal of the former and surprisingly little of the latter.

None of this answers whether Medicare for All is a good idea. In fact, that may be the wrong first question. The more useful question is what we mean when we say it:

• Does “all” mean citizens, lawful residents, everyone physically here, or whatever a future administration decides “resident” means?
• Does “Medicare” mean the familiar program for seniors, or a substantially new national insurance architecture?
• Does “free at the point of care” mean costs have disappeared, or that we have moved them somewhere less visible?
• Does “simplified administration” mean bureaucracy is eliminated, or that a different bureaucracy takes its place?
• Does a four-year transition describe an achievable migration or merely the date on which Congress has ordered the old servers turned off?

Those are not arguments for or against universal health coverage. They are what emerge when you start unpacking the slogans. “Medicare for All” is three words. S. 1506 is a proposal to reorganize a significant fraction of the American economy.

There is quite a distance between the two.