The market that can’t be a market

In the summer of 2023, a 58-year-old electrician in Memphis named Robert Clark was diagnosed with Type 2 diabetes. His doctor prescribed insulin — a drug first synthesized in 1921, whose core patents expired decades ago, and whose production cost per vial has been estimated by researchers at Yale at under $10. Clark paid $300 a month. His insurance, through his employer, had a high deductible he had not yet met. He began rationing: taking half his prescribed dose, checking his blood sugar less often than he should, skipping follow-up appointments to save the copays. His brother, who had a different employer and therefore different insurance, paid $35 a month for the same drug after Congress capped insulin prices for Medicare enrollees in 2022.

Same drug. Same disease. Same city. Different price — by a factor of nearly nine — because of which employer each man happened to work for.

This story is not unusual. It is routine. In 2024, US health spending reached $5.3 trillion — $15,474 per person. Countries that spend far less cover everyone. The United States still does not: Census data for 2024 found that 92.0 percent of people had health insurance for some or all of the year, which also means roughly 27 million did not. The Commonwealth Fund’s Mirror, Mirror 2024 comparison again ranked the US last overall among the ten wealthy nations it studied.

What is striking about these facts is not what they say about any particular policy position. It is what every position in the American healthcare debate does with them. Market advocates cite innovation — American pharmaceutical companies produce a disproportionate share of the world’s breakthrough drugs, and high prices fund that pipeline. Single-payer advocates cite the outcomes data — if we spent what Germany or Canada spends per person and covered everyone, we would get better population health. Private equity defenders cite consolidation efficiency. Critics cite mortality data from PE-owned hospitals. Every position looks at the same system and finds confirmation.

This is not because the evidence is ambiguous — though it sometimes is. It is because every healthcare dispute in America is downstream of a structural contradiction that no individual reform has managed to resolve: medicine is organized as a commodity and understood as infrastructure. The maps Ripple has been building across this cluster, taken together, show why that contradiction is so durable, and what it costs.

The hidden common structure

Mapped individually, the debates in this cluster look like separate policy questions. Universal healthcare is a debate about payment mechanisms — single payer versus multi-payer, Medicare for All versus the Affordable Care Act. Drug pricing is a debate about patents and pharmaceutical industry profitability. Private equity in healthcare is a debate about financial regulation and hospital ownership. Mental health policy is a debate about parity, coverage, and whether community care can replace institutional psychiatry. End-of-life care is a debate about physician-assisted dying, hospice, and how much the final months of life should consume the healthcare budget. Global health governance is a debate about TRIPS waivers and who bears the cost of pandemic preparedness.

But every one of these maps, looked at structurally, is an argument about a single unresolved question: is medical care a commodity to be allocated by willingness and ability to pay, or infrastructure to be allocated by need?

What the cluster now makes harder to ignore is that this question does not stop at the clinic door. It extends into the years when someone is not curable, not fully independent, and not well served by episodic medicine. Once dependency becomes a durable condition rather than a temporary interruption, the American system does something revealing: it stops calling the problem healthcare and starts routing it to daughters, spouses, low-wage aides, Medicaid spend-down rules, and exhausted households.

The question sounds like a values dispute, and it is. But it is also an empirical one — because markets have preconditions, and healthcare doesn’t satisfy them. A functioning market requires voluntary exchange (you can opt out), informed buyers (you can evaluate quality and compare prices), competitive suppliers (alternatives exist and you can switch), and products whose value you can assess before purchase. Medical care fails all four tests in the cases that matter most.

You cannot comparison-shop in an emergency. You cannot evaluate the clinical quality of a surgical procedure as a layperson. You cannot exit the market for insulin if you are a Type 1 diabetic. You cannot accurately assess the value of a treatment before receiving it, because the information asymmetry between patient and physician is, in complex cases, essentially uncloseable. Kenneth Arrow identified this in 1963 — his foundational paper “Uncertainty and the Welfare Economics of Medical Care” argued that healthcare was structurally different from other goods in ways that made standard market analysis inapplicable.

And yet the system is structured as a market — priced by negotiation between insurers and providers, supply-constrained by licensure and hospital consolidation, distributed by ability to pay, with profit extraction at every node. The resulting prices, access gaps, and administrative overhead are not market failures waiting to be corrected. They are the predictable output of applying market mechanisms to something that doesn’t behave like a market.

The dependency problem the cluster keeps rediscovering

Healthcare debate in the United States is usually narrated around acute episodes: getting coverage, filling prescriptions, seeing a specialist, surviving a hospitalization. But the maps in this cluster increasingly point to a different center of gravity: the long middle stretch in which people need help not for a week, but for years. The family-burden question is not a side issue inside elder-care policy. It is one of the main places where the healthcare system reveals what it really assumes about obligation.

KFF’s 2024 overview of long-term services and supports shows the design clearly. The United States spent $415 billion on LTSS in 2022. Medicaid paid 61 percent. Out-of-pocket spending accounted for another 17 percent. Medicare, the program most older Americans assume will cover old-age care, generally does not pay for ongoing long-term support outside limited circumstances. HHS estimates that an average American turning 65 today will incur $120,900 in future LTSS costs, with families paying more than one-third themselves. This is not a bug at the edge of the system. It is one of the system’s governing design choices.

The hidden subsidy that keeps this design politically survivable is unpaid care. AARP’s March 26, 2026 update to Valuing the Invaluable estimates that 59 million family caregivers provided 49.5 billion hours of care in 2024, worth about $1.01 trillion at replacement cost. That labor is doing system maintenance. It keeps hospitals able to discharge patients into homes that are not actually staffed for recovery. It keeps frail elders out of institutions that are too expensive or unavailable. It keeps the long-term-care crisis looking like a private family struggle rather than a first-order healthcare financing problem.

This is why the healthcare cluster and the welfare cluster keep touching. Dependency is both a medical reality and a distributional question. Someone has to absorb the time, income loss, administrative burden, and emotional vigilance that chronic illness, disability, dementia, and dying require. The central fight is not only over whether care should be covered. It is over whether those costs should remain privatized inside households until they become catastrophic enough to qualify for public rescue.

Structural tension one: the commodity/infrastructure contradiction

The United States has never settled the question of what healthcare is for, at the level of institutional design. The result is a system that is simultaneously — and incoherently — both.

The commodity logic is visible in how hospitals are financed and governed, how pharmaceuticals are priced, how insurance is structured around employment, how prior authorization works, how billing operates. Steward Health Care, the largest private for-profit hospital chain in the United States, filed for bankruptcy in May 2024 after its private equity owners — Cerberus Capital Management and then Medical Properties Trust — had extracted hundreds of millions in sale-leaseback transactions and management fees over the preceding decade. Hospitals in eight states closed or faced closure. The private equity in healthcare map documents this pattern: a 2023 study in the Journal of the American Medical Association found that PE acquisition of hospitals was associated with a 25 percent increase in hospital-acquired adverse events — falls, infections, bedsores — and higher charges per patient. The commodity logic was operating exactly as intended. The infrastructure logic was failing exactly as predictably.

But the infrastructure logic is also visible — in the Emergency Medical Treatment and Labor Act of 1986, which requires emergency rooms to stabilize patients regardless of ability to pay; in Medicare and Medicaid, which cover the elderly and the poorest Americans through collective financing; in the political backlash that greets every proposal to cut Medicaid or end ACA protections for pre-existing conditions; in the near-universal moral intuition, shared across the political spectrum, that someone should not die because they cannot afford treatment.

The infrastructure intuition is so strong that even committed market advocates do not argue for pure triage by ability to pay — they argue instead that markets will produce better outcomes for everyone, including those who cannot pay. The argument is consequentialist, not principled. Which means the moral consensus — care should follow need — is broader than the policy consensus. Everyone agrees, in principle, on the infrastructure framing. The institutional structure operates on the commodity framing. That gap is not an accident or an oversight. It is a durable political equilibrium produced by the fact that the commodity structure generates enormous rents for organized interests who are effective at defending them.

The contradiction becomes even sharper once care shifts from cure to maintenance. Acute medical treatment is partially socialized through emergency obligations, Medicare, Medicaid, and public subsidies. Dependency is not. Ongoing help with bathing, feeding, medication management, transportation, supervision, and respite is treated as a family problem until a household runs out of money or stamina. The system socializes rescue but privatizes endurance. That is why long-term care feels simultaneously central and invisible.

The healthcare access map shows this contradiction in its purest form. The four positions — market competition advocates, incremental universalism advocates, single-payer advocates, and community health infrastructure advocates — are not primarily disagreeing about goals. They are disagreeing about whether the commodity structure can be reformed to achieve the infrastructure goal, or whether the institutional structure must be replaced to achieve it.

Structural tension two: the surplus capture problem

The United States healthcare system generates an enormous surplus — the gap between what care actually costs to produce and what is charged for it. That surplus does not disappear. It goes somewhere. Where it goes is the hidden subject of every healthcare policy debate.

The drug pricing and pharmaceutical patents map makes this explicit. The core dispute is between the innovation argument — high prices fund the R&D that produces breakthrough therapies, and caps on prices would reduce investment in future drugs — and the access argument — prices that bear no relationship to production cost exclude the people who need drugs most. Both arguments contain something real. But the innovation argument is doing double duty: it is simultaneously a genuine observation about R&D incentives and a shield for practices that have nothing to do with innovation.

Patent evergreening — filing secondary patents on minor reformulations of existing drugs to extend exclusivity periods beyond their intended term — generates revenue without generating new therapies. Pay-for-delay agreements, in which branded manufacturers pay generic competitors to stay off the market, were estimated by the FTC to cost consumers $3.5 billion annually. Pharmaceutical companies, as a sector, consistently spend more on share buybacks and dividends than on R&D. The surplus generated by pricing power flows substantially to financial returns, not to new drugs. The innovation argument is true of some of the surplus. The rent-capture argument is true of the rest. Because these are bundled together — in the same companies, the same legislative frameworks, the same pricing structures — challenging rent-capture can always be reframed as attacking innovation. This bundling is not incidental. It is the political economy of the pharmaceutical lobby.

The same pattern appears in insurance. A 2019 study published in JAMA found that administrative costs accounted for 34.2 percent of total US healthcare expenditures — roughly $812 billion in 2017 dollars. The comparable figure in Canada’s single-payer system was 12 percent. The difference — more than $350 billion annually in administrative overhead — represents the transaction costs of running a multi-payer system: prior authorization reviews, claims processing, billing coding, utilization management, network negotiations. None of this produces care. It produces infrastructure for the allocation of the right to receive care based on ability to pay. The universal healthcare and single-payer map captures this: the efficiency argument for single-payer is not primarily ideological. The administrative overhead is a measurable cost of the commodity structure.

In the mental health policy map, the surplus capture problem takes a different form. Mental healthcare is structurally underfinanced relative to physical healthcare — insurers pay lower rates for mental health services, providers are more likely to be out of network, wait times for psychiatrists in the US average 25 days nationally and can exceed six months in many states. The Mental Health Parity and Addiction Equity Act of 2008 required equivalent coverage, but enforcement has been weak and the gap persists. The gap is not because mental health care is less effective — psychotherapy and psychiatric medication have strong evidence bases. It is because the return on investment for insurers is less immediate and harder to measure: mental health benefits accrue over time, across payers, as people change jobs and therefore insurers. No insurer captures the full downstream value of treating depression or psychosis early. The surplus from undertreatment flows to lower premiums and higher margins; the cost flows to emergency rooms, to incarceration, to lost productivity, to families.

Long-term care adds a third layer to the same pattern. The system suppresses public spending by relying on unpaid labor and underpaid labor, then treats the resulting shortages as if they were natural facts. Medicaid underwrites much of the sector, but families still face spend-down rules, waiting lists, fragmented home-care markets, and nursing home chains trying to extract margin from publicly financed dependency. The surplus is not captured only through billing complexity or pharmaceutical pricing. It is also captured when the system acquires an implicit trillion-dollar labor subsidy from family caregivers while refusing to count that subsidy as part of healthcare infrastructure.

The maps in sequence

Watching the commodity/infrastructure contradiction and the surplus capture problem appear and reappear across the maps in this cluster shows how the same structural tension generates different-looking surface debates.

End-of-life care is where the infrastructure framing becomes impossible to avoid. Roughly a quarter of Medicare spending occurs in the final year of life. The debate over physician-assisted dying, over hospice versus aggressive intervention, over who decides when treatment becomes futile — these are debates about what medicine is for at the moment when the commodity logic offers the least coherent guidance. You cannot optimize care for dying people by reference to future returns. The question of what we owe someone who is dying is, almost definitionally, not a market question. And yet the institutional structures governing end-of-life care — who gets palliative care, who gets aggressive intervention, who gets referred to hospice and when — are substantially shaped by insurance incentives, hospital revenue models, and liability exposure. The commodity structure operates even where the commodity logic has no normative purchase.

The end-of-life map also clarifies something the cluster now sees more clearly than it did at first: dependency is often what people are actually arguing about when they say they fear loss of dignity or becoming a burden. Hospice and palliative care are not only clinical services. They are supports for households carrying the last, most concentrated phase of care. A system that offers formal choice at the end while underbuilding the caregiving conditions around that choice is not neutral. It is asking families to improvise moral decisions inside an infrastructure deficit.

Global health governance makes the contradiction global. The TRIPS Agreement — the WTO’s intellectual property framework, adopted in 1994 — extended pharmaceutical patent protection globally, giving wealthy-country pharmaceutical companies the legal authority to charge monopoly prices in low- and middle-income countries. The HIV/AIDS crisis of the late 1990s was the first major confrontation: antiretroviral drugs that cost less than $100 to produce were being priced at thousands of dollars per patient per year in sub-Saharan Africa, where the epidemic was worst. South Africa’s government faced legal action from pharmaceutical companies for seeking to produce generics. The Doha Declaration of 2001 established that countries could override patents in public health emergencies — but the political and economic pressure to not use that flexibility has been relentless. The COVID-19 pandemic replicated the argument at global scale: vaccine intellectual property waivers were proposed, delayed, partially adopted too late to matter during the acute phase. The innovation argument and the surplus capture argument appeared again, globally, in the same bundled form.

Care work and elder care sits at the intersection of the healthcare cluster and the welfare cluster. That map turns the hidden premise of the whole cluster into the subject itself. Long-term care — the nursing homes, home health aides, assisted living facilities, adult day programs, and family labor that support people who cannot live independently — is not peripheral to healthcare. It is what healthcare becomes when cure is no longer the main task. Yet financing remains means-tested, fragmented, and reliant on women’s unpaid labor plus a low-wage workforce whose compensation is largely downstream of Medicaid rate-setting. The surplus capture problem appears here too: nursing home chains and home care agencies operate on the commodity model, extracting margin from Medicaid rates that are too low to fund quality care. The welfare cluster synthesis identified this as the visibility asymmetry — the cost of cutting collective provision does not disappear, it relocates to household budgets and unpaid family labor. In elder care, that relocation is happening at scale, and the demographic math makes it unsustainable. By the time the healthcare system names the problem, families have usually been carrying it for years.

What the commodity/infrastructure contradiction makes impossible

The most important thing the commodity/infrastructure contradiction produces is not any specific policy failure. It is reform-resistance at the system level.

Every serious healthcare reform proposal of the last four decades has run into the same pattern: a reform that addresses one part of the system creates incentives that are exploited in another part, generating new problems that require further reform. The ACA expanded coverage and prohibited exclusions for pre-existing conditions — real achievements. But it preserved the multi-payer structure and its administrative costs. It created insurance exchanges that have faced chronic instability. It expanded Medicaid in states that chose to participate and left gaps in states that did not. It generated a cottage industry of insurance products designed to satisfy the letter of minimum essential coverage requirements while providing minimal actual coverage — what critics call “junk insurance.” Each fix created a new surface for the commodity logic to exploit.

The private equity wave in healthcare followed the ACA’s consolidation incentives. The ACA rewarded scale — larger systems could negotiate better rates, invest in electronic health records, manage population health more efficiently. Private equity acquisitions pursued scale efficiently, extracted surplus through sale-leaseback transactions and management fees, and left degraded infrastructure behind. The reform produced the condition for the extraction.

This is not an argument against reform — the alternative, not reforming, allows existing harms to continue. It is an observation about what kind of reform is possible within the commodity/infrastructure contradiction versus what requires resolving it. Incremental reforms that preserve the fundamental structure are systematically exploited by interests whose viability depends on that structure. Structural reforms that would resolve the contradiction — single payer, all-payer rate setting, direct negotiation of pharmaceutical prices on a Medicare-for-All scale — face political opposition proportional to the size of the rents they would eliminate.

Long-term care makes this reform trap especially visible. The United States can expand insurance coverage, cap some drug prices, and regulate hospital billing while still leaving dependency financed through household depletion and Medicaid after spend-down. That means reforms can improve access to treatment without resolving who absorbs the cost of the years after treatment: the daughter who cuts her hours, the spouse who becomes a night-shift aide, the worker who leaves the labor force, the older adult who enters a nursing facility only after exhausting savings. A system built this way can look more generous at the point of service than it actually is over the full course of illness.

The rent value at stake is substantial. A 2020 study by researchers at Yale, Harvard, and other institutions estimated that a single-payer system would save approximately $458 billion annually in administrative costs and $236 billion in pharmaceutical prices — nearly $700 billion per year, most of it currently flowing to insurance companies, pharmaceutical manufacturers, and private equity. The political power of those interests in the US legislative system is proportional to those flows. The campaign finance cluster synthesis is relevant here: the political economy of healthcare reform is inseparable from the political economy of who funds elections.

What the maps reveal that individual debates obscure

Looking at the healthcare maps as a cluster rather than individually reveals three things that no individual map can show.

First, the innovation argument and the access argument are not symmetrically valid. The innovation argument is real but bounded: returns to investment do drive pharmaceutical R&D, and patent protection does create incentives for drug development. But the innovation argument is being deployed to defend practices (evergreening, pay-for-delay, administrative overhead, PE extraction) that do not produce innovation. The burden of proof runs asymmetrically: the innovation argument must demonstrate not just that it applies to some pharmaceutical activity, but that the specific practices under challenge are necessary for the innovation that matters. The evidence on evergreening, pay-for-delay, and administrative overhead does not support that claim. The debate has been structured as if these are symmetrical positions when they are not.

Second, the political durability of the current system is not evidence of its legitimacy or efficiency. It is evidence of the power of the interests that benefit from it. The comparison with peer nations — the same health outcomes at half the cost, with universal coverage — is not seriously contested empirically. The debate about whether the US should converge toward those systems is not primarily a debate about evidence. It is a debate about whether the political economy permits it. Naming that honestly is a precondition for productive disagreement.

Third, the cluster is not only about medical prices and insurance design. It is also about where dependency gets assigned. Elder care, mental health, disability, and end-of-life care all expose the same transfer: when collective systems are thin, the remaining work moves into families and low-paid care labor. That means “market versus obligation” is too narrow unless it also asks obligation of whom. The real distribution question is whether dependency will be treated as a shared civic condition or a private household problem that becomes publicly visible only after families break.

A bridge framing

The positions in this cluster are not symmetric, but they are not irrational. The market position is protecting something real: innovation incentives, the dangers of monopsony in government purchasing, the inefficiencies of bureaucratic allocation, the value of consumer choice where choice is meaningful. The single-payer position is protecting something real: the observable efficiency gains of pooled risk, the moral unacceptability of pricing people out of care, the demonstrated outcomes of peer-nation systems. The private equity position is protecting something real: capital formation, consolidation efficiency, the genuine operational improvements that professional management can produce. The reform position is protecting something real: the evidence that financial extraction consistently degrades care quality.

What the maps reveal is that these positions are arguing past each other because they are responding to different parts of the system. The innovation argument applies to drug discovery; it does not apply to evergreening. The efficiency argument applies to genuine management improvement; it does not apply to sale-leaseback extraction. The single-payer efficiency argument applies to administrative costs; it does not automatically apply to clinical quality. The bridge framing is not “both sides have a point” — it is “these arguments have different domains of validity, and conflating them is the source of the talking-past-each-other quality of American healthcare debate.”

The same clarification is needed for dependency. Family-centered advocates are protecting intimacy, continuity, and distrust of bureaucratic standardization. Care worker advocates are protecting the claim that the system runs on labor it refuses to value. Fiscal skeptics are protecting the warning that long-term-care promises become cruel if they are not durably financed. These are not identical claims, but neither are they random. They are arguments about what kind of infrastructure dependency requires, and about how much of that infrastructure can be left to households before “choice” becomes another name for exhaustion.

Arrow’s 1963 insight matters here. Once you accept that healthcare does not satisfy the preconditions for markets to work efficiently, the question becomes not “market or government” but “what combination of market and non-market mechanisms produces the best outcomes for each part of the healthcare system?” Pharmaceutical innovation may genuinely require market incentives. Administrative functions may genuinely benefit from consolidation. Clinical care delivery may genuinely require infrastructure logic to work at all. These are separable questions. Treating them as one question — as the “market vs. government” framing does — is the source of most of the heat and very little of the light in this debate.

References and further reading

• Kenneth Arrow, “Uncertainty and the Welfare Economics of Medical Care” (1963). The foundational paper on why healthcare is structurally different from other markets.
• Centers for Medicare & Medicaid Services, National Health Expenditure Fact Sheet (published December 2025 / accessed April 2026). Current official topline on US health spending: $5.3 trillion and $15,474 per person in 2024.
• US Census Bureau, Health Insurance Coverage in the United States: 2024 (September 9, 2025). Current official coverage figures showing 92.0 percent insured for some or all of 2024.
• Commonwealth Fund, Mirror, Mirror 2024 (September 19, 2024). Comparative ranking placing the United States last overall among the ten wealthy nations studied.
• David Himmelstein et al., “Health Care Administrative Costs in the United States and Canada”, JAMA (2020). Documents the administrative-overhead gap between US multi-payer and Canadian single-payer systems.
• Erin Fuse Brown, “Private Equity and Patient Safety”, JAMA (2023). Documents the 25% increase in adverse events following PE hospital acquisitions.
• Robin Feldman, “May Your Drug Price Be Evergreen”, Harvard Law Review (2018). Analysis of evergreening practices and their effect on drug pricing.
• Galvani et al., “Improving the prognosis of health care in the USA”, The Lancet (2020). The Yale/Harvard study estimating $458 billion in administrative savings and $236 billion in drug pricing savings under single payer.
• KFF, 10 Things About Long-Term Services and Supports (LTSS) (July 8, 2024). Clear current overview of LTSS financing, including Medicaid’s dominant role and the limits of Medicare coverage.
• HHS Office of the Assistant Secretary for Planning and Evaluation, Long-Term Services and Supports for Older Americans: Risks and Financing Research Brief (December 2024). Source for the estimate that an average American turning 65 today will incur $120,900 in future LTSS costs and pay 37 percent out of pocket.
• AARP Public Policy Institute, Valuing the Invaluable 2026 (March 26, 2026). Current estimate of unpaid family caregiving: 59 million caregivers, 49.5 billion hours, and $1.01 trillion in annual economic value.
• Centers for Medicare & Medicaid Services, Ensuring Access to Medicaid Services Final Rule (April 22, 2024). Includes the HCBS payment-adequacy provision requiring most states to direct at least 80 percent of certain Medicaid HCBS payments to worker compensation.
• CMS, Hospice Monitoring Report (2025). Useful current overview of hospice utilization, payments, and the growing scale of end-of-life care under Medicare.
• Mariana Mazzucato, The Value of Everything (2018). On how rent-seeking is distinguished from genuine value creation in healthcare and pharmaceutical industries.
• Elisabeth Rosenthal, An American Sickness: How Healthcare Became Big Business and How You Can Take It Back (2017). Accessible narrative account of how the commodity logic operates across sectors of American healthcare.