← All explorations

How is the US healthcare system structured, why does it cost twice as much as peers, and what would it take to fix

Why Does American Healthcare Cost So Much, and What Would Actually Fix It?

| 120 nodes · 472 edges
↓ .md ↓ .db Take this into your AI — the full analysis + graph as markdown, ready to paste into ChatGPT, Claude, Gemini or any AI.

Based on analysis of a 120-node, 472-edge knowledge graph mapping the structural causes of US healthcare costs and the barriers to reform.

The Short Answer

The US spends about twice as much per person on healthcare as comparable wealthy countries. The graph does not show this is because Americans are sicker, or go to the doctor more often, or receive dramatically better care. It shows the gap is almost entirely explained by prices — what hospitals, drug companies, and specialists charge. Understanding why those prices are so high, and why they have stayed high for decades, requires understanding a system where the people who benefit from high prices also have the power to prevent the changes that would lower them.

Think of It Like a Town With One Grocery Store

Imagine a small town where there is only one grocery store. The owner charges twice what stores in neighboring towns charge. Residents complain, but there are reasons they can’t easily shop elsewhere — the roads to other towns are legally restricted, competing stores aren’t allowed to open nearby, and the town council (which sets those rules) receives most of its campaign donations from the grocery store owner.

This is not a perfect analogy for US healthcare, but it captures the core structure the graph describes: a system where high prices persist because the actors who set prices also have significant influence over the rules that could constrain them.

The Main Problem: Too Many Middlemen, No Shared Price List

Most wealthy countries have either a single government health insurance program (like Canada) or a tightly regulated set of insurance funds that all pay the same prices for the same services (like Germany). When everyone pays the same price, the price can be negotiated down.

The United States has neither of these. Instead, it has hundreds of different insurance companies, employer health plans, government programs, and direct-pay arrangements, each negotiating separately with hospitals and doctors. The graph calls this “multi-payer fragmentation,” and it treats it as the foundational architectural feature from which most other cost problems emerge.

Here is why fragmentation is such a multiplier. When a hospital bills many different payers at many different rates, it needs a large billing department to manage all of those relationships. When a doctor’s office accepts twenty different insurance plans, it needs staff dedicated entirely to paperwork. Studies have estimated that roughly a quarter of US healthcare spending goes to administrative work — billing, coding, prior authorization, appeals — that simply does not exist at the same scale in countries with simpler systems. The graph shows that a major government investment in electronic health records (about $38 billion) made this worse rather than better, because the software had to serve all of the different billing requirements simultaneously.

The Closed Loop: Why High Prices Don’t Get Fixed

In most markets, when prices are too high, competition eventually brings them down. In US healthcare, the graph identifies a self-reinforcing loop that prevents this from happening.

Hospitals in many regions have merged and consolidated, reducing competition. Fewer competing hospitals means higher prices. Higher prices generate more revenue. More revenue funds lobbying. Lobbying maintains the rules that prevent new competition (like laws requiring government approval before a new hospital can open) and blocks price regulation. The rules maintaining consolidation and the lobbying funded by consolidation point back at each other in a circle.

The graph identifies a node called “Healthcare Industry Political Capture” as the most connected single concept in the entire map — 44 connections, touching nearly every major cost mechanism. The key structural finding is that this is not simply an obstacle sitting in front of reform. It is built from the same mechanisms it enables. The node is both cause and effect, which is why the graph describes it as “load-bearing” — removing it would require weakening the mechanisms that sustain it, which are themselves sustained by it.

Doctors, Supply, and Pricing Power

The number of medical residency training slots in the United States — the positions new doctors must complete before practicing independently — has been effectively capped since 1997, when Congress froze Medicare funding for residency programs. This was not an accident of neglect. The graph shows multiple connections between this cap and the American Medical Association, the physician organization that also controls the committee that recommends what Medicare pays for different medical services.

Fewer doctors, particularly in primary care and rural areas, means the doctors who practice have more pricing power. The same organization that benefits from physician scarcity has influence over physician prices. This is a separate feedback loop from the hospital consolidation loop, but it connects to it — both ultimately feed the same political capture mechanism.

The graph also notes that laws in many states restrict what nurse practitioners and physician assistants are allowed to do without direct physician supervision. These scope-of-practice laws reduce the supply of care providers who could offer lower-cost alternatives. Six different nodes in the graph describe variations of the physician supply cap, and three describe scope-of-practice restrictions, all pointing toward the same effect: constrained supply supporting higher prices.

A Safety Net Program That Funds Consolidation

One of the more counterintuitive findings in the graph involves a federal program called 340B, designed to help hospitals serving low-income patients afford drugs by allowing them to buy medications at steep discounts. The program appears six times in the graph, each time analyzed from a different angle.

What the graph shows is that the gap between what hospitals pay for drugs under 340B and what they charge insurance companies or patients for those same drugs generates substantial profit — and that profit has been used to fund hospital acquisitions. The safety-net subsidy becomes a consolidation subsidy. A program intended to improve access for vulnerable patients ends up funding the consolidation that drives prices higher for everyone.

The ERISA Knot

A federal law called ERISA, passed in 1974, allows large employers who self-insure their employees’ health benefits to operate outside state insurance regulations. This was intended to let multistate companies offer consistent benefits across all their locations without having to comply with fifty different sets of state rules.

The graph identifies a striking structural consequence: the same legal provision that prevents states from regulating healthcare prices (because large employers can simply exempt themselves from state rules) also gives large employers the power to bypass insurance companies entirely and contract directly with hospital systems and drug manufacturers. ERISA is simultaneously the primary barrier to state-level reform and the primary vehicle for the only proven market-based workaround.

There is a further complication. When large employers successfully negotiate lower prices by contracting directly, they reduce their own healthcare costs — and also reduce their incentive to advocate for broader systemic reform. The partial solution may weaken pressure for the complete solution.

The MLR Regulation That Backfired

The Affordable Care Act required that insurance companies spend at least 80 to 85 percent of premium revenue on actual healthcare (the “medical loss ratio” requirement). The intent was to limit insurer profit. The graph identifies an unintended consequence: because profit is calculated as a percentage of total spending, an insurer’s absolute profit grows when total spending grows. A rule designed to constrain profit ends up giving insurers a structural reason not to aggressively contain costs. One large insurer in the graph is shown explicitly exploiting this dynamic by acquiring the companies that provide care, creating a situation where spending more on care they own generates more revenue across the combined entity.

What Would Actually Change Things

The graph identifies several mechanisms that countries with lower costs have used effectively: all-payer rate setting (one set of prices that all payers must use), negotiated drug pricing, primary care investment that reduces expensive specialist and emergency care, and simplified billing. It also identifies why these mechanisms have not been adopted in the United States — each faces a specific structural barrier that the political capture loop maintains.

One state, Maryland, operates a version of all-payer rate setting and shows measurably different cost trajectories. The graph notes that Maryland’s model depends on a unique federal waiver that no other state has successfully replicated, and that ERISA exemptions limit its scope even within Maryland. The model works; its replication is blocked.

The Bottom Line

The graph’s structural findings can be summarized in four points.

First, the core driver of US healthcare costs is prices, not how much care Americans use. Americans do not visit doctors significantly more often than people in comparable countries; they pay dramatically more per visit.

Second, high prices persist because the actors who set prices have accumulated enough political influence to block the policy changes that would constrain them. This is not a static obstacle — it is a self-reinforcing loop where high prices fund the influence that maintains high prices.

Third, the system’s architecture — hundreds of separate payers negotiating separately — multiplies administrative costs and undermines every mechanism that has reduced costs elsewhere. Fragmentation is not one problem among many; it is the structural condition that makes most other problems worse.

Fourth, several targeted reforms (340B, electronic health records, medical loss ratio rules) were designed to improve the system and instead produced unintended effects that reinforced existing cost drivers. The graph suggests that reforms introduced into a system with strong self-reinforcing dynamics tend to be captured or redirected by those dynamics rather than correcting them.

The graph does not identify a simple fix. It identifies a system where the interventions most likely to reduce costs face the strongest structural resistance from the actors most capable of blocking them.