Anyone who has received a medical bill in the United States knows the experience: a document listing charges that bear no relationship to anything, an insurance statement that isn't a bill, and a final amount that arrives weeks later and can't be predicted in advance.
It's tempting to attribute this to incompetence. It's more accurately understood as the necessary output of how prices are set.
There is no price
The foundational fact. A medical procedure does not have a price in the ordinary sense. It has many prices, depending on who is paying.
A hospital maintains a list of charges — historically called the chargemaster — which is essentially a starting point that almost nobody pays. Each insurer then negotiates its own discounted rates, separately, confidentially.
So the same procedure at the same hospital on the same day has different prices for a dozen different insurers, plus a different price for government programmes, plus a different one for the uninsured.
Which means the question "what does this cost" has no single answer, and any billing system built on top of that will be complicated by necessity.
The network problem
The second structural layer. Insurers contract with networks of providers, and treatment from a provider outside the network is covered differently or not at all.
That's manageable when you're choosing a doctor. It becomes unmanageable in a hospital, where the facility may be in network while individual practitioners working there — anaesthetists, radiologists, pathologists, emergency physicians — are separately contracted and may not be.
This produced the surprise billing phenomenon: a patient who did everything right, at an in-network hospital, receiving a large bill from a practitioner they never chose and could not have identified in advance.
Federal legislation has restricted this practice for many situations, which was a genuine improvement. The underlying structure that made it possible remains.
Why the numbers look absurd
The eye-watering figures on an itemised bill exist for reasons that have little to do with cost.
Because insurer payments are negotiated as discounts from list prices, a higher list price produces a higher negotiated amount for any given discount percentage. That creates an upward ratchet on list prices that's disconnected from actual costs.
List prices also matter for certain payment formulas and for uninsured patients, who are the only group ever billed at the full amount — which is to say the people least able to pay face the highest prices, an outcome nobody designed and everybody could have predicted.
The administrative cost
All of this requires people. Coders translating clinical activity into billing codes. Billing staff submitting claims. Insurance staff adjudicating them. Providers appealing denials. Patients calling to understand statements.
Studies comparing administrative spending in healthcare across countries consistently find the US at the top by a substantial margin, with billing and insurance-related administration accounting for a significant share of total health spending.
That's not waste in the sense of anyone being idle. It's the labour required to operate a system with thousands of separately negotiated price schedules, and it would largely not exist under a single fee schedule.
Price transparency rules
Regulations requiring hospitals and insurers to publish negotiated rates have been introduced, with the aim of enabling comparison.
Compliance has been uneven and the published data is frequently in formats that are technically compliant and practically unusable — enormous machine-readable files with inconsistent structures.
There's also a genuine question about whether transparency alone changes behaviour. Healthcare purchasing is frequently urgent, referral-driven, and geographically constrained. A patient with an acute problem is not comparison shopping, and for many services there's no realistic alternative provider within reach.
Transparency helps most for planned, shoppable procedures, which are a minority of spending.
What patients can actually do
Limited, and some things genuinely help.
For planned procedures, request a written estimate in advance and ask specifically whether every practitioner involved is in network.
Always request an itemised bill. Billing errors are common, and duplicate charges or services not received appear regularly.
Do not pay before the insurer has processed the claim. The initial statement from a provider is frequently not the final amount.
Ask about financial assistance. Nonprofit hospitals generally have obligations to provide charity care, and eligibility thresholds are often higher than people assume. These programmes are real and are not well advertised.
And appeal denials. A meaningful proportion of denied claims are overturned on appeal, and the appeal rate is low, which suggests a lot of people are paying for things that would have been covered had they pushed.
The employer link
One structural feature that shapes everything else and is largely a historical accident: employment-based health coverage.
It emerged substantially from wartime wage controls, when employers competed for workers by offering benefits rather than pay, and was then entrenched by tax treatment favouring employer-provided coverage.
The consequences run deep. Coverage is tied to a job, which affects mobility and bargaining power. The purchaser negotiating with insurers is your employer rather than you, so the product is designed for their priorities. And changing jobs means changing networks, deductibles and sometimes doctors.
Nobody designing a system from scratch would produce this. It persists because unwinding it would disrupt coverage for an enormous number of people at once, which is a genuine constraint rather than an excuse.