Jordan Hale: Hey, before we start — I have to tell you that I read something this week and it made me feel like the entire healthcare system is built on a kind of collective hallucination, and I need you to either confirm that or talk me down.
Alex Mercer: That's a high bar. What specifically?
Jordan Hale: The University of Michigan Institute for Healthcare Policy and Innovation — the IHPI — they estimated that cancer screenings saved the U.S. healthcare system at least six-point-five trillion dollars over twenty-five years. Trillion. And nobody ever saw that money. It's a modeled estimate of costs that didn't happen. We're celebrating savings that exist only in a counterfactual.
Alex Mercer: Huh. No, that's the right instinct — that number does a lot of work in the prevention argument and almost nobody interrogates how it's constructed.
Jordan Hale: And then you look at something like the Disease Control Priorities benchmarking data — breast cancer screening via clinical breast examination, five hundred and fifty-two dollars per life year saved. Treatment for the same cancers? Thirteen hundred to over a hundred and sixty-three thousand per life year saved. Same disease, completely different orders of magnitude. So if those numbers hold up, why is every structural incentive still pointing toward treatment?
Alex Mercer: That's the core of it, yeah — and I think part of the answer is that those numbers are measuring different things than people assume. Cost-effectiveness and cost-saving aren't the same claim. An intervention can produce health gains at acceptable cost per life year — that's cost-effectiveness — without actually lowering the total spending curve. The fee-for-service payment system doesn't care about cost-effectiveness ratios. It cares about billable encounters, and a prevented illness produces none.
Jordan Hale: So the numbers feel like they settle it, but structurally nothing changes. That's — yeah, that's the hallucination.
Alex Mercer: To some extent. The question is whether that's a market design problem, a measurement problem, or something else entirely — and I'm not sure we've been asking it precisely enough.
Jordan Hale: But wait — the slipperiness is actually the problem, right? Like, you said cost-effective and cost-saving aren't the same thing, and I think I nodded along, but I'm not sure I actually felt the difference. What does that gap look like in plain terms?
Alex Mercer: Think of it like buying a gym membership. It might make you healthier and spare you a hospital bill someday — but it's not free, and you might have stayed healthy anyway. Cost-effective means the value is real; cost-saving means you actually spend less overall. Those are genuinely different claims.
Jordan Hale: Oh. Oh, that's — yeah, okay. So prevention can be worth it without the total bill going down.
Alex Mercer: Exactly. And the IHPI six-point-five trillion figure — that's a modeled estimate of avoided costs, not an audited accounting. It's subject to the same probabilistic uncertainty that undermines prevention's market value in the first place. Which doesn't mean it's wrong, but it means we shouldn't treat it like a bank statement.
Jordan Hale: So what's a cleaner example — like, where the evidence is actually tight?
Alex Mercer: The Oregon Tobacco Quit Line trial, 2007 — four thousand six hundred and fourteen adult smokers, randomized. Smoking cessation counseling plus nicotine replacement therapy, measured against long-term cardiovascular outcomes. It demonstrated the intervention is highly cost-effective. Not that it collapsed the total cost curve — cost-effective. That distinction is doing real work there, and I think it's a more honest example than the trillion-dollar headline because it's actually empirical, not modeled backward from a counterfactual.
Jordan Hale: And you can't just scale that finding up to everything, either — I mean, what's cost-effective for four thousand smokers in Oregon in 2007 might not hold when you're running a national program across completely different health system contexts.
Alex Mercer: That's basically the scalability problem — the evidence explicitly flags it. Context-dependence erodes the estimate. And the danger is, if we keep saying 'prevention saves money' without that precision, we hand ammunition to anyone who wants to defund a public health program the moment one trial doesn't replicate. The directional truth is strong — prevention generally wins — but the imprecision is a real liability.
Jordan Hale: And that imprecision isn't just academic, you know, because the structural problem underneath it is — okay, like, picture a radiologist submitting a billing code for a chemotherapy infusion. There's a patient, there's a chart, there's a revenue cycle, the whole machine turns. Now picture a community health worker in rural Nairobi completing a cardiovascular screening — and because of that, some guy never develops hypertension worth treating. That second thing generates a statistical non-event. No code. No claim. No payment.
Alex Mercer: It disappears from every ledger.
Jordan Hale: Completely. And that's not an accident — fee-for-service is literally architected that way. Billable encounter exists, revenue flows. No encounter, nothing flows. So a prevented illness is, like, structurally invisible to the payment system.
Alex Mercer: The pharmaceutical patent system compounds it. Treatment innovations get exclusivity — you can point to the molecule, the manufacturer, the patient who took it. Attribution is clean, so the revenue capture is clean. Try doing that for a prevention program. Who owns the outcome that didn't happen?
Jordan Hale: Wait — so even if a pharma company developed a perfect prevention drug, the patent gives them less leverage than a treatment drug?
Alex Mercer: Basically, yes. And then layer in the insurer churn problem. An insurer pays for your prevention program today — the payoff comes in fifteen years. By then, statistically, you've switched insurers. So the company absorbing the cost now is funding a benefit that lands on a competitor's balance sheet later. A well-meaning insurer still has no financial reason to invest.
Jordan Hale: That's — I mean, that's not a market failure in the sense of something going wrong. That's the market working exactly as it's designed.
Alex Mercer: Right — and that's the time horizon problem in one sentence. Private insurers face churn. Companies face quarterly earnings. Governments face electoral cycles. None of those windows align with prevention's payoff, which might be twenty years out.
Jordan Hale: Which is what makes the thing we haven't gotten to yet so strange — because if this is purely a fee-for-service and patent problem, systems without those incentives should've solved it, and they haven't.
Alex Mercer: That's the thing that actually unsettles me — the UK's National Health Service, no fee-for-service, no shareholder pressure, still spends the majority on treatment. The WHO has been flagging this structural misalignment for decades. The diagnosis isn't new. So if the payment model isn't the ceiling, what is?
Jordan Hale: And then — wait, this is where the COVID vaccine thing breaks my brain a little — because governments absolutely CAN override the market when they want to. Advance market commitments, push funding, expedited pathways. They did it. Fast.
Alex Mercer: Right — and that proves the tools exist. AMCs, direct R&D subsidies, the CDC vaccine schedules functioning as a sustained pull mechanism. Political will materialized, and the market got overridden. Which makes the persistence of the imbalance everywhere else almost harder to explain, not easier.
Jordan Hale: Okay but antiretroviral therapy for HIV — that's treatment, but it also reduces transmission, so it's functionally prevention. And it attracted enormous investment. Why? Because it generates visible, attributable, recurring revenue. A patient, a pill, a claim.
Alex Mercer: Which confirms it's not really about the biological category at all.
Jordan Hale: Exactly — like, the incentive follows attribution and revenue structure, not whether something is technically a vaccine or a drug. The label doesn't matter. What matters is whether someone billable gets better in a way you can point to.
Alex Mercer: And national health services — I mean, they still have budget cycles, they still have ministers defending line items, they still need to show outputs someone can count. A prevented illness still doesn't generate that. The attribution problem survives the removal of profit motive entirely.
Jordan Hale: So it's not — wait, it's not that the market is wrong. It's that every institution we've built, market or not, measures rescue and ignores the statistical person who never got sick.
Alex Mercer: Basically. And until the measurement changes — not the payment model, the measurement — the incentive follows the visible patient. Every time.
Jordan Hale: And that's — you know, the thing is we started with that six-point-five trillion figure, this ghost number of savings that never appeared on any ledger. And after all of this, it's still a ghost. Not because the tools don't exist — like, COVID vaccines, advance market commitments, push funding, CDC schedules holding up demand — we proved we can do it. The capability isn't the mystery.
Alex Mercer: I think the mystery is whether institutions can ever be rewired to see a non-event as an output. The statistical person who never got sick — that's not legible to a budget cycle. It wasn't legible before fee-for-service, and it survives the removal of profit motive entirely. That's not a fixable incentive. That's an architectural limit.
Jordan Hale: Which is maybe the most unsettling place to land — not that it's broken, but that it might be working exactly as every institution is built to work. I don't know. Thanks for sitting in the discomfort of that with me.
