The Economics of Giving Up

The numbers describe a market that has stopped fighting. Across HealthCare.gov plans in 2023, insurers denied 86 million in-network claims; consumers appealed 376,508 of them — an appeal rate below 1%.¹ On the provider side, the industry's own estimate is that up to 65% of denied claims are never resubmitted at all.² The claims simply die, written off as the cost of participating in American billing.

The usual reading is fatigue, or ignorance of the appeal right. The record suggests something less flattering to the system and more flattering to the practices: the math. When providers do fight, they win more often than they lose — Premier's survey found 54.3% of private-payer denials ultimately overturned and paid.³ A majority-win game that almost nobody plays is not evidence of apathy. It is evidence that the entry fee exceeds the pot.

The cost side is measured, repeatedly and consistently. MGMA puts the cost of reworking one denied claim at $25.20 at the practice level.⁴ Premier, surveying providers on full claims adjudication, measured $43.84 per claim in 2022 and $57.23 in 2023.⁵ For formal appeals in hospital settings, the figure MGMA cites runs to an average of $118 per denied claim.⁶ These measure different depths of the same fight — see the methodology — but they bracket the entry fee: roughly $25 to $118 of staff time per denial.

The revenue side, for the visits that fill a small practice's day, is public. Medicare's 2026 physician fee schedule pays a national non-facility rate of $95.19 for a 99213 office visit⁷ and $135.61 for a 99214.⁸ Put the two ledgers on one scale and the problem is visible before any algebra:

Take a staff cost of $50 per appeal — a round figure inside Premier's measured $43.84–$57.23 range. If winning were guaranteed, any claim above $50 would be worth fighting. It is not guaranteed: at the 54.3% overturn rate, the expected recovery on a claim worth V is 0.543 × V, and the expected-value break-even rises to $50 ÷ 0.543 ≈ $92. A 99213 — the workhorse visit of American primary care — clears that threshold by about three dollars of claim value, which is to say: appealing the most common denial a small practice sees returns less than $2 of expected profit.

The table locates exactly where the giving-up concentrates: not on the expensive imaging study or the surgical claim, where the fight funds itself several times over, but on the low-hundreds E/M visits that make up most of a primary-care practice's revenue. A practice whose denied claims are mostly 99213s is not failing to defend its revenue when it writes them off. It is defending its payroll. And the threshold moves with the cost: at the $118 formal-appeal figure, the expected-value break-even climbs to about $217 — a level at which even a 99214 loses money in expectation, win rate and all.

This is the mechanism behind the opening statistics. A 54.3% win rate coexists with a sub-1% appeal rate because the winnable fights are not worth having at the prevailing price of fighting. Payers do not need denials to survive review; they need them to cost more to contest than the claim is worth. On claims under roughly $92, at current costs, that condition holds automatically.

The break-even has only two inputs: cost and win rate. The win rate is the weaker lever. It is already above half, it cannot exceed one, and even a perfect 100% win rate at $50 of cost still abandons every claim under $50 — which excludes a large share of the lab work, imaging components, and brief visits that get denied in volume.

The cost lever scales the threshold linearly, all the way down. At $25 of cost per appeal, the expected-value break-even falls to about $46 — below the 99213, which flips the largest single category of small-practice denials from write-off to worth fighting. At $10, the threshold is about $18. At $5, about $9, and nearly every denied claim in the practice becomes rational to contest. Nothing else in the system has to move: not the payer's behavior, not the win rate, not the fee schedule. The economics of giving up are an equation with one dominant variable, and the equation is indifferent to how that variable falls — cheaper staff time, better tooling, batching, delegation. What it is not indifferent to is whether it falls. Until it does, a sub-1% appeal rate is not a failure of nerve. It is arithmetic, done correctly, by people who cannot afford to do it wrong.