The Economics of Value-Based Care

Most conversations about value-based care are conversations about medicine. Care teams. Risk stratification. Closing gaps. Meeting patients where they are.

All of that matters. None of it explains why value-based care moves quickly in one organization and stalls for a decade in another.

The explanation is usually arithmetic.

Value-based care is a change in how a business earns money. It changes what counts as revenue, when that revenue arrives, who absorbs the cost of being wrong, and which expenses actually fall when volume falls. Leaders who understand the clinical model but not the financial one tend to make a series of decisions that look reasonable one at a time and add up to years of expensive standing still.

So it is worth walking through the arithmetic plainly.

You Are Running Two Businesses That Want Opposite Things

During the transition (and it's long), an organization operates two income statements at once.

In the first, an occupied bed is revenue. In the second, an occupied bed is a cost. In the first, a well-managed heart failure patient who stays out of the hospital is a lost admission. In the second, that same patient is the entire point.

These are not two views of one business. They are two businesses with opposing incentives, sharing the same doctors, the same buildings, and the same budget process. Every capital request, every service line decision, and every hiring plan has to be argued before both.

This is why so many organizations end up with a value-based care department rather than a value-based care strategy. A department is what you create when you want to participate without letting the second business threaten the first. It is a rational hedge. It is also the reason the numbers rarely move.

The uncomfortable part is that the hedge has a cost. You pay for the infrastructure of the second business while continuing to be measured on the first. You get the expenses of transformation and the economics of the status quo.

The financial exposure behind that tension is becoming harder to treat as peripheral. According to CMS’s 2026 Medicare Accountable Care Organization Initiatives Participation Highlights, 82.8% of Medicare Shared Savings Program ACOs are in BASIC Level E or ENHANCED in 2026, both performance-based risk tracks that qualify as Advanced Alternative Payment Models. CMS says this is the highest percentage since the program began in 2012.

The Money Shows Up Late, and Only if Everyone Else Performs Too

The cost structure of value-based care is immediate and fixed. Care managers, pharmacists, analytics, data feeds, quality abstraction, contracting talent. You hire these people in January, and you pay them every month.

The revenue is delayed and contingent. Shared savings are calculated after the performance year closes, reconciled against a benchmark, adjusted for quality, and paid out somewhere between nine and eighteen months later. You are financing a full year of new operating cost, then waiting most of another year to learn whether any of it was rewarded.

Contingency is the sharper problem. Your care management team can do excellent work on its own panel and still receive nothing, because the payment depends on the entire attributed population beating a benchmark. Individual performance does not guarantee a settlement. Aggregate performance does.

This is not a conviction problem or a clinical problem. It is a working capital problem, and it should be planned for the way any other multi-year capital commitment is planned. Organizations that fund value-based care out of operating slack tend to quit in year two, which is precisely when the curve is at its lowest, and the learning is at its highest.

Your Revenue Is a Number Someone Else Calculated

In fee-for-service, revenue is something you generate. You perform a service, you bill for it, you collect a knowable amount.

In value-based care, revenue is the difference between what was spent on a population and what a benchmark said should have been spent. You do not set the benchmark. You often cannot fully predict it. In many programs, it is built partly on your own historical spending, which has the effect of quietly discouraging more organizations than any other single feature of these programs.

Perform well, and your benchmark resets lower. Perform well again, and you are asked to find savings against a baseline that already reflects the savings you found. The organization that has been efficient for twenty years starts the race with less room than the organization that has been wasteful for twenty years.

CMS explicitly refers to this dynamic as the rebasing ratchet effect. In CMS’s Medicare Shared Savings Program benchmark changes, the agency introduced a prior savings adjustment to mitigate the effect by returning value to an ACO’s benchmark based on its earlier success in lowering expenditure growth.

Program designers know this and have introduced regional adjustments and efficiency adjustments to soften it. The softening is partial. The strategic conclusion stands: in value-based care, benchmark methodology is not a technical footnote. It is the product specification for your revenue. It deserves the attention a hospital would give to its chargemaster or a payer would give to its rate filing, and in most organizations it receives a fraction of that.

Fixed Costs Are Why Hospital Economics and Value Economics Collide

Here is the single calculation that explains more about the value-based care landscape than any other.

Suppose you avoid an admission. The admission would have been reimbursed for $12,000. The variable cost of delivering it, including supplies, drugs, and incremental staffing, was $5,000. You have therefore given up $7,000 in contribution margin toward the fixed costs of the building, equipment, and salaried staff who are there whether the bed is full or not. In return, the population spent $12,000 less. At a 50% sharing rate, you receive $6,000. You did exactly what the model asked. You are $1000 worse off. And next cycle, the benchmark comes down.

Nothing about this is an argument against value-based care. It is an argument about who is structurally suited to it. Organizations with a high ratio of fixed to variable cost lose real margin when volume declines. Organizations built mostly around clinician time lose much less. This is not my opinion. CMS reported that in the 2024 performance year of the Medicare Shared Savings Program, accountable care organizations made up largely of primary care clinicians produced net per capita savings of roughly $400, against roughly $220 for those with fewer primary care clinicians.

The gap is not a difference in clinical talent. Read as an operator, it is a difference in cost structure.

For a health system, the implication is not to retreat. It is that avoided utilization alone will not carry the business case. The case has to be built on backfilling freed capacity with higher-margin work, on shifting care to lower-cost settings you own, on retaining patients who currently leave the network, and on holding a large enough share of the population so that the sharing rate is worth the margin given up.

Scale Is Not Ambition. It Is Statistics

Population health results are noisy. A small panel with a handful of catastrophic cases in a single year will miss its benchmark for reasons that have nothing to do with how the population was managed.

This is why minimum population thresholds exist, and why the practical floor for meaningful risk is higher than the regulatory one. Below a certain size, you are not measuring performance. You are measuring variance and paying consultants to explain it.

The regulatory threshold itself reflects the statistical problem. The Medicare Shared Savings Program requires an ACO to have at least 5,000 assigned Medicare fee-for-service beneficiaries. CMS states this is intended to support reliable and accurate assessment of ACO financial and quality performance.

Two consequences follow. First, reinsurance and stop-loss are genuine line items, not afterthoughts, and their costs must be modeled into the business case from the beginning. Second, aggregation has real economic value, which is the entire reason enablement companies exist. They pool lives, spread variance, and amortize infrastructure across many small practices that could never afford it alone.

The Share of Revenue at Risk Is the Number That Decides Everything

Take an organization where value-based arrangements cover 8% of revenue. Assume it performs superbly and earns a large share of the savings on that 8%. The effect on enterprise economics is close to nothing.

Meanwhile, the disruption to physician workflow, referral patterns, documentation, and scheduling is felt across the entire practice because you cannot run one clinical model for 8% of patients and a different one for the rest.

This is the central economic trap. Below a threshold, the cost of transformation is spread across the entire organization while the benefit is confined to a sliver of it. The transformation looks expensive, and the returns look thin, and both readings are correct.

The math only turns when enough of the book is under aligned incentives that changing behavior once produces returns across most of the population. That threshold is different for every organization, but it is almost always higher than where hedging instincts naturally settle. Which produces an awkward conclusion for anyone advising caution.

Partial commitment is not the low-risk option. It is the option that reliably produces cost without return. The two defensible positions are staying in fee-for-service deliberately and going far enough into risk that the economics can actually work. The middle is where money goes to die quietly.

The Value Is Captured in One Place and the Work Is Done in Another

Almost every failed value-based care program has the same internal signature. The system captures the savings. The physicians absorb the effort.

A primary care physician managing risk well does more unpaid work. Longer visits with complex patients. Documentation that is complete rather than merely sufficient. Calls, coordination, follow-up that generates no billable event. If the compensation model still pays by relative value units, the system is asking clinicians to reduce their own income to increase the organization's.

The compensation data shows why the conflict persists. According to the AMA’s 2024 Physician Practice Benchmark Survey, 55% of physicians received at least some compensation from productivity metrics, while productivity accounted for 28.1% of physician compensation on average.

They will decline, politely and indefinitely, and the program will be described as a change management failure. It is not. It is an incentive design failure, and it is fixable with internal economics that passes a real share of the upside to the people generating it, early enough that they experience the connection between the behavior and the reward.

Programs Are Shorter Than Investments

The last piece of the arithmetic is timing risk of a different kind.

Building the capability for risk is a five-to-seven-year investment. Many of the programs that capability is built to run operate in three-to-five-year cycles, then change or end. ACO REACH is scheduled to conclude at the end of 2026, while Kidney Care Choices now runs through 2027. The Transforming Episode Accountability Model became mandatory for selected hospitals in 2026, with different participation tracks providing different paths into downside risk. National direction has held steady across administrations, but the specific vehicles turn over regularly.

The response is not to wait for stability, because stability is not coming. It is to invest in capabilities that survive a change in program design. Accurate longitudinal data on your population, a clear view of total cost of care by patient and by physician, functioning care management, and the ability to model a new contract quickly are valuable under every model that has existed and every model likely to follow. Investments tightly coupled to one program's rules are the ones that get written off.

What This Actually Adds Up To

Value-based care is not a moral question. It is a business model question, and the businesses best suited to it look different from the ones that dominate healthcare delivery today.

The organizations that have made it work share a short list of traits. They understand their marginal cost, not just their average cost. They know what their benchmark will be before they sign. They treat delayed and contingent revenue as a financing problem rather than a forecasting problem. They concentrate enough of their population under aligned incentives for behavior change to pay off. And they route real money to the clinicians doing the work.

One useful property of that list is that none of it is wasted if you decide to stay in fee-for-service. Knowing total cost of care by patient and by physician, being able to model a contract quickly, and holding clean longitudinal data on your population improve your negotiating position with payers, your service line decisions, and your denial performance regardless of which way you go. The capabilities are what hold their value. The specific contract is what expires.

That list contains no clinical innovation at all. Which is roughly the point. The medicine has been well understood for years. Finance is where the transformation is actually won or lost, and it is the part most often delegated downward, while strategy conversations stay comfortably clinical.

The reason to get the arithmetic right is not that it makes value-based care easy, but because arithmetic is the only thing standing between a genuine model of care and an expensive department that exists to prove you tried.

I will be at HLTH USA 2026 at The Venetian Expo, Las Vegas, from November 15 to 18. If you are pressure-testing the economics of value-based care in your own organization, Coditas’ ClarityVBC provides real-time analytics across attribution, risk, quality, cost, and savings so teams can see what is changing and where attention is needed. Let’s continue that conversation at Booth 4051, Level 2, Exhibit Hall, AI Zone.