Junto Analytics

There is something about the relationship between physicians and the systems that pay them that does not occur in any other professional market. A lawyer at a firm knows roughly what their billable hours generate in revenue. A consultant at McKinsey or Bain can calculate the daily rate the firm charges for their work. A software engineer at Google can look up roughly what their team’s products generate. The economic value created by their labor is visible to them, and the share of that value flowing to them as compensation is calculable.

Physicians work without this visibility. A surgeon performing a knee replacement at a major academic medical center generates a specific dollar amount of revenue for the institution. The amount is known with precision to the institution’s revenue cycle department, the contracting office, the chief financial officer, and a small number of senior administrators who manage the institution’s economics. The amount is not known to the surgeon performing the procedure.

The physician knows what they are paid. They know roughly how much work they do, often expressed in work relative value units, the standardized measure of physician work that the Medicare program established and that has since become the universal currency for measuring clinical effort. They may know that their compensation is benchmarked against survey data from organizations like the Medical Group Management Association or the Association of American Medical Colleges, and they may have a general sense of where they fall in those benchmarks.

What they almost never know is what their work generates. The price the institution receives from commercial payers, Medicare, Medicaid, and self-pay patients for the procedures and visits the physician produces. The dollars per work relative value unit that flow into the institution from the payer side, before any administrative costs, facility costs, or supply costs are netted out.

There are, in other words, two prices for every wRVU. The price the institution receives. The price the physician receives. Both prices are real. The first is paid by payers to systems. The second is paid by systems to physicians. The relationship between them defines the economics of physician compensation in ways that almost no physician currently understands, because almost no physician has access to both numbers.

Junto Analytics exists to make those two prices visible.

Healthcare in the United States operates as one of the most information-asymmetric markets in the modern economy. Prices are negotiated bilaterally between payers and providers in confidential contracts. The terms of those contracts have historically been protected as trade secrets. Patients rarely know the price of their care before receiving it. Employers paying for that care through insurance benefits rarely know what specific services cost. Physicians, despite producing the work that generates the revenue, are typically the last to see the prices their labor commands.

This is not a description of a free market with some imperfections. This is a description of a market that has been deliberately structured to be opaque, with information asymmetries protected by contract terms, by regulatory frameworks that until recently permitted such opacity, and by the operational complexity of healthcare’s billing systems.

The legacy benchmark vendors who have served the physician compensation market for decades have built their business models around this opacity. Organizations like MGMA, AAMC, and SullivanCotter conduct annual surveys of healthcare employers, asking those employers to report what they pay their physicians by specialty and region. The vendors aggregate the responses, publish the medians and percentiles, and sell the resulting reports back to those same healthcare employers, who use them to set the compensation of their next year of physicians.

This system has logic, but it has limitations that are worth being explicit about. The data is collected from employers, not from physicians. The data reflects what was paid, not what should have been paid. The data is annual, with publication lag of one to two years. The methodology behind the survey aggregations is proprietary; the underlying microdata is not publicly available. The benchmarks become self-reinforcing, because employers paying within the benchmark range can point to the benchmark as justification, and employers exceeding it have reason to compress toward the median.

Most importantly, the legacy benchmarks tell physicians what they are paid relative to their peers. They do not tell physicians what their work generates relative to the revenue stream from which they are paid. The first question is interesting; the second question is essential. The legacy benchmarks answer only the first.

Beginning in 2021, the Centers for Medicare and Medicaid Services began requiring U.S. hospitals to publish standardized files containing their negotiated rates with commercial payers. The requirements have evolved through multiple rule iterations, with the most significant updates taking effect in early 2026, requiring hospitals to publish actual median allowed amounts, tenth and ninetieth percentile allowed amounts, and counts of allowed amounts, all derived from real claims data over the prior twelve to fifteen months.

A parallel set of requirements for payers, known as Transparency in Coverage rules, has required commercial health insurers to publish similarly detailed files containing their negotiated rates with in-network providers across the country.

These rules have produced an extraordinary volume of public data. Approximately 6,000 U.S. hospitals each publishing detailed rate files. Major commercial payers publishing files that, in aggregate, contain hundreds of millions of negotiated rate observations. Updated periodically. Available to anyone who knows where to find them and has the technical infrastructure to make sense of them.

For the first time in the modern history of American healthcare, the price the institution receives is observable. Not perfectly, not without gaps, not without compliance unevenness across hospitals and payers, but observable in ways that were inconceivable five years ago.

What this enables, methodologically, is the calculation of the first of the two prices for every wRVU. For any procedure performed at any hospital that has published its rates, for any payer that has signed a contract that has been disclosed, the median allowed amount divided by the procedure’s wRVU value yields a payer-to-system dollars per wRVU figure. Aggregated across the procedures a specialty actually performs, weighted by frequency, stratified by site of service and payer mix, this produces a defensible estimate of what an institution receives per wRVU of work performed by physicians in that specialty in that market.

This calculation has not been broadly available to physicians before. It is now possible.

If the price the institution receives per wRVU can now be calculated from public data, and the price the institution pays the physician per wRVU is approximately known to the physician (it is, after all, embedded in their employment contract), then a third number becomes calculable: the ratio between the two.

We call this ratio the yield, by analogy to the financial concept. In financial markets, yield is the relationship between what an asset produces and what it costs. In physician compensation, yield is the relationship between what the physician produces for the system and what the system pays the physician. A physician working in a specialty where the institution receives, on average, two hundred dollars per wRVU from payers, and where the institution pays the physician sixty dollars per wRVU as compensation, has a yield of thirty percent. The other seventy percent flows to the institution to cover overhead, facility costs, supplies, and to generate the operating margin the institution needs to function.

There is no objectively correct yield. The right share of revenue flowing to the physician versus the institution depends on a great many factors: the specialty’s typical overhead structure, the institution’s mission and financial position, the local labor market, the physician’s individual productivity and experience, the contract terms negotiated by either party, the regulatory environment, and the relative bargaining power of each side at the time of contracting.

What is true is that yield varies meaningfully across specialties, across institutions, across geographies, and over time. A primary care physician in a high-overhead specialty may have a yield close to or above one hundred percent because the institution loses money on primary care and offsets the loss through downstream revenue. A surgical specialist in a high-revenue specialty may have a yield of fifteen or twenty percent because the institution captures the substantial facility revenue from the procedures the surgeon performs. The variation is not chaos; it reflects the underlying economics of how different specialties contribute to institutional finances.

What is also true is that almost no physician currently knows their yield. The information asymmetry that has historically protected the institution from physicians having this information is the same asymmetry that prevents the labor market from functioning efficiently. A physician deciding between two job offers has limited basis on which to evaluate them beyond the headline compensation numbers. A physician negotiating a contract renewal has limited basis on which to argue for changes beyond what the survey benchmarks suggest. A physician choosing between specialties in training has limited basis on which to anticipate how the choice will translate into long-term economic outcomes.

This is the gap that price transparency, properly applied, begins to close.

Junto Analytics is building the infrastructure that turns price transparency data into useful intelligence about physician compensation. The architecture has several components, each addressing a specific aspect of the calculation that produces the two prices.

We aggregate hospital price transparency files from across the United States into a normalized analytical dataset, allowing per-procedure median allowed amounts to be queried by hospital, payer, geography, and site of service. We aggregate Transparency in Coverage data from major commercial payers, providing additional perspective on negotiated rates from the payer side of the contracting relationship. We integrate data from the Centers for Medicare and Medicaid Services, including the Physician and Other Practitioners utilization file, the Physician Fee Schedule relative value file, and the National Plan and Provider Enumeration System, to provide the structural context against which the price transparency data is interpreted.

We compute, for each specialty in each metropolitan area, the median payer-to-system dollars per wRVU across the procedures that specialty actually performs, with appropriate weighting and confidence intervals. We allow this calculation to be stratified by site of service, recognizing that a procedure performed at a hospital outpatient department generates different revenue than the same procedure performed at an ambulatory surgery center or a physician office. We allow the calculation to be stratified by payer segment, recognizing that commercial payer rates differ meaningfully from Medicare and Medicaid rates. We surface the confidence intervals around these estimates honestly rather than hiding them behind point estimates.

For the system-to-physician side of the calculation, we provide tools for physicians to enter their own contract terms and productivity data, producing personalized estimates of their yield. We supplement this with calibration data from publicly disclosed compensation observations, including IRS Form 990 Schedule J filings for nonprofit health systems, which provide observed compensation for the highest-compensated employees at thousands of U.S. hospitals.

Both sides of the calculation rest on public data sources. The methodology is documented and will be published as a peer-reviewed working paper. The confidence intervals are surfaced rather than hidden. Every published figure traces to its source data and the methodology version that produced it. We treat the methodology as a scientific contribution to the field, not as a proprietary trade secret.

The question of physician compensation is often framed as a question of how much physicians should earn. We think this framing is incomplete. The more useful question is how the value created by physician labor should be distributed between the physicians who create it and the institutions that organize and capture it.

This is not a question with a single correct answer. Different specialties have different economic structures. Different institutions have different missions. Different geographies have different labor markets. Different career stages, different practice models, different individual circumstances all produce different reasonable distributions of the value created.

What the question requires, to be answerable at all, is information. Physicians cannot evaluate the fairness of their compensation without knowing what their work generates. Institutions cannot defend their compensation structures without showing the economics they reflect. Markets cannot function efficiently when one side of every transaction operates without access to the underlying economics.

For most of the modern history of American healthcare, this information has been unavailable to physicians. The legacy benchmark vendors have provided one piece of it: a view of compensation distributions across peers. The price transparency rules have made the other piece available for the first time: a view of revenue distributions across hospitals and payers. Junto Analytics exists to bring both pieces together, to apply rigorous methodology to the analysis, and to make the resulting intelligence accessible to the physicians whose work the prices represent.

The market that emerges when both prices are visible will not be the same market that has existed when only one was visible. We do not predict the specific changes; the relationship between transparency and outcomes is complex and depends on many factors beyond data availability. We do believe, with some confidence, that physicians who understand the economics of their work make better decisions than physicians who do not. We believe institutions that defend their compensation structures with reference to the economics those structures reflect operate with more integrity than institutions that obscure the economics. We believe the labor market for physicians functions better when both sides of the contract have access to the underlying numbers.

Healthcare in the United States is in the early years of a meaningful transition from an opaque pricing environment to a transparent one. The transition is uneven, contested, and will produce winners and losers in ways that are not yet clear. Junto Analytics is one of the institutions building the infrastructure that this transition requires. We are building it carefully, with rigorous methodology, with honest treatment of uncertainty, with public documentation of our work, and with explicit recognition that the value of what we build comes from being trusted by the physicians, institutions, researchers, and policymakers who eventually use it.

The two prices of your work are about to become visible. We are building the tools that will make them so.