Physician Practice Ownership and Consolidation Statistics

What the 2024 survey found

The American Medical Association's 2024 Physician Practice Benchmark Survey estimated that 42.2% of physicians worked in private practices. The comparable estimate was 60.1% in 2012. This is an 18 percentage point change over twelve years, not an estimate that 18% of practices closed. The survey follows physicians and the characteristics of their main practice, so the physician is the analytical unit.

The same report estimated that 34.5% of physicians worked in hospital-owned practices in 2024, compared with 23.4% in 2012. Another 12% were employed by or contracted directly with a hospital. Those groups describe different arrangements. A physician can work in a practice owned by a hospital, while direct hospital employment refers to the physician's relationship with the hospital itself.

Private equity was a smaller reported ownership category. The AMA estimated that 6.5% of physicians worked in private-equity-owned practices in 2024, up from about 4.5% in both 2020 and 2022. The result does not mean that 6.5% of all legal medical-practice entities are owned by private equity. Larger organizations employ more physicians, which can give them a larger physician-weighted share than an establishment count would show.

These figures document a national shift in where physicians work. They do not predict whether a particular independent practice will remain independent, sell, merge, or grow. They also do not measure the quality of care, financial health, or working conditions produced by any ownership form.

Definitions that change the answer

"Private practice" in the AMA report refers to a practice wholly owned by physicians. It is not a synonym for every outpatient office outside a government building. A physician working in an office acquired by a hospital or insurer may still see patients in a setting that looks like a conventional private office, but the survey ownership category has changed.

Employment status is another variable. Owners, employees, and independent contractors are physician-level relationships. Practice ownership describes the organization. A physician owner can work beside employed physicians, and an independent contractor can provide services in an organization owned by someone else. An analysis that treats every employed physician as evidence of hospital ownership will misclassify physician-owned groups that employ associates.

Practice size also needs a stated unit. The number of physicians in the main practice is different from the number of clinicians, sites, billing identifiers, or people in a parent health system. A multisite group may have one ownership structure and many establishments. A federal establishment dataset and a physician survey can both be accurate while returning different counts.

The AMA survey also reports an additional group of physicians who said they or another physician retained an ownership stake but described the practice as owned by a hospital, insurer, or private equity group. This is why ownership percentages should come from the published mutually exclusive categories, not from adding appealing numbers found in separate paragraphs.

What consolidation studies measure

Research on consolidation uses several designs. Transaction databases count announced or completed acquisitions. Claims-based studies infer changes in billing relationships. Business registries count establishments or legal entities. Surveys ask physicians about their main practice. Each reveals part of the market, but none is a universal census of independent practices.

The AHRQ Compendium of U.S. Health Systems maps hospitals, physician organizations, and systems using multiple data sources. Its purpose and denominator differ from the AMA Benchmark Survey. The Federal Trade Commission studies competition and enforces antitrust law. Government Accountability Office reports may examine acquisition activity and data gaps. Peer-reviewed studies often focus on a particular specialty, ownership type, payer, or observation period.

Dates require care. A study published in 2024 may analyze acquisitions completed several years earlier. An ownership survey fielded in 2024 describes respondents during its survey period. Neither should be relabeled as a live count for August 2026. For a trend, compare waves that use sufficiently consistent questions and sampling methods.

Percent changes and percentage point changes are not interchangeable. The private-practice share fell 17.9 percentage points from 60.1% to 42.2%. Relative to the 2012 level, that is about a 30% decline in the physician share. The percentage point description is usually clearer because the reported values are already percentages.

How an independent practice can use the figures

National ownership statistics can frame a strategic discussion, but they cannot supply the practice's answer. Start with a simple ownership map: legal entity, voting rights, equity interests, management agreements, leases, billing relationships, and any parent or affiliated organization. Have qualified legal and accounting advisers interpret contracts and tax consequences. A research article cannot replace that review.

For operating comparisons, define practice size consistently. Record physician full-time equivalents, other clinician full-time equivalents, locations, active patient panels, visits, and payer mix separately. Do not compare a head count from one practice with full-time equivalents from another. Note whether locums, contractors, and part-time physicians are included.

If the practice is considering a transaction, keep descriptive market evidence separate from valuation. A national ownership trend does not establish a fair price, expected revenue, or future autonomy. Transaction terms, local referral patterns, payer contracts, liabilities, real estate, technology obligations, and governance rights can matter more than a national percentage.

The statistics are also useful for vendor claims. A seller may cite consolidation as proof that a solo model is impossible, or cite independence as proof that every small practice can thrive. Neither conclusion follows from the AMA estimates. The figures describe the distribution of physicians across ownership arrangements. They do not test the performance of a specific service or business plan.

For an internal annual snapshot, preserve source documents and definitions. Record the date, survey wave, unit, population, and exact category. If a later report changes its classification, keep an overlap table rather than silently rewriting the old series. That makes the comparison reproducible.

Methodology and limits

We reviewed ten sources spanning the AMA national physician survey, federal provider and health-system data, competition oversight, government review, and peer-reviewed consolidation research. The AMA report is the main source for the three headline percentages because it publishes the relevant 2024 physician-weighted categories and historical comparisons.

The sources do not share one denominator. CMS data describe enrolled clinicians and organizations for program purposes. AHRQ identifies health systems and their components. Transaction research may count deals or acquired locations. Academic articles may study selected specialties or markets. We did not pool these measures.

Survey estimates are subject to sampling, response, recall, and classification error. Ownership structures can also be layered. Published categories simplify contracts that may divide clinical, financial, and governance control. The estimates are national and do not describe every state, specialty, or local market equally well.

The defensible conclusion is narrow. A smaller share of physicians worked in wholly physician-owned practices in the 2024 AMA survey than in 2012, while hospital-owned arrangements accounted for a larger share. Private equity ownership remained a separate, smaller category in the survey. An independent practice can use those findings as context, provided it keeps physician shares distinct from practice counts and does not turn a national trend into a forecast for one office.