For Profit vs Nonprofit Hospitals: Who Owns Them, Who Profits

Who owns America's hospitals, and which ownership model actually makes money? The federal cost reports answer both questions at once. Every Medicare-participating hospital files an annual cost report on form CMS-2552-10, and that single document records both the facility's type of control — for-profit, nonprofit, or government — and the revenue and expense lines needed to compute its operating margin. Joining the two across all 6,019 hospitals in the 2026-05-24 HCRIS snapshot produces a clean ownership-and-margin map, with no external data and no modeling assumptions.

The headline is stark. For-profit hospitals are the only ownership class with a positive average operating margin — +0.19%. Nonprofit hospitals average −4.75%, and government hospitals average −62.38%. But the averages hide more than they reveal, and the honest story is steadier — and more interesting — than the one-line version.

For-profit is the only ownership class in the black

The 6,019 hospitals split into three clean classes from the cost-report "Type of Control" field: 3,034 nonprofit (voluntary), 1,785 for-profit (proprietary), and 1,200 government. On the simple average of each facility's operating margin, only the for-profit class clears zero.

The cleanest single comparison is the share of each class that earns money at all. 65.0% of for-profit hospitals post a positive operating margin, against 44.1% of nonprofits and just 20.8% of government hospitals. A for-profit hospital is more than three times as likely as a government hospital to be in the black. That ratio does not depend on any outlier; it is a simple head-count, and it is the most robust way to state the finding.

For-profit hospitals are not the most profitable because they are the best run. They are the most profitable because they pick the patients and the service lines the other two models are obligated to keep.

Why the averages mislead: the median tells a steadier story

A simple average of per-facility margins treats a $2-million-revenue rural clinic the same as a billion-dollar system, so a handful of tiny hospitals with catastrophic ratios can swing the mean. The median is the cure: it is the typical hospital, immune to outliers.

On the median, the picture sharpens rather than softens. The median for-profit hospital runs +6.91% — solidly profitable. The median nonprofit runs −1.77%, a thin operating loss. The median government hospital runs −12.30%. The ordering is identical to the average, but the levels are sane: the typical for-profit makes about seven cents on the patient-care dollar, the typical nonprofit loses two, and the typical public hospital loses twelve.

The financial-distress rate — the share of each class with an operating margin below the −5% mark that rating agencies and MedPAC treat as a warning line — tells the same story a third way. 27.1% of for-profit hospitals are distressed, against 39.5% of nonprofits and 67.6% of government hospitals. Two in three public hospitals are running operating losses deep enough to count as distress.

The −62% government number is a measurement artifact, not a collapse

The −62.38% government average is real arithmetic, but it is not a description of how public hospitals operate. It is the footprint of a small group of facilities that the Medicare cost report is the wrong instrument to measure.

Eighty-four government hospitals report an operating margin below −100%, with the most extreme running into the thousands of percent negative. They are overwhelmingly standalone state psychiatric hospitals — institutions funded almost entirely by state appropriations rather than patient billing. These are not failing businesses, and the reason traces to a specific federal rule.

A Medicare cost report records these hospitals' full operating expenses but almost none of their revenue, because Medicare net patient revenue — the denominator of the margin — is a sliver of what actually funds them. The result is a margin of negative several thousand percent that says nothing about financial health. Stripping those 84 facilities' influence by reading the median (−12.30%) instead of the mean gives the honest answer: government hospitals are genuinely the weakest of the three classes on patient-care operations, but they are not collapsing at −62%.

Revenue-weighted, every class is profitable — but the order never changes

There is one more lens, and it is the one a financial analyst would reach for. Instead of averaging each hospital equally, weight by dollars: total net income over total net patient revenue, per class. This answers "across all the patient-care dollars in this ownership class, what is the margin?"

On that revenue-weighted basis, every class is in the black. For-profit hospitals earn 16.39%, nonprofits 9.02%, and government hospitals 7.55% across $168.5B, $1,092.3B, and $237.8B of net patient revenue respectively. The large, well-capitalized hospitals that hold most of the dollars are profitable in all three classes; it is the long tail of small facilities — disproportionately rural and public — that drags the unweighted averages negative.

That is the most important honesty point in this study: the for-profit → nonprofit → government ranking is identical under all four measures — average margin, median margin, distress rate, and revenue-weighted margin. The levels move enormously depending on the lens, but the order never does. For-profit hospitals are reliably the most profitable ownership class and government hospitals reliably the least, and that conclusion does not depend on a single fragile statistic.

Where for-profit hospitals concentrate

Ownership is geographic. For-profit hospitals are the majority in a cluster of Sun Belt and Mountain West states: Nevada (62.3% for-profit), Louisiana (57.7%), Puerto Rico (57.4%), Texas (55.8%), and Florida (53.9%) all run majority-proprietary among hospitals with 40 or more reporting facilities. The Northeast and Upper Midwest remain overwhelmingly nonprofit. This is the legacy of state certificate-of-need regimes, Medicaid-expansion choices, and decades of for-profit chain acquisition concentrating where commercial payer mixes are strongest — the same selection that shows up in the margins.

Why ownership predicts margin: payer mix, service lines, and the safety-net role

The margin gap is structural, not a story about management talent. Three forces explain almost all of it.

Medicare pays everyone below cost, so the gap is in the commercial book. MedPAC's analysis puts the Medicare margin at −12.7% in FY2022, and around −3% even for hospitals it judges relatively efficient. Because every ownership class loses money on Medicare, the difference between classes comes almost entirely from their commercial patients — and for-profit systems carry the best-insured patient mix.

For-profits select markets and service lines. They site in better-insured suburbs, carry lighter Medicaid exposure, and emphasize higher-margin elective and surgical lines while avoiding the low-margin services — trauma, burn units, inpatient behavioral health — that anchor public hospitals. Government and many nonprofit hospitals are obligated to keep those services regardless of margin.

Nonprofits trade margin for a tax exemption — and that bargain is contested. Nonprofit hospitals are exempt from income tax in exchange for community benefit, reported on IRS Form 990 Schedule H. KFF estimates the exemption was worth about $28 billion in 2020, more than the roughly $16 billion of charity care nonprofits delivered that year — the core of an ongoing Senate Finance Committee and ProPublica critique. The American Hospital Association counters that total community benefit averages 15.1% of hospital expenses, and that the exemption underwrites cheaper tax-exempt bond financing. Either way, a nonprofit's thin operating margin is partly by design: surplus is reinvested or spent on community benefit rather than booked as profit.

Read together, the three forces say the same thing the data does. For-profit hospitals are the most profitable class because the system lets them choose the profitable parts of it, while government and nonprofit hospitals absorb the rest.

How this differs from our hospital-margin snapshot

This study is the ownership cut. Our companion study, The hospital margin gap: a national snapshot, reports the overall margin distribution without splitting by who owns the hospital, and Hospitals running on fumes: the days-cash-on-hand crisis measures liquidity rather than profitability. The closure-risk lens lives in Rural hospital closures, by the numbers, which slices the same HCRIS margins by rural Critical Access status instead of ownership. All four draw on the same CCN-keyed cost-report backbone; this one is the only one that answers "does ownership predict the margin?"

Methodology

Every figure is a direct aggregation over two HCRIS-derived tables joined on the CMS Certification Number (CCN): hospital_directory (ownership type, from form CMS-2552-10 Worksheet S-2 Part I "Type of Control") and hcris_facility_summary (operating margin, total margin, net patient revenue, net income). Both tables descend from the same cost report (HOSP10FY2024.ZIP), so ownership and margin are internally consistent — there is no cross-source entity-matching risk. All 6,019 directory rows join 1:1 to a summary row; margin statistics are computed over the rows with a non-null operating_margin_pct (5,787 nationally). Operating margin is (net patient revenue − total operating expense) / net patient revenue × 100, as published in the cost report. Class averages, medians, positive-margin shares, and the −5% distress rate use the per-facility margin; the revenue-weighted margin uses sum(net_income) / sum(net_patient_revenue) per class. The exact query is in the reproducibility block below and resolves every headline number to specific rows in a specific frozen federal snapshot. Methodology version: hospital-ownership-margins/v1.

Limitations

• Cost-report margins are not GAAP audited financials. They are computed from the Medicare cost report and reflect the Medicare-cost-reporting view of revenue and expense, not a hospital's full audited income statement.
• State psychiatric hospitals distort the government average. As shown above, 84 appropriation-funded public facilities push the government mean to −62.38%. The median (−12.30%), distress rate (67.6%), and revenue-weighted margin (7.55%) are the robust government figures; the mean should always be read with this caveat.
• Operating margin excludes investment income. It is the right metric for comparing operations across ownership types, but it understates the total financial position of large nonprofits with endowments, whose total margins run higher.
• Ownership is a three-way cost-report flag. "Type of Control" does not distinguish, for example, a small physician-owned for-profit from a national investor-owned chain, or a county hospital from a federal facility. The classes are broad.
• Aggregate-only, and not a quality measure. Margins are a financial signal, never a measure of care quality. Every figure is an ownership-class or state-level aggregate; no individual hospital is named, ranked, or scored, and a negative margin is a pressure indicator, not a prediction of closure.

Sources

• MedPAC — March 2024 Report to Congress: Medicare Payment Policy — Medicare and all-payer hospital margins by ownership; the −12.7% FY2022 Medicare margin.
• AHA — "Many hospitals continue to face significant financial challenges" (MedPAC, March 2024) — the quotable Medicare-margin figures and the rating-agency negative outlook.
• KFF — The estimated value of tax exemption for nonprofit hospitals was about $28 billion in 2020 — the exemption-vs-charity-care gap behind the nonprofit-margin debate.
• AHA — Fact sheet: nonprofit hospitals' tax-exempt status (January 2025) — the industry response: 15.1%-of-expenses community benefit and tax-exempt bond financing.
• Manhattan Institute — US psychiatric hospitals under Medicaid's IMD exclusion — confirms state psychiatric hospitals are state-appropriation funded, the artifact behind the government average.