File 04 · Long-term care · 2000–2026
Nursing homes
Nursing homes are paid largely by Medicare and Medicaid, housing people who cannot leave if the care thins out. That is a dangerous place to put a leveraged owner. A study by Atul Gupta, Sabrina Howell, Constantine Yannelis, and Abhinav Gupta followed private-equity deals against millions of Medicare patients and found a grim, specific result: short-term mortality up about 10%, on the order of 20,150 lives over twelve years, while taxpayer spending per episode rose and nursing staff declined. Costs shifted toward the items that pay a fund — monitoring fees, interest, and lease payments. The residents did not share in the multiple.
The economists’ phrase for where the money went after buyouts: a shift toward monitoring fees, interest, and lease payments — away from the patient.
Findings
- Mortality
- +10%
- The bill to Medicare
- +11%
- Staffing
- Fewer nurses
PE ownership increased the short-term mortality of Medicare patients by 10% in the NBER study, about 20,150 lives over the twelve-year sample. The authors also estimated on the order of 160,000 life-years lost.
Taxpayer spending per patient episode rose about 11% in the working paper’s abstract. Net income of the facilities did not simply boom — the surplus was rerouted into fees, debt, and rent.
The same research found declines in nurse availability and in compliance with standards. A later federal analysis, cited in Sen. Murphy’s 2026 report, associated PE investment with about 12% fewer registered-nurse hours per resident day and a higher deficiency score.
The playbook, in this industry
- 01
Separate the real estate from the care
The property is sold, often to a related party, and leased back. The operating company that employs the aides looks asset-light and lawsuit-poor. The rent is senior to staffing.
- 02
Thin the floor
Certified nursing assistants and registered nurses are the cost. Residents feel it as unanswered call lights, missed turns that prevent bedsores, and less time for eating and walking.
- 03
Bill the public harder
Medicare pays for post-acute stays. The study found spending up even as frontline care intensity fell. The public pays more for a worse stay.
- 04
Use chemical quiet
The researchers found a sharp increase in antipsychotic use — drugs discouraged in older people because they are associated with higher mortality. It is a grim substitute for staff.
Incidents
2000–17
The study window
Gupta and colleagues match about 7.4 million Medicare patients to buyouts: 1,674 homes acquired in 128 deals, inside a universe of more than 18,000 homes. This is not a handful of horror stories. It is the portfolio.
2007
Carlyle buys ManorCare
HCR ManorCare, one of the largest chains, is taken private by Carlyle in a debt-heavy deal, then put through the real-estate split common in the industry. The chain later goes bankrupt, in 2018.
2020–22
COVID makes the staffing visible
The pandemic kills nursing-home residents at a scale the country could not ignore. Homes that had already cut staff had less slack. The White House, in 2022, names private equity’s nursing-home buying directly.
2023–26
Regulators count the hours
Federal staffing analysis and Senate reports keep finding the same operational signature: fewer nurse hours, more deficiencies, financial costs pointed upstairs.
Case files
1,674 homes, twelve years
What the national study actually counted
The paper is careful in a way anecdotes are not. It tries to separate the homes private equity chose from the effect of being owned, and to separate which patients ended up there. After those corrections, death during the stay and the following 90 days is higher by about 1.7 percentage points — 10% above the baseline. Mobility worsens. Antipsychotic use rises about 50% relative to the prior rate in the authors’ economic summary. Nurse time per patient falls. This is the file’s center of gravity: not a theory of ‘greed,’ a measured change in whether people survive a stay their family thought was rehab.
The public bill. About 20,150 additional Medicare deaths in the sample window, against spending that went up, not down.
A national chain, a single sponsor
ManorCare
Carlyle’s purchase of HCR ManorCare is the company-scale version of the pattern the paper measures across many smaller deals. A huge chain, a leveraged buyout, a real-estate separation, and a 2018 bankruptcy. Residents do not move easily when a chain fails. Their ‘choice’ was often the home with an open Medicaid bed. Bankruptcy court is a strange place for a promise about bathing, meals, and night shift.
The public bill. A flagship chain in bankruptcy, and a template widely copied.
What the public lost
Cost
- Medicare spending per episode up about 11% under PE in the NBER study
- Family out-of-pocket charges and ancillary fees on top of public payment
- Lease and monitoring fees that do not show up on a brochure
Access
- Lower nurse hours per resident
- Higher deficiency scores — the regulatory record of missed care
- Homes that remain ‘open’ while the night shift cannot do the work the admission assumed
The owners’ argument
Operators argue that Medicaid rates are too low to staff well, that private equity invested in failing homes, and that mortality reflects the patients, not the owner. The Gupta study’s design is an attempt to answer exactly that objection, comparing within facilities and instrumenting for who is admitted. Low Medicaid rates are a real problem. They are not a reason to add interest, rent to a related landlord, and a monitoring fee, then cut the nurse.