Cutting hospital reimbursements didn't slow research. It sped it up and changed its focus.

In 1997, the Balanced Budget Act cut Medicare's payments to the teaching hospitals that anchor US medical research. The expected collapse of that research never came. A new study in Research Policy finds that the most exposed centers, the Academic Medical Centers (AMCs) that combine medical schools, teaching hospitals, and research labs, did more research, not less. Research output at the hardest-hit AMCs rose roughly 6%, and the new hiring tilted toward translational researchers, scientists who move lab findings toward patients.

The paper, distributed earlier as NBER Working Paper 27943, looks at what happened to AMCs after the 1997 Balanced Budget Act (BBA) shrank Medicare's inpatient reimbursements and graduate medical education subsidies. Medicare is the federal health insurance program for Americans 65 and older. Its payments to teaching hospitals underwrite both patient care and the research that happens inside the hospital. Cutting them was supposed to slow the science. It did the opposite at the centers that could least afford to lose them.

The authors, led by Pierre Azoulay, find that research output at the most exposed AMCs rose by roughly 6% after the reform, an estimate reported in the paper's abstract on PubMed and presented with the paper's own caveats about the precision of that figure. The increase is not in spite of the financial shock. It is, the paper argues, a response to it.

The mechanism is composition, not just volume. The Medicare cuts hit the part of AMC budgets that pays for clinical care and residency training. The part that pays for research, much of it flowing through the National Institutes of Health (NIH), the federal agency that funds most basic biomedical research in the United States, was untouched. Hospitals that could not recover lost clinical revenue by cutting research leaned the other way. They redirected hiring toward researchers whose work was most likely to win NIH grants, especially translational researchers who can frame a question in a way that survives peer review and that can be packaged as a clinical-trial pipeline. Incumbent investigators, the established scientists already on faculty, were pushed to expand their research portfolios and publish more.

That tilt matters because translational work is the bridge from bench to bedside. A hospital system that needs NIH money has an incentive to hire people who can produce it, which means a premium on scientists who can turn a molecular finding into a fundable intervention. The 1997 reform, the paper suggests, did not invent that pressure. It sharpened it.

The authors flag a limit on what the result implies. The same paper finds little measurable effect of the BBA on clinical outcomes. Patients at the most exposed hospitals did not get measurably better or worse care in the years that followed. The reform changed what got studied and who got hired. It did not, on this evidence, change how well patients fared.

The lesson is institutional. In US biomedicine, the money that pays for clinical care and the money that pays for research are run by different agencies with different budgets, but they flow into the same institutions. When the clinical money shrinks, the research money becomes more important to the institution, not less. Hiring tilts, grant-writing intensifies, and the kind of science that gets done shifts in the direction of what the remaining funder will pay for.

That dynamic is unlikely to stay in 1997. Every round of Medicare reimbursement reform, every NIH budget fight, and every teaching-hospital merger is, in effect, a vote on which questions American biomedical researchers will spend the next decade chasing.