For years, Medicare’s typical payment arrangements, or the terms under which facilities and other healthcare providers get reimbursed by Medicare for the services they provide, have been the subject of debate. Traditional payment arrangements use a fee-for-service approach. One service, one fee. Another service, another fee. Some critics say this rewards volume: If providers do more, they get paid more. Economists note this response is pretty standard supply-side behavior. The problem that might arise is if providers did more when payments are cut, which would mean that efforts to control spending by fee cuts might not work.
But how do providers respond to fee changes? To know the true effects of paying providers more, researchers would ideally like to design an experiment. Some providers in some places get paid more, while other providers in other places get paid less: do the providers in those places then act differently, and how? But since Medicare is a national program that follows well-known formulae for payments, it’s actually very rare that such experiments present themselves.
In an article appearing in the June 2020 issue in Health Economics, W&M professors Peter McHenry and Jennifer Mellor, along with Swarthmore professor Daifeng He, identified such an experiment. A seemingly random policy change in the complex formula that Medicare uses to pay skilled nursing facilities took place, and it increased the payment rates for some facilities, while decreasing the rates for others. These changes created a “natural experiment” with exactly the type of variation that policy analysts crave.
The study focused on skilled nursing facilities, which are facilities that provide nursing and rehabilitative services to Medicare beneficiaries following hospital stays. Spending on skilled nursing facility care has been one of the fastest growing categories of Medicare spending over the past few decades. The research team used seven years of data from more than 12,000 nursing facilities nationwide. They found that in facilities that received higher payments, the number of days of skilled nursing care provided to Medicare residents increased. And while Medicare payment increases affected Medicare resident days, payment decreases did not, suggesting that providers have some reluctance to reduce the amount of care provided.
The results mean that facilities are very nearly following the standard economic model of how sellers behave. “What’s useful about this right now is that CMS recently bumped up their payment rates to skilled nursing facilities in rural areas,” said Peter McHenry. “Based on our results, we might expect that those facilities would be in a position to deliver more skilled nursing services to their residents. We didn’t see any evidence that facilities did more to undermine the effects of price cuts.”The study was funded by the Agency for Healthcare Research and Quality, and is the lead article in the June 2020 issue of Health Economics. You can read it here.