Historically, physicians worked independently of—if in partnership with—hospitals. But increasingly, doctors are selling their practices to hospital systems, as well as private equity firms and insurance companies. What does it mean for patients and what they pay for healthcare when their doctors are employed by a hospital system?
The effects of hospital systems acquiring physician practices are hard to determine, because until now, there has been no comprehensive source of data about these mergers and the effects of integration on pricing can be hard to isolate. New research by Yale SOM economist Fiona Scott Morton tackles these challenges, determining the scale of hospital-physician mergers and examining their effects on pricing. In their study, Scott Morton and her co-authors—Yale economist Zack Cooper; Stuart V. Craig and Ashley T. Swanson of the University of Wisconsin–Madison; Aristotelis Epanomeritakis of Harvard University; Matthew Grennan of Emory University; and Joseph R. Martinez of the University of California, San Francisco—found that these integrations significantly increase prices for consumers, because competition decreases.
To Scott Morton, these findings are a wake-up call for regulators. “States haven’t been enforcing [antitrust laws] to the extent that they could,” she says. “I think there’s a lot of good that could be done with some incremental enforcement.”
Hey, I’m AInsights Ask me questions about the article and its underlying research. Let’s chat! Ask Insights powered by AI ...
Mergers between hospital systems and physicians in private practice are what antitrust experts call “mergers of complements” or “non-horizontal mergers.” These aren’t horizontal mergers, because doctors aren’t competing with hospitals, but they aren’t quite vertical either, because doctors aren’t suppliers to hospitals.
In general, scholars expect mergers of complements to have positive effects. Imagine, for example, a car maker decides it wants to own the software that will go into its vehicles. Rather than keeping its future design plans a secret from its software maker, “I can talk to them, and they can start thinking about what cool software could be used with my hardware. So we could do things together that we can’t do apart,” Scott Morton explains. Non-horizontal mergers can also reduce prices: “If I mark up my car and you mark up your software, then the [buyer] has two markups to pay, whereas if the hardware buys the software and makes a bundle, then they internalize that problem.”
But Scott Morton and her co-authors suspected healthcare might function differently. For example, there’s no reason to believe healthcare quality would improve after integration; after all, between navigating different insurance companies and referring patients to different specialists and hospitals, “doctors are rather good at providing care across corporate boundaries,” she says. And because healthcare is geographically constrained—patients don’t want to travel long distances to see a doctor—regional consolidation could increase prices.
To study hospital–physician mergers more closely, the researchers first had to contend with an informational challenge. “Everybody knew, anecdotally, that these transactions were happening and that there were a lot of them,” Scott Morton says, “but we just couldn’t measure them.” Hospital acquisitions of private practices are generally small enough transactions that hospitals aren’t required to report the purchases to regulators.
So, in the absence of a single comprehensive source of data, the team developed a workaround. Using administrative data from Medicare, hospital surveys, physician directories, and filings with the Securities and Exchange Commission, they trained a machine-learning algorithm to identify when a physician had moved from private practice to a hospital system. The researchers tested the algorithm’s results against verified data about mergers and found it achieved 97% accuracy.
You may say, the increase is only 3%. But this is 3% of a large hospital bill. And healthcare is already 19% of GDP. It’s giant.
... continue reading