A Yale economist on taming rising hospital prices while maintaining quality

Over the last two decades, the U.S. hospital sector has seen nearly 2,000 mergers and significant consolidation. During this period, prices for hospital services have grown markedly faster than inflation, driving up health spending and insurance premiums. Concerns about rising health care costs have led to proposals in various state houses and in Congress to regulate these prices.

Policymakers considering hospital price regulation, however, must balance the goal of reducing prices with maintaining the quality of care, which requires a better understanding about the complex relationship between hospital market power, provider prices, and clinical quality.

Yale economist Zack Cooper recently co-authored a National Bureau of Economics Research working paper that tests whether patients get better health outcomes when they are treated at higher-priced hospitals and explores how the relationship between providers’ prices and quality varies in concentrated and unconcentrated hospital markets. The paper was co-authored by John Graves of Vanderbilt University, and Joseph Doyle and Jonathan Gruber of the Massachusetts Institute of Technology.

Cooper, an associate professor of health policy at Yale School of Public Health and of economics in the Faculty of Arts and Sciences, and associate faculty director of the Tobin Center for Economic Policy, recently spoke to Yale News about the research, its implications for public policy — and when these price increases are actually cost effective. The interview has been edited and condensed.

What causes rising hospital prices?

Zack Cooper: The preponderance of evidence shows that the increase in hospital prices we’ve seen over the last 20 years has been driven by the profound increase in consolidation that has occurred in the hospital sector. It really is striking. There have been nearly 2,000 mergers among the approximately 5,000 hospitals in the United States. Unfortunately, we’re at a point right now where most hospital markets are “highly concentrated,” according to the Department Justice and Federal Trade Commission’s Horizontal Merger Guidelines. This is extremely problematic. We must remember that the U.S. hospital sector is 6% of the country’s GDP and the largest contributor to health care costs, so these changes in hospital markets affect all of us.

Why should the public be worried about this consolidation?

Cooper: We have really good evidence that hospital competition raises providers’ quality. We also can see that hospital mergers allow hospitals to raise their prices. So, taken together, the changes we’ve seen in the U.S. hospital sector have meant weaker incentives for hospitals to improve their clinical outcomes and their prices. This isn’t what I think we want for the American public. When hospital prices go up, this raises insurance premiums, and my team and I are really focused now on analyzing how rising insurance premiums are actually leading to pretty big changes in non-health sector employment and inequality.

What should policymakers be thinking about as they consider whether to regulate hospital prices?

Cooper: Economists tend to be reflexively cautious about price regulation. First, it’s really hard to regulate prices effectively — you have to decide on a price level to set and that’s just not easy. Second, it could be that providers’ high prices reflect their quality. The automobile market, for example, is quite competitive and Ferraris have high prices. The high price of a Ferrari isn’t a function of consolidation; their high prices reflect the fact that they’re high-quality, high-cost cars. So, if you regulated the price of Ferraris, we’d all be worried that their quality would fall.

The same thing could be happening in the U.S. hospital sector. High-quality hospitals might be providing higher quality care. Their high prices could reflect patients’ willingness to pay for quality and strategic investments by the hospitals to improve quality. If that’s the case, ill-designed price regulation in the hospital sector could actually lead to reductions in quality. This is why we need to know whether going to a higher priced hospital causally results in better outcomes.

How did you and your colleagues go about studying that relationship?

Cooper: Analyzing the causal relationship between hospital prices and quality is hard because seriously ill patients probably seek out better, higher-priced hospitals. This leads to selection bias. As a researcher, you worry that if you just did a simple analysis and regressed quality on prices, you’d find that hospitals with higher prices have higher mortality rates. Of course, those rates could be higher because the high-priced hospitals treat sicker patients.

This is where some of the tools in econometrics come in. What we do in this study is exploit the fact that in medical emergencies, ambulance companies are randomly assigned to pick up patients and different companies have differing preferences regarding the hospitals where they transport patients. This generates a random assignment in where patients get care, which we can exploit. It’s called an instrumental variable analysis.

Here’s how I like to think about it: If you’re in a big city and open your Uber app, you’ll see a bunch of cars circulating around you. When you hit “confirm,” the car that picks you up, in a sense, is randomly assigned. It’s just a function of which driver happened to be closer to you when you pressed the button. The same thing happens with ambulances, and it turns out that different ambulances will take you to different hospitals. By using ambulance data, we can compare two patients who were collected from the same ZIP code at the same time by randomly assigned ambulances that took them to either a high-priced or low-priced hospital. This allows us to see whether patients who go to high-priced hospitals have better outcomes.

What did you discover?

Cooper: Honestly, the results surprised me. In both concentrated and unconcentrated hospital markets, we found that receiving care from high-priced hospitals increased health spending by approximately 52%, which is quite a lot. What was striking is that when patients went to high-priced hospitals in concentrated markets, their spending went up by 52% but they did not get better outcomes. So this spending was really just wasteful.

Juxtapose that with what we found in unconcentrated or more competitive markets. In these markets, attending a higher-priced hospital led to higher spending, but it also led to a 47% reduction in mortality for time-sensitive conditions like heart attacks. So, in these competitive markets, you spend 52% more, but you’re 47% less likely to die. And what we found is that because of the scale of the gains at high-priced hospitals, their high price increases are likely cost effective.

What does this mean for the policy discourse?

Cooper: To take a step back, we think that rising health care costs are probably the most important issue on the domestic agenda. We know that high hospital prices drive rising health care costs. Our findings directly inform the debate on how to solve this problem.

The solutions it suggests are threefold: First, we need more vigorous antitrust enforcement. We’ve got to stop hospital markets from becoming consolidated and stave off anticompetitive transactions. Our work highlights that competition can lead to pretty sensible pricing.

Second, we probably ought to be cautious about regulating hospital prices in competitive markets. It doesn’t mean that we absolutely shouldn’t do it, just that we should proceed with caution. Our analysis shows that high prices in competitive markets are cost effective, which might strike people as kind of wild. These hospitals are spending $1 million to $4 million more for every life that they save. So even though the steep prices sound crazy, they’re probably justified in some ways given the reduction of death that they produce.

Third, given that our work suggests that high prices in consolidated markets are mostly hospital rents, we ought to seriously consider proposals to regulate hospitals prices in those markets. In fact, there are two bills under consideration in the U.S. House of Representatives that propose regulating prices in concentrated markets. Both are sensible and worth a serious look.

Can an academic paper inform public policy?

Cooper: I hope so. That’s really our goal at the Tobin Center. We want to drive world-class scholarship that’s policy relevant. We also want to get the paper in the hands of policymakers. And that’s what we’re doing here. We’re presenting the work to members of Congress; we’re presenting the work to state officials; we’re presenting the work to federal officials. We want the work to be used and we’re here to help.

We actually took the time to post hospital concentration measures for every hospital in the United States by name. It’s up on our website. You should be concerned about hospitals with a Herfindahl Hirschman Index (HHI) of greater than 4,000 — that’s above the threshold that raises concerns with federal regulators. I’d encourage readers to go look at our data and see where their hospital stands.