By James Kwak
Driving home from school today, I listened to a Fresh Air interview from two months ago with Atul Gawande, by now perhaps the most famous doctor in the policy intelligentsia. The interview was based on a New Yorker article discussing how some doctors and even some health care payor organizations are trying to reduce health care costs for the most expensive people while improving outcomes. In Camden, New Jersey, one doctor found that one percent of people generate thirty percent of health care costs.
One refrain you heard incessantly during the health care reform debate was that we have high health care costs because of overconsumption and we have overconsumption because people don’t bear a high enough share of their marginal health care costs, so the solution is to increase copays and deductibles. This is what Economics 101 would tell you: people respond to incentives. But Gawande discussed one large company that tried this year after year, but only saw their costs going up. The problem was that while most members responded to the higher copays and kept their costs more or less steady, the 5 percent of members who generated 60 percent of the costs behaved differently. Or, rather, they also reduced consumption (of doctor’s visits and prescription medications), but as a result they often had catastrophic outcomes. These were people with heart disease on cholesterol-lowering medications, and when they went off their medications they ended up in the hospital with heart attacks and then with congestive heart failure.
If incentives worked on this level, we should have solved the problem already. Employers all want to bring health care costs down, so if any insurer could bring health care costs down they would have a competitive advantage, and so insurers should be trying to bring health care costs down. But it’s not working. One explanation is that insurers don’t have enough market power compared to providers (like large hospital chains); I believe Uwe Reinhardt has explained the situation this way.
Another way of looking at the problem is to note that there is no one who is trying to brings costs down directly. Sure, insurers try to do it, but they do it through the types of monetary incentives that economists love: higher copays, lower payments for various procedures, etc. But that’s not actually what most companies do when they have a cost problem. If you run an auto company and it’s costing too much to build a car, you don’t lower the transfer price that you pay to that factory and let incentives solve the problem. You go and figure out what the problem is and you engineer a solution, whether by redesigning the manufacturing process, reengineering the product to use cheaper parts, negotiating lower wage costs, negotiating lower input costs, or something else. That’s how you solve most problems in the business world — not by tweaking some clever incentive scheme.
This is a high-level analogy for what Gawande is talking about: doctors and health care organizations identifying their most expensive patients or members, figuring out what’s wrong with them, and getting them the right treatments. In the few examples that Gawande discusses, it results in cost reductions on the order of 20 percent with better outcomes. It seems that for the people who consume the most health care dollars, you can save money simply by focusing on giving them better care — because right now their big problems are things like coverage gaps that prevent them from getting basic care, not being on the right medications, and ending up in the emergency room for catastrophic problems. Maybe for most people you would not save money simply by providing better care, but for the few people who consume most of the system’s resources, maybe you would save money. The problem is that with few exceptions, no one is trying to do that. That’s what we need an incentive for.