Under the Affordable Care Act, insurers can no longer deny or price medical coverage based on pre-existing conditions. This offers enormous benefits to those who could not obtain or afford adequate coverage—and potentially enormous risks to health insurance markets.
In an informal talk “that raised more questions than answers,” Becker Friedman Distinguished Fellow Pierre-André Chiappori showed that we don’t know the ramifications of this change. And we don’t have the data to estimate them.
By eliminating risk-based price differences, the ACA shifts demand for insurance. “People who know they are sick want to buy a lot of insurance, but the employer is not allowed to price differently,” Chiappori said.
This introduces massive amounts of adverse selection—an insurance pool filled with high-risk, high-cost individuals. In the worst-case scenario, skyrocketing payouts and regulated prices might cause health insurance markets to collapse.
If this seems far-fetched, Chiappori asked the audience to consider a situation where a $1 million term insurance policy costs the same for a healthy 25-year-old or someone with a month to live. The terminally ill would snap up policies, and the insurance company would quickly go bankrupt paying off claims.
With the ACA, “We’re talking about a large scale reform, and the damage can be huge. Crucial issues seem to have been underestimated. What’s very surprising is the lack of precise empirical studies” underlying the reform, he noted.
Chiappori said we lack the necessary data on joint distribution of risk, wealth, and risk aversion to calculate the costs and impact of these changes. We also have no data on the nature of existing insurance contracts.
“That’s shocking. The ACA is aimed at solving a problem we can’t even measure."
Chiappori, a premiere theorist in economics of the family, applied his expertise in contract theory to raise many unanswered questions of the ACA.
The primary motivation for health care reform was to address the 50 million Americans without any health care coverage. That presumes the remaining 250 million are covered. “This is not the case. Plenty of people are statistically ‘covered,’ but if you look closer, they are very far from being anything like what we would call sufficiently covered. We don’t know how many are in this situation.”
Furthermore, he continued, “What we mean by health coverage is ambiguous. If you have a policy with a $10,000 cap, is that actually coverage? Are you actually ‘covered’ if you get cancer and your employer can lay you off?”
Chiappori pointed out that competitive insurance markets require price differentiation for coverage based on individual risk. That includes two components: the risk of actually developing a disease, and the risk or likelihood of being classified as high-risk.
Bypass surgery, for instance, is likely to result in reclassification as bad insurance risk. With a genetic disease like Huntington’s disease, the risks of carrying the gene and developing the disease are well known and don’t change over time.
Is it ethical and fair to charge a high premium to those predisposed for a disease? Discrimination against those with a predisposition for a genetic disease is now generally prohibited, but we don’t really know the long-term consequences of that, Chiappori said.
“In a situation where a small share of the population has an expensive genetic disease, the market remains robust," he explained. “But now the cost of doing complete genetic analysis is low, and we know a lot about the genetic contributions to many other diseases.” As our knowledge grows, so does the pool of people who carry genetic risk.
Many people would presumably like to buy insurance against the “cancer then layoff” scenario. That protection does exist through policies with guaranteed renewal when purchased directly—but not when employer-provided.
“Why does the market provide a solution when insurance is purchased directly but not when employer-provided? What is the welfare cost of this market failure? Can we estimate it? No! We don’t have the data."
He highlighted a few other potential problems in insurance markets. One is a death spiral. When everyone is charged the same price for coverage, this subsidizes high-risk individuals and effectively taxes the healthy. The healthy then decline to buy insurance, so the costs incurred by the remaining insured pool increase. This requires higher premiums, so more people drop their insurance, further shrinking the pool and raising premiums, and so on until the market collapses.
In a more subtle problem, insurers offer a menu of contracts, and high-risk individuals self-select into high-coverage plans, driving the costs and premiums up.
“Under such systems, in the end we might end up in a situation where the level of price discrimination is not far from what we started with.”
Finally, he noted potentially important labor effects. The Congressional Budget Office recently released an estimate of reduced labor supply, because the ACA reduces incentives to work full time.
Less data is available on how the ACA will affect productivity, although there is some evidence that health coverage increases productivity. “It will affect matching between employees and firms. We don’t want really gifted people to apply to work (unskilled jobs) at a call center because of the great health insurance provided,” he said.
Some data is available from the Massachusetts health care reform, but it’s not completely useful because the commonwealth is not representative of the whole U.S. In response to a question from the audience, he explained that European markets could not provide helpful data for analysis.
A few recent studies from the U.S. provided some insight, but they followed the reforms.
“I would like to think the reform would be designed using such studies, not be the source of them,” he said.
“My main message is that we can go to one extreme, with European coverage, or we can pursue a market solution, which is very complex.
“People think there’s too much regulation; I think there’s not enough. We should regulate contracts and deductibles as well.”