Address to HBS Health/YPO-WPO
Health Reform at the Intersection of Government and Markets
May 4, 2009
The leading Democratic health care reform proposals share a common feature. They all rely on greater regulation to make health insurance more accessible and more secure. The plan on which President Obama campaigned, Senate Finance Committee chairman Max Baucus' "white paper," and Sen. Ron Wyden's "Healthy Americans Act" would mandate that Americans purchase coverage, would dictate its terms, and would dictate its price.
They take this approach because the prevailing view in health policy – and certainly in the Democratic party – is that health-insurance markets fail. Absent regulation, health-insurance markets will not provide secure coverage, making regulation necessary.
What I want to argue today is that the prevailing view ignores a growing body of economic literature suggests that lightly regulated health insurance markets perform a lot better than most people think. If we charge headlong into more regulation without taking that literature seriously, we could end up making ourselves less secure.
Critics lodge two principal complaints against private health insurance: sick people can't get it, and once people become sick private health insurance often disappears. Interestingly, the main cause of both problems is not market failure but government failure.
Initially through wage and price controls and subsequently through tax policy, the federal government created our current employment-based health insurance system. Princeton's Uwe Reinhardt muses that only the devil could have designed a system so perverse that you lose your health insurance at the very moment you get sick and lose your job. At that point, of course, most people cannot afford private health insurance. Not only do they have no income, but their premiums are both sky-high and no longer tax-exempt. Wharton's Mark Pauly found that high-cost individuals covered by small employers are nearly twice as likely to end up uninsured as similar individuals who purchased coverage directly from an insurer. All five of the uninsured Ground-Zero rescue workers whom Michael Moore took to Cuba for medical care had health insurance on September 11th, but lost their insurance when they lost their jobs. Yet Moore – and even many responsible critics – blame those failures on the market rather than government.
Moreover, a growing body of research suggests that lightly regulated health-insurance markets do a better job of pooling risks and providing secure coverage than critics suggest. Some scholars argue that health-insurance markets could do even more amazing things if government would just get out of the way.
Susan Marquis and her colleagues at the RAND Corporation found that in California's individual health-insurance market, there are few regulations and yet "a large number of people with health problems do obtain coverage."
Marquis and her colleagues confirm what Mark Pauly and his colleagues have found working separately, which is that premiums in the individual market do not vary much with risk. The variation in premiums, Pauly finds, reflects only 15 percent of the variation in risk. Why? The answer appears to be that when people purchase health insurance, yes, they are charged a premium based on their risk. When their health status worsens, however, their premiums do not rise to reflect that change. Their premiums rise along with the average of the group. Pauly concludes that markets therefore achieve most of the risk averaging that regulation seeks to achieve — yet without the added burden.
Many critics object that there is any underwriting at all. Yet underwriting performs a crucial function in health insurance markets.
Health insurance companies are essentially intermediaries between members of the pool. They charge higher premiums to enrollees who purchase more extensive coverage, because those members will draw more money from the pool. They require members to pay part of the cost of their own medical care (through deductibles, coinsurance, and copayments) to ensure that members aren't careless with other members' money. They look over physicians' shoulders to ensure physicians are being careful with their members' money. And they calibrate each new member's premium to her expected claims. If an individual waits until she is sick to join the pool, her premiums will therefore be much higher than if she joined while healthy. Risk-rating actually promotes compassionate behavior by encouraging people to contribute to the pool while they are still healthy, so their premiums can help save the lives of strangers.
Unfortunately, there will always be some people who did not obtain health insurance before they became sick.
Assuming they cannot afford medical care, individuals with expensive preexisting conditions require subsidies, which is not to say they need insurance. Insurance is merely one way—and actually, a very expensive way—of subsidizing preexisting conditions. More than other types of subsidies, insurance resembles a blank check. In general, strangers do not voluntarily give blank checks to other strangers, again with good reason: strangers are difficult to monitor, and the beneficiaries (encouraged by their health care providers) may take more than they need. Other ways of subsidizing the needy include limited amounts of cash, vouchers, or in-kind subsidies from providers, private charities, or government.
Risk-rating and exclusions for preexisting conditions do not indicate a lack of compassion by markets or insurance companies or consumers. They are the insurance market's way of telling us that, compared with the alternatives, the added costs of subsidizing preexisting conditions with insurance outweigh the added benefits. They are the market's way of telling us that consumers do not want to subsidize people with preexisting conditions through insurance. They do not preclude other options for subsidizing the needy.
Moreover, Mark Pauly's research suggests that banning risk-rating and exclusions for preexisting conditions accomplishes little. Rating restrictions do little to increase pooling, because the market achieves so much risk averaging on its own. Yet rating restrictions do increase the number of uninsured by discouraging the healthy from purchasing insurance.
So let's stipulate for the moment that risk rating, while painful in some instances, performs an important function. And let's say that markets do a good job of risk averaging once people have purchased insurance.
Critics lodge one more complaint against private insurance markets, which is that once you become sick, an insurers will drop you because you're costing them a lot of money but you're only paying an average premium, and/or the healthy people in your pool will drop out because they can find a lower premium elsewhere. Again, Mark Pauly provides a response, with a little help from his colleagues including John Cochrane of the University of Chicago.
John Cochrane offers this illustration. Suppose you're a healthy 25-year-old male and your expected medical expenses over the next year are $2,000. A health insurer will charge you that much to cover your medical expenses for the next year. So far, health insurance looks a lot like auto insurance.
But, of course, health insurance is a little more complicated. Suppose further that there's also a 1-percent chance that you will contract a long-term, costly illness this year that would make your premium rise to $10,000 next year and all subsequent years. To protect you against the risk that your premiums would jump by $8,000 per year, assuming a 5-percent interest rate, your insurer would need to collect an additional $150,000. That comes to just $1,500 per enrollee. So for a total premium of $3,500, you are covered against the risk of needing medical care this year and the risk that a high-cost condition will cause your premiums to jump.
In the early 1990s, Pauly and his colleagues theorized that if insurers "front-loaded" premiums this way, neither the insurer nor the healthy people in the pool would have an incentive to defect. The insurer would have collected all the money it needs to cover those long-term costs, and in a competitive market, the force of reputation would prevent insurers from trying to run sick people out of their pools. The healthy people in the pool would have no incentive to leave because they are not being asked to subsidize people with known medical conditions. They already subsidized those folks in the previous period, before those conditions were known. In 2006, Pauly published a paper where he essentially expressed his surprise that insurers had been front-loading premiums that way all along.
Still, some critics say the forces of competition are not enough to keep insurers from reneging on their commitments to the sick.
Enter John Cochrane. Cochrane is a Booth School finance professor who argues that a less-regulated market – where government neither favors job-based coverage nor regulates premiums – would develop even greater assurances that insurers would not defect.
For example, Cochrane argues that health insurers could "settle up" with their sick customers at the end of the year. In essence, if you contract a costly, long-term illness, then your insurer would put that $150,000 – the present value of the $8,000 increase in your health premiums – in a special trust account that you control but that can only be used toward health premiums. The insurer cannot defect if he's already settled your claim and cut you a check. And you would have complete freedom either to stay with that insurer or to choose another that treats you better.
After Cochrane had written a paper for the Cato Institute about this concept – but before we had published it – the market took a large leap in Cochrane's direction. UnitedHealth Group announced its Continuity product, which allows people to buy the right to purchase health insurance in the future at standard rates, regardless of their future health status.
I'm not here to tell you that you have to agree with Marquis or Pauly or Cochrane about how lightly regulated markets already perform, or how less-regulated markets would perform. But here's why I think we all need to take them seriously.
The leading Democratic proposals try to cover the sick and make coverage more secure by mandating that all Americans purchase health insurance and by restricting risk-rating. Those measures have some pretty undesirable side effects. They increase moral hazard, health care spending, and thus the overall cost of coverage. They narrow the range of health insurance choices by eliminating economical as well as comprehensive plans. They encourage insurers to avoid and provide lousy service to the sick. And the enforcement costs – of monitoring people to make sure they have health insurance or to prevent insurers from competing to avoid the sick – are considerable.
If the research findings of Marquis and Pauly and their colleagues accurately reflect reality, then those regulations will impose considerable costs for almost no benefit. And if Cochrane is correct, then we could have more secure health insurance in a market where insurers compete to serve the sick, rather than to avoid them.