Skip to main content

Making a Plan and Not Sticking To It

When sick people change health plans, it can muck up the insurance market. A new paper suggests ways for fixing it.

People switch from one health plan to another all the time. But unlike, say, changing to a different grocer or movie theater, signing up for new health insurance might actually hurt other customers. That's because less healthy people cost insurers more and drive up premiums.

When Congress returns from its recess to resume wrestling with health care reform, legislators will have to overcome that predicament if they hope to cover some of the tens of millions of currently uninsured Americans. A new working paper examines when and why people switch plans and suggests ideas for maintaining insurance options without allowing the market to get out of whack.

For decades, researchers and insurance companies have known that sicker people tend to choose more generous health plans. But experts still struggle to understand why patients make a switch, or how it affects the insurance marketplace — partly because insurers are tight-lipped about premium levels and costs.

To better understand the situation, Harvard Kennedy School of Government professor Richard Zeckhauser and David Cutler, a Harvard applied economist, worked with Bryan Lincoln of the Massachusetts attorney general's office to analyze health insurance in the Bay State.

They looked at the 225,000 covered by its Group Insurance Commission, most of them state employees and their families. The claims data covered the fiscal years 1994 through 2004, before the state instituted a public insurance plan, and showed patients' movement between a generous fee-for-service insurance plan and a less generous health maintenance organization (HMO) plan.

The research team tested for several possible explanations for people's plan decisions.

The first was adverse selection, in which sicker people disproportionately join a generous health plan, driving up an insurer's expenses. The insurer consequently increases premiums, causing healthier people to leave the risk pool. The result is a vicious cycle of ever-increasing premiums for those left in the plan, with healthy folks who might otherwise join remaining in a different plan, rationing their health care or avoiding insurance altogether.

Two other possible mechanisms affecting insurance are adverse retention, where sick people choose not to switch plans so they can remain with current doctors, and aging in place, in which plan members grow older, costing their plans more as their bodies slowly become more susceptible to injuries and illness.

Zeckhauser's group found that people who spend more — a proxy for the less healthy — were more likely to move to the generous plan. They also saw premiums in that fee-for-service plan rise as its composition changed, initiated by adverse selection. There was evidence, too, of aging in place: Younger individuals were 50 percent more likely than older ones to leave the generous plan.

Most people didn't switch during that decade, which shouldn't have been too surprising. Massachusetts heavily subsidizes the government-insured, paying about 85 percent of premiums, blunting the cost difference between the plans and severely limiting adverse selection. The researchers estimated that if the subsidy dropped to 50 percent, only 13 percent of people would remain in the fee-for-service plan; if it decreased to zero, just 8 percent would remain.

In the private market, some large employers, who are essentially self-insuring their workers, use a similar strategy.

That's not an option, however, for the tens of millions of Americans either buying into individual plans or getting benefits by working at small businesses. And if Congress ends up instituting public-option insurance, which may cover as much as one-fifth of Americans, adverse selection will be an even greater concern. The government might solve this problem by encouraging insurers to raise or lower an individual's premium based on his or her risk of getting sick.

"My incentive to enroll in a more generous plan goes away when that more generous plan charges me more based on how sick I am," says William Vogt, a RAND Corporation economist.

Yet such a scheme probably wouldn't go over well with the public because it would force people to pay for conditions over which they have little or no control. "From an economic point of view, that sounds reasonable," says Stanford University medicine and health policy professor Jay Bhattacharya. Politically, it's "potentially very dangerous."

Currently, insurance companies that charge smokers higher premiums remain wary of charging obese people more, partly because of state regulations and partly due to fear of discrimination lawsuits. Instead, they opt for indirect ways to attract healthy, low-risk individuals into their plans, like offering generous maternity benefits.

A more realistic way to address adverse selection, say the authors, is "risk adjustment," a strategy already applied to Medicare and included in several bills on Capitol Hill. The general idea is to pay plans more to take on sicker and older people, reducing the overall premiums for individual insureds. An alternative version might transfer money from the government directly to those with chronic illnesses — diabetes, for instance — so they can afford the high premiums of generous plans.

Another more radical approach, health-status insurance, would actually set up separate insurance plans that kick in after diagnosis of chronic illness. (The author of health-status insurance, John B. Cochrane at the University of Chicago's Booth School of Business, recently addressed the issue of insuring for pre-existing conditions in a Wall Street Journal editorial here.)

Experts are still debating which line of attack and what dollar amounts make the most sense, and the Obama plan (assuming it passes) is likely to leave those decisions up to technical bureaucrats. However, one promising finding of the Massachusetts research is that age and gender predicted plan mobility better than how much people spend on health care. That means the relatively simple measure of spending additional money on plans that enroll older men would make a significant impact.

Yet that's only a partial solution. Correcting for disease-specific risks will be much more complex without patient information that most insurers just don't have.

"I may know your complete medical history," Cutler says, "but I have no idea if your diabetes is well under control or not. It's fundamentally impossible to know as much as the individual about his own health."

Until plans find a way to alleviate that information shortage without spending too much money, adverse selection will continue to challenge plan administrators and add to patient costs.

Sign up for our free e-newsletter.

Are you on Facebook? Become our fan.

Follow us on Twitter.

Add our news to your site.