Equally important to Hendren’s argument is the idea that sufferers of heart disease and cancer have greater self-knowledge than healthy people in terms of what their likely medical care costs will be. The market for insurance unravels, in Hendren’s model, when patients have a clear view of their future health care costs and people who anticipate lower-cost futures self-select out of insurance coverage.
Hendren tests his assumption using the Health and Retirement Study, which has surveyed Americans 55 and older since 1992. The HRS asks a battery of questions about participants’ futures, including the probability that they’ll end up in a nursing home, be disabled, or dead sometime in the next decade. Crucially for Hendren, the survey also asks about existing medical conditions, so it’s possible to sort HRS respondents based on whether they have conditions which, according to insurance underwriting guidelines, would be grounds for rejection. For example, those with strokes or previous bouts of home care can’t get long-term care insurance that would cover nursing homes; people with back conditions or obesity are ruled out of disability insurance; and stroke and cancer victims aren’t eligible for life insurance.
Hendren then examines whether those suffering from illnesses that freeze them out of insurance markets are better at predicting their medical futures. Across all three markets, he finds this to be the case. On the prospect of nursing homes, insurable respondents’ predictions are no better than random guesses, after accounting for age, gender, and other things that an insurance company can use to calculate customer risk. They’re not much better at predicting future disability or the arrival of the Grim Reaper. By contrast, across all categories, predictions made by uninsurable respondents do much better than random, and always out-predict their insurable counterparts by significant margins. This backs up Hendren’s theory that the reason insurers won’t cover patients with pre-existing conditions isn’t that they’re too sick—it’s that they’re too knowledgeable about their likely future medical costs.
If a voluntary market for insurance for pre-existing conditions is doomed to unravel, what’s to be done to accomplish the Obama administration’s goal of accessible health care for all? For the time being, there’s a government-run Pre-Existing Condition Insurance Plan that covers those denied coverage in the recent past. But that program’s history hints at the enormous—and unexpected—costs that come from insuring people who have already had cancer, heart attacks, and strokes. The cost per participant of PCIP is projected to be nearly $30,000 in 2012, more than double what government actuaries projected. This is exactly what Hendren’s model would have predicted: Only the highest-cost cases choose to purchase the insurance.
Come 2014, the Affordable Care Act will prevent insurers from discriminating based on pre-existing conditions: cancer victims and stroke survivors will be able to buy insurance at the same price as otherwise similar applicants. Insurance companies may take a hit to profits, but part of the cost will surely be passed on to the lower-cost counterparts to this high-cost pool. Healthy people might be tempted to opt out, but under the new law, they’ll be required to have insurance. This individual mandate is a natural fix to the problem of adverse selection in health insurance: It keeps the lowest-cost participants from opting out, and as a result the market doesn’t unravel.
Perhaps unsurprisingly, these solutions rankle the likes of Ron Paul and other libertarians who see the heavy hand of government at work here. More thoughtful alternative proposals from the free-market-is-best school of thought suggest creating a market for the right to buy insurance in the future: That way, you could enjoy your individual liberty by not buying health coverage today, but still keep open the possibility of exercising the option to buy insurance in the future. This “forward contract” to buy insurance wouldn’t be undermined by the problems Hendren highlights, since purchasers would still be making the decision before they experience the stroke or heart attack that gives them inside knowledge on what their future costs will be.
But that’s asking for an awful lot of foresight for the average 20-year-old, who doesn’t really have much reason to think about his likely health status a few decades in the future. And indeed, the least forward-looking are also probably those that can’t or don’t buy insurance.
You can say that’s their own fault—if insurance contracts are available and people don’t buy them, that’s a choice and they should deal with the consequences. Or we can accept that there are some situations—health care being one of them—where the market doesn’t know best, and the guiding hand of government needs to step in to ensure fair treatment for all.