Researchers at USC recently published a study designed to find out how much people are willing to pay for better drug coverage from their health insurance plan. The question they posed to the general public was straightforward: How much extra money would you pay per month for a health insurance plan that would pay for “specialty drugs” if you need them?
Specialty drugs are expensive new treatments for diseases like leukemia, multiple sclerosis and rheumatoid arthritis. These drugs often cost tens of thousands of dollars, and in some cases even run into six figures per patient. But these high costs can be accompanied by significant benefit. Gleevec for example can dramatically increase life expectancy for people with otherwise fatal leukemia.
Keep in mind that not only are specialty drugs expensive but they are being used with increasing frequency. According to the USC team, 3 out of 100 people in the United States will use at least one specialty drug in the following year.
How much would you pay to make sure you aren’t responsible to pay for these drugs out of pocket? Would you be willing to give your insurance company an extra $5 per month? $10? Maybe even $20?
The USC team found that, on average, people were willing to spend around $13 extra per month to make sure their health insurance plans cover such specialty drugs. (The study was published in the April issue of Health Affairs, and was led by John Romney.) To put that into perspective, the actuarial cost of such coverage—how much insurance companies would expect to spend per person if everyone obtained such coverage—is around $5 per month.
Sounds like a win/win/win situation. Win #1: Insurance enrollees could fork over, say, $10 a month and get coverage they value at a higher price. Win #2: That $10 would leave insurance companies with enough money to profit from this expanded coverage. Win #3: Pharmaceutical companies would be happy, because more people would have affordable access to their products.
But this deceptively simple survey is, in my view, deceptive in its simplicity. My background in behavioral economics has taught me that if the USC team wanted to obtain a high estimate of the public value of specialty drugs, they could not have found a better way to do so.
The problem began with the way they asked their willingness to pay question. To illustrate the problem, imagine the following situation. I gather together ten groups of people. I ask the first group how much money they would give to a charity that would save the lives of thousands of children in Africa. I ask the second group how much money they would give to a charity that would promote clean water in their communities. I ask other groups about other causes, supporting children’s healthcare, rebuilding local churches, etc. Let’s suppose, on average, that people are willing to give $50 to whichever charity I mention to them.
Do you think if I gathered an eleventh group of people and described all ten charities, that they’d be willing to give $50 to each of the ten charities? That they’d be happy to fork over $500 additional dollars for good causes?
Not even close.
In fact, studies in behavioral economics have consistently shown that people overweigh the ideas that have been made most prominent in their minds. Ask people to think only about specialty drugs, and they will be glad to give $13 a month to make sure they’re covered by insurance. Ask them about any one of a dozen things that aren’t fully covered by health insurance presently—long term care, for example, or co-pays for physical therapy appointments—and they might fork over $13 for any of these services. Describe all dozen such things to people however, and let’s see—12 items, each worth $13 per month, . . . that would add up to an extra $156 per month. Do you think they’d be willing to pay that much more for broader health insurance coverage?
How did an accomplished research team from an elite university fall prey to such a basic survey blunder? A cynic might assert that the researchers, who were funded by Bristol-Myers Squibb, looked for a way to get the kind of answer that their funder wanted to see. But I don’t share such skepticism. I favor an alternative explanation.
The research team was made up of people trained in what you might call traditional economic methods. Perhaps this team is not thoroughly knowledgeable about what behavioral economists have learned about willingness to pay measurement. Or perhaps they simply do not believe the behavioral economic findings. In academia, you see, most of us are prone to intellectual myopia. We know our own field, and barely have time to recognize (much less appreciate) anyone else’s research. Traditional economists have long believed that willingness to pay questions yield useful data. Many of the flaws in willingness to pay methodology have been uncovered by folks who don’t “hang out” in traditional economics circles.
But as good scientists, or as good citizens, we need to see beyond our own biases. The USC team was addressing a very important issue. As a society, we need to figure out, as they set out to do, what kind of insurance to make available to people so they can spend their money in ways that fit their own tolerance for financial risk.
But we don’t come close to figuring this out by focusing people on one single class of healthcare services, as if their willingness to pay for such things in isolation reflects the true value such interventions hold for them.
We need to measure public willingness to pay in a manner informed by the insights of behavioral economics.
Peter Ubel is a physician, behavioral scientist and author of Pricing Life: Why It’s Time for Health Care Rationing and Free Market Madness. He teaches business and public policy at Duke University. Peter’s new book, Critical Decisions will be available in the fall of 2012. You can follow him on his personal blog.