The EPA's 'cancer premium' shows how fear can trump real risks
Imagine you are told you will contract either of two fatal illnesses, cancer or heart disease. Which way would you choose to die?
Most people opt for heart disease. Understandably. They believe cancer brings greater pain and suffering, and while we can control the risk of heart disease, we don't feel we can do much to reduce the risk of cancer. So it's scarier.
But death by heart disease is far more likely. The illness is the leading killer in the United States. In 2007, the last year for which final data are available, heart disease killed 616,000 Americans. Cancer was second, killing 563,000. And heart disease kills people across a broader range of demographics.
Wouldn't it make more sense to be more afraid of what's more likely to kill you?
Yes, but that's not how we perceive risks. Risks have psychological and emotional characteristics that make some feel scarier than others, the probabilities notwithstanding. A long, painful cancer death may not be any worse than a long, painful heart disease death, but we think it would be, and feel we can't control it, and that makes cancer more feared.
That is precisely what a new proposal at the Environmental Protection Agency is trying to acknowledge. When assessing whether a new regulation would be worth the money, the agency projects how many lives it would save vs. the costs of implementing it. But now, the EPA suggests that death by cancer is so frightening to the public, cancer deaths should carry greater weight in its calculations than deaths by other causes.
This is an approach that may have some ethical and emotional appeal, but it carries serious dangers for us all.
As part of the cost-benefit analysis required for all regulations whose impact on the economy would be at least $100 million, economists developed a system that places a dollar amount on lives saved, to make an apples-to-apples comparison to the cost of the rules and regulations. It's called the value of a statistical life, or VSL.
VSL is economically rational to some, emotionally repugnant to others and poorly understood by most. Here's how it works:
People who would be affected by the regulation are asked in "willingness to pay" surveys what they would spend to reduce the risk of dying by, say, 1 in 100 or 1 in 10,000. If people in the affected population are willing to pay an average of $100 for a risk reduction of 1 in 10,000, and there are, say, 10 million people in the population, then the population as a whole is willing to pay $1 billion for that risk reduction. Out of the affected population of 10 million, a 1-in-10,000 risk reduction will save 1,000 lives. So each of those lives is worth $1 million - the $1 billion everyone's willing to pay divided by 1,000 lives saved. (You can relax. The math part is done now.)
These aren't real people with faces and arms and legs. These are statistical, abstract lives. VSL is just a mathematical way to convert people's willingness to pay for safety into a quantitative measure for rationally comparing the costs and benefits of a regulation.
But while it may be scientific, VSL doesn't feel, well, human, because the psychology of risk perception involves more than math and probabilities. You're dead at the end, no matter how you die, but not all deaths are the same. How scary a fatality risk feels depends not just on whether you might die, but how.