The primary drawback of the VSL is that it fails to account for future generations’ valuation of benefits and costs.
I appreciate that Professor W. Kip Viscusi took the time to respond to my recent essay in The Regulatory Review, titled “Rethinking the Value of a Statistical Life.” In general, I would invite Viscusi, as well as anyone else who is interested, to read the entire paper on the topic of the value of life on which my essay draws. It provides more in-depth discussion of the issues raised in my essay and also addresses some of those concerns raised in Viscusi’s response. Among other things, I explain why a variant of the cost of death approach commonly used prior to the value of a statistical life’s (VSL) widespread adoption is vastly superior to the VSL along a wide range of ethical, economic, and also practical dimensions.
For the purposes of this response, however, I want to focus on the primary area of disagreement between Viscusi and myself, which I believe relates to the treatment of future generations in analysis. To be perfectly blunt, I think their preferences matter as much as our own; Viscusi does not appear to agree. Viscusi also makes several inaccurate claims in his response to my essay and correcting them should provide further clarity.
First, Viscusi states that the VSL represents “society’s willingness to pay for the mortality risk reduction.” This is objectively not true. In fact, the VSL is an estimate of what some people in the current generation are willing to pay for mortality risk reduction. The present generation and society as a whole (which would presumably include future generations, as well as those who have been born but are not in a position to make economic decisions yet) are two different things.
It may well be true that senior citizens today, or any other group for that matter, are willing to pay something for risk reduction that roughly corresponds with the VSL value. What Viscusi misses is that people in the future would also be willing to pay something for present-day policies that will eventually benefit them, or to avoid present-day policies that will harm them. A more universal measure of willingness to pay should therefore account for their preferences too. But future generations obviously cannot trade in our markets.
When resources are exhausted in wasteful ways, it may well be in line with present preferences. Much like cashing in a well-performing stock to go on a decadent vacation, there is an immediate benefit. But in exchange for this ephemeral gain, there is an unseen future cost to consider as well: how the stock would have performed had it been held on to. And although spending some of society’s present-day wealth to cover the cost of regulation is inevitable to a certain degree, we should strive to understand the opportunity cost for our children.
Viscusi is correct when he states that policy-making “should be restricted to situations where there are demonstrable, consequential negative externalities directly generated by individuals’ own consumption decisions.” The decisions made by present consumers in the marketplace do impose such externalities on their successors. The VSL is estimated from these market conditions, which is precisely why the VSL misrepresents society’s willingness to pay for risk reduction—because it is built upon a foundation of market failure.
For similar reasons, Viscusi is also factually incorrect when he says that using the VSL “enables policies to be guided by the preferences of the citizenry who are affected.” People in the future are very much affected, but their willingness to pay to prevent or encourage the unseen effects described above goes overlooked.
The implicit model that Viscusi is using holds that the preferences of present citizens should be adhered to irrespective of any external costs imposed on future generations. In this sense, the present generation is like a dictator that gets whatever it wants.
If we took that to the extreme, and the present generation wanted to consume 100 percent of society’s wealth and leave behind nothing for the next generation—essentially forcing civilization to start over—should the government work to fulfill those desires through coercive regulation? After all, it would be consistent with the preferences of current citizens.
The fact that cost-benefit analysis could, in theory, produce such a horrifying recommendation—and the fact that it makes less-extreme versions of this recommendation a routine practice—should give economists everywhere pause and lead them to reconsider how analysis is presently conducted.
Viscusi is correct that the problems I have identified have broader implications than just how lives are valued in cost-benefit analysis. But this does not justify ignoring the VSL’s problems. The VSL provides the basis for some of the largest benefit estimates in regulatory impact analysis. If problems with the value of life are addressed, perhaps other troubling aspects of cost-benefit analysis will be addressed too.
As I said in my earlier essay, the VSL directs policymakers to “give people what they want now!” But “right now” is not all that matters. Those who care about fostering a civilization for the long term should think carefully about what it would mean to follow the VSL to its logical conclusion.
This essay is part of a seven-part series, entitled The Value of the Value of a Statistical Life.