On his first day in office, President Joe Biden signed an executive memorandum requiring his administration to modernize the regulatory review process. Among other things, the Biden Administration will be updating regulatory analysis guidelines that have been in place since 2003, guidelines that could shape how all kinds of regulations—including those stemming from Biden’s ambitious environmental agenda—are analyzed for years to come.
One important input in the economic analysis of regulations is the dollar value economists attach to the benefit of reducing risks of death. The issue at stake is figuring out how much society should spend (in the form of compliance or other costs) on regulations targeting health, safety or environmental hazards that cost lives. These are uncomfortable questions, but also unavoidable ones in certain situations.
One common metric economists rely on is the value of a statistical life (VSL), which is a measure of how much money a group of individuals is willing to pay to reduce the probability of its members dying. For example, workers in a relatively dangerous industry like construction receive higher wages than they might otherwise get elsewhere. Given the statistical realities involved, one can add up the excess wages earned by workers in exchange for taking a riskier job to ascertain how much as a group they jointly place on a life lost in their industry.
One claim sometimes made by public health experts, regulators, and even from some economists is that when they use the VSL, they are only putting a dollar value on “risk,” not on anyone’s actual life. Even the Office of Management and Budget, in the regulatory analysis guidelines now being updated, makes a version of this claim. According to those guidelines, using the term “value of life” to describe the VSL can be “misleading because it suggests erroneously that the monetization exercise tries to place a ‘value’ on individual lives.”
Is it true that when economists employ the VSL, they are not placing a dollar value on individual lives?
To start, consider a hypothetical: Let’s say for the sake of argument that air pollution in a city affects a population of 1 million people. Over the course of a lifetime, four people in the population die early from causes related to air pollution, and the deaths are more-or-less random. Thus, each member of the population faces a 1-in-250,000 risk of death.
Now let’s say that government planners decide to implement a public health regulation that they expect will reduce the number of people who die prematurely from air pollution from four to two. Furthermore, let’s say the city’s 1 million people are each willing to pay $20 for this reduction, which lowers any particular individual’s chance of dying from 1-in-250,000 to 1-in-500,000. Collectively, therefore, the city’s population is willing to pay $20 million, or $10 million for each person saved.
The first thing to note about this particular example is that “risk” for any individual changes when there is a corresponding change in the total number of people in the population who die early due to air pollution. As a group, the population is indeed placing a dollar value on specific, individual lives. Whether its “risk” or “lives” that is valued is semantics in this case.
Of course, whose life is on the line is generally not known ex-ante. Even after the fact, the individuals involved are often unidentifiable, but they are no less real just because they can’t be identified.
The number of unidentifiable individuals whose lives are at stake can also vary, even holding a change in survival probability constant. In epidemiological data, analysts can’t always identify the exact number of deaths a hazard like air pollution moves forward in time; they may only be able to identify how the probability of dying changes from year to year at different levels of exposure. Thus, a single VSL applied in regulatory analysis can imply a range of values attached to individual lives.
One advantage of popular alternatives to the VSL—based on people’s productivity or earnings, for example—is they are invariant to the seemingly irrelevant factors that lead to wide swings in the value of life with the VSL. The fact that the same exact life saved can be valued dramatically differently with the VSL—based on the size of the risk involved, whether the individual’s identity is known, and even whether the wellbeing of foreigners is counted in analysis or not—should make people think twice about the method.
Explicitly or implicitly, economists are indeed placing a dollar value on life when employing metrics like the VSL in the context of environmental or other policies. The Biden Administration is now in the process of updating the government’s regulatory analysis guidelines. As it does so, it should be thinking carefully about how to value lives in regulatory analysis. A good first step would be to make clear that valuing human lives is precisely what its economists are doing.