As the VSL Turns...: In Value of a Statistical Life Debate at EPA, Moral Decisions Hide Behind Technical Jargon

A report yesterday from Inside EPA offered a fascinating overview of the agency’s struggle to update the way it assigns dollar values to the suffering and premature death that its regulations prevent. Seriously, as far as economic esoterica goes, this stuff is riveting. What’s more, your life may depend on it.

Currently, EPA values each statistical human life saved by its rules at $7.9 million. This number is derived from so-called “wage-risk premium” studies that examine large data sets on employment and occupational risk. The idea is that, if you control for education, job sector, geographic region, and other relevant factors, then you should be able to come up with a number representing the portion of an employee’s wage that compensates for higher on-the-job health or safety risks. Depending on how a worker values health and safety compared to other goods, he – and he is an important distinction here since the value-of-life studies tend to only look at male-dominated blue collar jobs – might be willing to take a higher wage in exchange for accepting higher levels of occupational risk. In theory, then, the studies can pull out the amount at which workers themselves value risk exposure, which can then be converted into a uniform “value of a statistical life” (VSL) for policy analysis. By using the VSL number to value the health and safety benefits of regulations, EPA can avoid the messy task of government deciding on its own how much protection is worth investing in.

According to the Inside EPA report, staff experts are recommending a new, updated methodology, but the agency’s Environmental Economics Advisory Committee (EEAC) cautioned that the new method might be “too complicated for non-specialists to understand.” This claim is a real howler as it seems to imply that the current methodology is accessible to non-specialists. It is not. Deep and controversial value judgments are embedded within the current methodology, ones that lay persons can scarcely glean. For instance, studies show that union workers receive much higher wage-risk premiums than non-union workers – a finding that suggests bargaining power has a lot to do with the market outcomes that are supposedly capturing individuals’ true “preference” for life preservation. Should EPA use the higher union VSL, rather than the lower non-union VSL that economists tend to favor? This is not a matter of expertise. It is a value judgment that should include a full range of democratic inputs, but its import instead is buried deep within the technicalities of economic regression models.

Apparently the EEAC wants to push EPA even deeper into the weeds by asking the agency to compile a unique VSL figure for each regulatory context that the agency addresses. For instance, if a mercury emissions regulation would disproportionately benefit Native Americans (who eat far more contaminated fish than the general population), then the monetary value of reducing mercury exposure would be calculated using studies that find out how much Native Americans in particular are willing to invest in health and safety. In theory, this would bring the agency closer to the economists’ ideal world in which all values are assessed by the affected individuals themselves, rather than by collective democratic processes. In practice, however, it would involve the government intimately in the perpetuation of discrimination.

The VSL is affected not only by an individual or group’s willingness to invest in health or safety, but also by the ability to do so. This is made clear by the difference between union and non-union VSL data. It is also made clear by studies that show certain minority groups, especially African-Americans, actually receive significantly lower wage-risk premium than should be expected based on their occupational hazard exposure. We might say this represents a weaker “preference” for staying alive among those groups, so that if EPA’s cost-benefit calculations weigh benefits to them at a lower rate than non-minorities, then, well, that’s just giving the people what they “want.” Alternatively, we might say that the picture is messier than this, and that past injustices continue to impact deeply the social and economic opportunities available to individuals and groups today. Treating current market outcomes as somehow neutral and objective does not wash the government’s hands of this history.

The VSL debate is a gripping saga, one with more than a little fiction in it, but with all too real consequences. And it is anything but accessible to non-specialists. For an attempt to break it down in more detail, and for supporting citations, see Chapter 4 of my book, Regulating from Nowhere: Environmental Law and the Search for Objectivity.