Covid-19 and the worth of a human life

The scales of justice (courtesy Wikimedia)

Covid-19’s grim toll

The statistics are staggering: As of 1 June 2020, according to the Johns Hopkins University database, the U.S. had logged over 1.811 million confirmed cases of Covid-19 and over 105,000 deaths. The U.K. was next, with over 277,000 confirmed cases and over 38,000 deaths. Worldwide, over 6.3 million cases had been confirmed, with more than 376,000 deaths. If current trends continue, the U.S. death toll alone will soon exceed that of all wars in its history except for the Civil War and World War II.

The economic costs have been similarly astounding. On May 8, the U.S. Department of Labor reported that the U.S. unemployment rate had risen to 14.7% (adjusted by some economists to 19.5%), substantially higher than the peak (10.0%) of the 2008-2009 recession. Trillions of dollars (and pounds, euros, yen, renminbi and other currencies) have already been spent by governments worldwide in an attempt to prevent a major economic collapse, and even more stimulus will likely be required in the coming months. Much of the economic fallout will be long-lasting, as many businesses, large and small, particularly in retail services and travel, may never fully recover.

At the present time, national, state and local governments are grappling with the difficult decision of when to relax stay-at-home orders and reopen their economies. Clearly the tradeoff is difficult, and the risks are great: if a relaxation of restrictions happens too fast, then the city/state/nation risks a “second wave” resurgence, which may require a return to very restrictive measures and result in even more deaths and economic costs. As New York Times columnist Paul Krugman observes, “What good is increasing G.D.P. if it kills you?”

But costs cannot be completely ignored: How much economic devastation is a society willing to accept to save additional lives? Clearly there is no easy answer, particularly given the uncertainties in how the Covid-19 pandemic spreads, and in which measures are most effective in controlling it.

Is a human life literally priceless?

New York Governor Andrew Cuomo recently expressed his view, shared by numerous others, in these terms: “How much is a human life worth? … To me, I say the cost of a human life, a human life is priceless. Period.”

But a quick reflection shows that this view cannot be taken to its logical extreme. Suppose, for point of discussion, that some nation has reduced its number of new Covid-19 cases to very near zero (such as New Zealand and Australia, among others, have reportedly achieved). Suppose also that the cost of shutting down the economy of such a nation for one additional month is USD$1 trillion, and that such a measure is predicted to save roughly an additional ten lives, in conjunction with reasonable testing and contact tracing programs. Is it worth continuing a nearly full-scale economic shutdown to possibly save a handful of additional lives?

Clearly, in almost every major national economy, USD$1 trillion could save far more than a handful of lives if spent in other ways, such as for improved highway and transit infrastructure (which could save hundreds if not thousands of lives each year), or in conversion from coal or oil to cleaner forms of energy (which again could likely save thousands of additional lives each year in reduced air pollution and black lung disease). For that matter, even a modest boost in funding for anti-smoking programs could save many lives.

In fact, one can argue that a governmental body can and should utilize such reckonings in its planning, because to fail to do so, e.g., to make decisions on primarily political grounds, listening to various special interest groups instead of scientists, almost certainly will lead to public funds NOT being allocated in ways that save the most lives.

The value of statistical life (VSL)

All of this raises a basic question: How much is society willing to spend per human life saved?

Such reckonings have a long history, as partially recounted in this article by Adam Rogers. In 1968, Thomas Schelling, an economist who later won the Nobel Memorial Prize in Economics for his work on game theory, wrote a chapter in the book Problems in Public Expenditure Analysis provocatively entitled “The Life You Save May Be Your Own.”

In this article Schelling introduced the concept of the “value of statistical life” (VSL), namely how much money society is willing to pay to reduce by one the expected risk of untimely death.

In 1981, economist Kip Viscusi of Vanderbilt University recommended using VSL to make decisions of how much additional “hazard pay” a worker should receive for doing a job with a known risk of accidental death. He reckoned that if 1 in 10,000 workers died on the job in a given year, and in return each received an extra $300 per year in wages, then VSL for such workers is roughly $3M, which in today’s dollars would be roughly $8.9M.

This is not too far from the current reckoning, used in various calculations, which is a nominal $10M. The U.S. Environmental Protection Agency (EPA), for example, has been using the figure $9.4M in cost-benefit analyses of environmental measures such as reducing auto pollution and water pollution.

There are actually numerous venues where VSL analyses can be (and are being) applied:

  1. Reckonings of hazard pay: As mentioned above, VSL analyses can and should be used to calculate hazard pay premiums for workers in dangerous jobs. The consensus of researchers such as Viscusi is that many workers do NOT receive sufficient additional compensation commensurate with their job risk (see also the table below). Along this line, the U.S. Congress has considered (but not yet passed) measures to provide additional hazard pay for essential workers at risk during the Covid-19 pandemic.
  2. Liability insurance premiums: Insurance company actuaries use VSL analyses to set liability insurance premiums for businesses and individuals as protection from wrongful death lawsuits, say, for instance, if one of the owned vehicles is involved in a fatal accident. Again, the consensus of observers in the field is that many firms and individuals do NOT have sufficient insurance coverage commensurate with their risk.
  3. Life insurance premiums: Individuals should consider the value of their own life, in terms of what financial support would be required for one’s spouse or family in case of death. Again, many financial advisors report that their clients’ life insurance policies are often insufficient.
  4. Governmental environmental and safety standards: Here, again, some nominal VSL figures can and should be considered in setting environmental, health and safety standards. As noted above, the U.S. EPA currently uses the figure $9.4M.

Is “social distancing” for Covid-19 cost-effective?

In April 2020, a group of researchers at the University of Wyoming released a study, to appear in the Journal of Benefit-Cost Analysis, that analyzed the cost-effectiveness of social distancing measures, including shutdowns of large portions of the economy, currently being taken in the United States to combat Covid-19.

Here is a brief summary of their analysis: After reviewing studies in the literature, they assume that U.S. Gross Domestic Product (GDP) would decrease 2% ($6.5T) this year without social distancing and shutdowns, but that with social distancing and shutdowns the GDP will shrink by 6.2% ($13.7T). Thus the cost will be $7.2T. Next (also after reviewing published studies in the literature), they assume that social distancing and shutdown measures will save 1.24 million lives. Using a VSL figure of $10M, they conclude that the benefit will be $12.4T. Thus social distancing and shutdowns save $5.2T. Yes, that is over 5 trillion dollars net savings.

In other words, social distancing and shutdown measures taken so far and projected in the next few months are definitely cost-effective. As Kip Viscusi (the Vanderbilt University economist mentioned above) observes, “Unless you have a really catastrophic outcome, the health benefits of social distancing swamp the costs.”

Are workers in hazardous job categories being sufficiently compensated?

VSL considerations raise the question of whether workers in relatively hazardous job categories are truly being sufficiently compensated for the risk they assume.

To that end, here is a table of some high-risk U.S. occupations (taken from a 2019 USA Today press report), together with median annual wages and yearly fatal injuries per 100,000 (2017 data). Appended to this table, in the column “VSL premium,” is a calculation of what the wage premium should be for this class of worker, based on a reckoning of VSL = $10,000,000.

Occupation Median annual wage Fatalities/100,000 VSL premium
Fishers and related fishing workers $28,310 100.0 $10,000
Logging workers $38,840 87.3 $8,730
Aircraft pilots and flight engineers $111,930 51.3 $5,130
Roofers $38,970 45.2 $4,520
Refuse and recyclable material collectors $36,160 34.9 $3,490
Structural iron and steel workers $52,610 33.3 $3,330
Driver/sales workers and truck drivers $37,610 26.9 $2,690
Farmers, ranchers and other agricultural managers $69,620 24.0 $2,400
First-line supervisors of landscaping, lawn service and groundskeeping workers $47,030 21.0 $2,100
Electrical power-line installers and repairers $69,380 18.6 $1,860

Needless to say, these data indicate that many occupations are not being paid an equitable wage, based on their risk of fatality. Note, for example, that if one were to subtract the appropriate VSL premium ($10,000) from the median annual wages ($28,310) of fishers and related fishing workers, one would conclude that these workers are really only being paid $18,310, or approximately $8.80 per hour, which does not even meet the minimum wage in most U.S. states. Clearly these workers should be paid a significantly higher wage, commensurate with their risk.

Conclusion

In short, analyses based on nominal figures for the Value of Statistical Life (VSL) are useful in several contexts in today’s society, including reckonings of hazard pay and insurance premiums, and by governmental agencies in crafting environmental, health and safety programs. A recently published analysis of the cost-effectiveness of Covid-19 mitigation programs (social distancing and resulting shutdowns of large portions of the economy), summarized above, shows that these measures are indeed cost-effective, even though the costs are staggering.

Many recoil at the notion of assigning a numerical value to human life. The present author himself finds such reckonings rather disconcerting. So perhaps because of their controversial nature, these reckonings will mostly be relegated to academic studies, actuarial analyses, tort cases and governmental agencies, rather than directly in the public eye.

But given that glaring discrepancies in the value of life are implicitly in effect in numerous arenas of society, particularly in wages for workers in dangerous occupations, does it help to turn a blind eye to these injustices? Does it really help to refuse do some hard reckoning in this area? Probably not.

For additional details and discussion, see this Wired article by Adam Rogers, this FiveThirtyEight.com article by Amelia Thomson-DeVeaux, this USA Today press report, this New York Times essay by Paul Krugman and this University of Wyoming study, mentioned above. See also this analysis on risks and benefits for Covid-19 strategies, by Marcos Lopez de Prado (Cornell University) and Alexander Lipton (Hebrew University of Jerusalem and MIT).

[This also appeared at the Math Scholar blog.]

Comments are closed.