This note is prompted by a recent article on the consequences of raising the speed limits on American highways. The authors estimated that in the decade after 1995, states’ decisions to raise posted speeds increased road fatalities by over 3% — over 12,000 additional deaths. That’s about quadruple the death toll of 9/11. Yet, it is clear that Americans wanted and still want to drive that extra 10 or so miles per hour faster. What does that show about how invaluable each life is?
How much we consider a human life worth has increased substantially over the last century, by one estimate about 30-fold. Yet, we do not spend all we can to save a life. Moreover, how much we do spend to save a life is inconstant and fickle, rather than thought-out and planned; the number typically depends on whose life we are saving. The question is whether we can and will make studied decisions based on how we actually value a life.
We do not, in practice, treat lives as sacredly “invaluable,” no matter our rhetoric. Analysts have sought to figure out how much we, in practice, actually do value a life. It is important to insurance companies and to government agencies. The EPA, for example, needs to justify expensive interventions, say, in cleaning up pollution, by showing that more money – based on the monetary value of deaths avoided – would be saved by the project than would be spent on it. The agency could never defend its policies in Congress by arguing that saving even one life makes environmental action worthwhile; it has to use a cost-benefit analysis that tallies up the dollar worth of saved lives.
So, how much is a life worth? The answer can bounce around quite a bit depending on all sorts of assumptions. In one such estimate, the EPA calculated that an American life is worth, in current dollars, about $8,000,000. A 1990 review of relevant studies (cited here) came up with an average valuation of about half that, $4,000,000 in today’s currency. Blogger Matthew Yglesias reprints a graph by W. Kip Viscusi showing that in the 2000s federal agencies’ valuations of a single American life varied from about $5 million to $9 million.
One way scholars generate a dollar value is by seeing how much extra workers in particularly dangerous jobs make. We assume that the workers who take those jobs are using, perhaps unconsciously, a formula that includes how much they value their lives, something like: Extra hazard pay > (the extra risk) X (dollar value of my life). Playing with the numbers, we can get their valuations – sort of. Another technique for estimating the worth of a life is to survey people how much they might pay (or expect to get paid) to reduce (or to increase) their risks of dying by a certain amount. How much would I have to pay you to get you accept a 1% greater chance of dying in the next year?
However assessed, Americans today value life more than we once did. Economists Dora Costa and Matt Kahn, using the wages technique, estimate that in 1900 an American life was worth about a half-million dollars and in 2000 it was worth over $14 million (in 2010 currency). While the specific numbers are only suggestive, the direction and size of the change are meaningful. Life has become a lot more valuable, at least in monetary terms.
But not all lives are “worth” that 8 or 6 or 14 million dollars; some are – in practice – worth more than others, at least in dollars.
This fact drove one of the most important life-valuing exercises of recent years: the decisions by Kenneth R. Feinberg, “special master” of the Victim Compensation Fund for 9/11. His charge was to base the compensation for families of victims largely on what the deceased’s future earnings would have been. Thus, for example, a 42-year-old woman was judged to be “worth” $976,915 (see here), far below the estimates I mentioned earlier.
Even without the 9/11 claims, we know that not all lives are valued by society equally. The endangered child of a wealthy person gets all the treatment that a top-flight health plan and extra funds can provide; the similarly endangered child of a low-income person depends on emergency-room treatment. More critically, the first will have had a lot of preventative care and the second a lot less. The result is difference in mortality (see, e.g., here).
In her important book, Pricing the Priceless Child: The Changing Social Value of Children, sociologist Viviana Zelizer traces the history of life insurance companies’ and courts’ “pricing” of children. In the 19th century, the courts decided that children who had died in accidents were “worth” little, because they earned little, and so parents received only token compensation, if any. In the 20th century, courts decided that the children who had died were “worth” many dollars because of their emotional “value” to parents.
So, if – despite any revulsion we might have about it – we operate with some implicit economic valuation of life, the question is, should we do so systematically and calculatingly (and justly), or should we leave it as we do today, as a haphazard process?
This, of course, brings us to the debate about health care: Should we have evidence-based evaluation of treatments or try to give “everyone” “anything” he or she might need – knowing that we really won’t do the second? Should we counsel terminal patients about less intensive and less expensive options, or should we go pedal-to-the-metal in the hospital? Should we even talk about the expense of the last months’ of life care? It seems so inhumane; our modern sentiments tell us that costs should not dictate life-and-death decisions. But those modern sentiments do not fit our modern experience.
(Cross-posted on The Berkeley Blog, September 14, 2011.)