Is it economically rational to let people die?

A British friend of mine (and economist) pointed me to this rant from George Monbiot at the Guardian*:

When Sir Nicholas Stern published his study of the economics of climate change, environmentalists - myself included - lined up to applaud him: he had given us the answer we wanted. He showed that stopping runaway climate change would cost less than failing to prevent it. But because his report was so long, few people bothered to find out how he had achieved this result. It took me a while, but by the time I reached the end I was horrified.

...

Stern's methodology has a disastrous consequence, unintended but surely obvious. His report shows that the dollar losses of failing to prevent a high degree of global warming outweigh the dollar savings arising from not taking action. It therefore makes economic sense to try to stop runaway climate change. But what if the result had been different? What if he had discovered that the profits to be made from burning more fossil fuels exceeded the social cost of carbon? We would then find that it makes economic sense to kill people.

At issue is the use of money to value life.  But I would rephrase the last sentence slightly--and I think this rewording makes all the difference: We would then find that it makes economic sense to kill people fail to prevent people from dieing.  I'm OK with that and I'll try to explain why.

Update: I clarify a point I made earlier below (in red).

Allow me to start my defense with a story.  In the movie Soylent Green starring Charlton Heston, Heston's character realizes late in the movie that a state-supplied protein food supplement is actually ground up people.  He has a HOLY CRAP! moment and starts yelling "Soylent Green is people!"

Monbiot has a fairly typical non-economist reaction when he realized that economists are placing monetary values on lives.  I like to think of it as a Soylent Green moment.  "HOLY CRAP! Environmental costs are people!"  Then the arguments devolves into moral indignation.  You can't put dollar values on human life! or as Montbiot puts it:

Human life is not a commodity. It cannot be traded against profits or exchanged for convenience. We have no right to decide that others should die to make us richer.

But there are two problems with this indignant moral stance:

1. Moral indignation fails to recognize that valuation methods are not valuing individuals but are valuing changes in risk. 
2. States and individuals make implicit trade-offs of life versus wealth/consumption/production all the time.  Moral indignation arises when those trade-offs are made explicit. 

On the first point, I give a simplistic introduction to the method for valuing statistical lives here:

To value the reduction in mortality, we will calculate the Value of a Statistical Life (VSL).  VSL's are not an attempt to value a particular person's life, like the courts would do in a wrongful death lawsuit.  VSL's instead give an average value of a 'statistical' anonymous person. 

The method is simple, we observe trade-offs that people make between the risk of death and money.  For example, people will take riskier jobs in return for more pay, or people will accept lower prices for less than perfect meat inspection programs.  Using these risk/money trade-offs that individuals make, we can infer the average value that market participants place on an 'average statistical' life.  Better put, individuals are making the choices, economists just use those choices to infer the implied population value of a life.  So if moral indignation is in order it is to be aimed at the individual for being willing to make that trade-off and not the economist for using it.

To the second point, we can use Montbiot's trick of the contra-example to see that acting to protect human life at all cost involves just as morally reprehensible a judgment as monetizing the risk of dieing.  Suppose we pass a blanket policy stance that human life is sacred and all human life should be protected.  Despite invoking money, such a stance places an infinite monetary value on life and any cost is swamped by the implied benefit of protecting life.

In the early 1970's the U.S. EPA (under the auspices of the Clean Water Act) was charged with enforcing regulations to make U.S. waterways fishable, boatable and swimmable (and maybe drinkable, I forget) within a decade.  The defense for such strict regulation?  Protecting the public health regardless of cost.  In hindsight the laugh-ability of such a strict stance is obvious.  Many U.S. waterways are not boatable, fishable or swimmable.  Why?  It cost too much to even get started.  There was no prioritization and States were charged with simply finding a way.  But states couldn't find a way without breaking the bank.  The implicit infinite value of life led to an inability to act. 

There are less extreme examples of course.  In the U.S. some states do not have motorcycle helmet laws, some do.  Isn't there an implicit judgment there that individual choice is more or less valuable than the risk of death and the associated costs to the state?

And in the case of climate change, acting regardless of the value of lives lost, may (or may not, but the possibility at least needs to be considered) lead to an increased risk of dieing for others.  Changes in the economic infrastructure structure required for reductions of carbon emission may require massive public expenditure--through direct expenditures on infrastructure and indirect expenditures on incentives for research and development of new technologies.  From where will these expenditure come?  Social welfare programs?  Universal health care programs?  Other environmental programs?  Might not reducing such expenditures increase risks to other subpopulations.  How then do we weigh the tradeoffs of different lives? 

Any regulation that involves a change in risk to humans implicitly places some value on life.  Value of Statistical Life methods just make the trade-offs explicit.  And that's a good thing.  Because then we start debating policy solutions and stop debating perceived moral imperatives.

Comments

[Bob]

The problem with cyclists not wearing helmets is not that they die quickly but that they live for an extended period with costly disabilities. The public inevitably pays for it. This cost certainly outweighs the inconvenience of wearing a helmet. Helmet laws are a no brainer because they prevent no brainers.

[robertdfeinman]

In theory one might be able to put a value on human life (health care services in countries which have national plans do this when evaluating treatments), but I think many of the empirical tests are flawed.

Seeing how much extra someone would need to be paid to undertake a risky job is a good example. The flaw that I see is that the workers who are making an allegedly rational decision, aren't. People are very poor at estimating the likelihood of risk. They think unlikely things are more risky than they really are, and vice versa.

Political policies are also designed so that some lives are worth more than others. This was made explicit when the compensation formula for 9/11 victims was created. The families of those who died who had a higher earnings potential were awarded more money.

The problem is not that life is assigned a value, it is that it is not done rationally. As usually money talks.

[dWj]

There's some moral importance -- at least intuitively there is -- to the fact that we're talking about an aggregation of small risks, rather than picking a handful of people to throw in front of the train. Most people can pretty quickly be made aware that the trading of economic resources against small risks is both ubiquitous and necessary, and the abstract minded will typically accept fairly quickly that it's rational to do so at a linear rate for low probabilities; your VSL is not, then, the value of a particular person's life, but the ratio of costs to small risks.

Measuring the appropriate market VSL can be fraught with difficulties, but there are a surprising number of controversial discussions where any number with 7 digits in it -- or often 6 or 8 -- gives you the same result. Arguing over the right number can be contentious, but if we have one policy that allows people to die for $100,000 per life, and another that requires safety precautions that cost $60 million, there's an unambiguous improvement available. In fact, rather than see a cost-benefit analysis worked out in dollars for a proposed policy that has a significant prevention-of-death component to it, I'd like to see what the VSL is that would justify it.

[chaitanya]

Three points:

(1) In case of people making choices with their jobs on risk vs money, they are *personally* making the decision. In case of climate change, the decision is being made for a person who is potentially on the other side of the globe.. say in Bangladesh ..?

(2) What about making decision for 60 million minus one (or whatever the latest number is) species, who as far as i can see, get no benefit of our economic activity, but should bear the consequences. ?

(3) People choosing riskier jobs, maybe doing as the last option, to let their family just survive. But, when we are dealing with economic benefits of climate change, we may be dealing with incremental luxury benefits like owning a plasma TV. Is is justified to equate the two kinds of benefits. One for survival and one for luxury.

could you please address any or all of these three questions. thanks.

[Jack Schieffer]

Guido Calabresi has a great story about "accepting the gift of the evil deity" that he apparently uses with his law students to illustrate the tradeoffs that people make regarding the value of human life. Here's a link to a few pages from his book, Ideals, Beliefs, Attitudes, and the Law. (I think this site is legit.)

People make these tradeoffs implicitly all the time, and it seems to be making them explicit that bothers people most. A good example of "cognitive dissonance," if I remember my undergrad psychology classes accurately, i.e. holding several conflicting beliefs simultaneously. For example:
A) I am a good person.
B) Hurting other people is bad.
C) My lifestyle creates pollution which then kills people.

A typical response is "Hmm...let's just box up belief C) and store it somewhere dark and quiet so we don't have to look at it very often."

[Prasad Rao]

I myself am (actually, was) a practitioner of cost-benefit analysis in ESH regulations. But I have to come out against the gross insult in your blog that depicts EPA's objectives vis-a-vis the CWA as ridiculous. I believe it was the Congress that required the EPA to protect the waterbodies of the US so. If my memory serves me right, there is always an economic caveat to these public charters. And EPA has taken cognizance of economics albeit by the use of arbitrary 'cost-effectiveness' and 'cut-offs' as criteria to determine pollution abatement and control measures. Besides, if the past is any guide, it is prudent on EPA's part (and optimal from the industry's perspective) to set lofty goals well ahead of desired date of compliance, that serve to both warn and inform industries as to the environmental standards expected of them in future (presidencies and) decades. Besides, and contrary to the theme of your blog, the VSL has been increasing over the years (at a rate much faster than the inflation rate). Even excluding other use and non-use values for environment, health and safety, a higher VSL does imply a stricter environmental standard. So what's laughable about EPA's resolve?

VSLs are a number that we crunch out of econometric programs based on data from markets that typically have experienced government intervention several times over, and/or involve various non-competitive practices and operational inefficiencies (Competitive oil prices, yeah right!). The problem with VSLs is that they vary across countries on account of differences in income and price levels, knowledge and societal preferences. While VSL would be one among many criteria that a regulator/policy-maker would consider to judge a proposed regulation, that criterion fails in global regulations, in particular global warming. As much as a statistically reliable range of VSLs is desirable for policy-making, eventually, regulations, especially, international environmental agreements, are a political decision. And one President overturns the decisions of his predecessor (Kevin Ruud, in his ratification of the Kyoto Protocol, reversed years of anti-global warming policies pursued by the Howard government. Did the VSL change overnight?)

I am one who believes that humans will over the decades, find ways to adjust to global warming, and perhaps even benefit from it on a net basis. The question then boils down to non-anthropogenic impacts and the values we humans have for those impacts. A hundred years down the road, there could be two alternative worlds - a world with environmentally sustainable technology in which human, flora and fauna live almost in an utopian world, and, God forbid, a world in which humans live largely cut-off from the environment that wilts under a 'scorched earth' policy.

Polar ice cap? uhh?

[reason]

You seem to be missing his main point that the trade off is being made for other people who have no say in the matter. It is a moral issue not an economic one.

[reason]

And we are not failing to prevent them from dying. We are (potentially) removing the means of their livelihood. Not so subtle difference.

[Tim Haab]

Prasad,

I think you misunderstood my intent and I take that as an indication I didn't write it well. I was not calling out EPA for their resolve in enforcing strict regulation, but rather the piece of legislation put in place by Congress that tied EPA's hands at the time.

When the EPA was formed, and handed Clean Air Act and Clean Water Act of the early 1970's they were not allowed to use costs or benefits to make environmental enforcement decisions. The only criteria was to protect the public health. Because the standards were so strict under the original CWA, it was impossible for states to meet the standards without breaking the bank. So in teh first decade after the CWA little was done. In the early 1980's, Reagan issued an executive order (12291) requiring documentation of costs and benefits for all government outlays over $100m (I think that was the amount). This loosened the Congressionally tied noose on EPA and allowed them to prioritize and target clean-up projects. Since, EPA has been much more successful at getting state buy-in for clean-up projects.

My point was not to criticize EPA, but to point out that Congress' failure to allow the consideration of costs and benefits led to completely unrealistic and ineffective regulation. That's what I meant by laughable.

Tim

[Thomas]

Having laws forcing people riding motorcycles to wear a helmet makes sense as Bob points out, but when it comes to bicycles it's more tricky. Forcing people to wear helmets discourage them from using a bike, and more exercise has a health benefit. Studies from, e.g. Denmark has shown that helmet laws for bicycles are counterproductive, you get a net loss in health even if it prevents some visible injuries. (In addition, encouraging people to use a bike reduces pollution and congestion).

[reason]

I perhaps should have been clearer. Monbiot pointed out that benefits accruing to one group of people where being balanced against cost born by another and matched in dollar terms. He doesn't think this is a valid comparison, and I think he has a point.

[Steve Salmony]

Perhaps someone can assist me and my generation of elders by responding to the following questions related to gaining a more adequate understanding what behavior changes might be required of the family of humanity so that a good enough future can be granted to our children by responding to the following questions.

Please take a moment to explain what you expect will occur that results in the stabilization of population numbers of the human species on Earth in the year 2050, given the fully anticipated young age distribution of a global population of 9+/- billion people at that time? What do you suppose billions of fertile young people, who are expected to be capable of reproducing in mid-century, will be doing with their sexual instincts and drives other than what human beings have been doing during the past several thousand years?

Thanks,

Steve

Steven Earl Salmony
AWAREness Campaign on The Human Population, established 2001
http://sustainabilitysoutheast.org/

[piglet]

You are missing Monbiot's point. The Stern report actually does *not* put a money value on human lives. What it tries to measure is economic loss. Now, how would the death of a poor African or Asian be an economic loss? That person isn't likely to have life insurance. Economists "don't do body counts", you know.

reason: "And we are not failing to prevent them from dying. We are (potentially) removing the means of their livelihood. Not so subtle difference." Exactly What economists are debating is this: will the profits that are to be made from continuing on our current outweigh the environmental damage that it will cause? Some economists asnwer yes to that. They are in effect saying (I will slightly exaggerate to make this clear) that if the US has more to win than Bangladesh has to lose (which is likely), then it is OK for the US to go on and destroy Bangladesh. Of course nobody will put it like that, but this is the true meaning of these calculations.

[piglet]

Sorry: "will the profits that are to be made from continuing on our current *path of economic development* outweigh the environmental damage that it will cause?"

-------------------------------------
Here's Monbiot in detail on Sterns calculus: "On the other side [of the equation] are the costs of climate change. Some of them - such as higher food prices and the expense of building sea walls - are financial, but most take the form of costs that are generally seen as incalculable: the destruction of ecosystems and human communities; the displacement of people from their homes; disease and death. All these costs are thrown together by Stern with a formula he calls "equivalent to a reduction in consumption", to which he then attaches a price.

Stern explains that this "consumption" involves not just the consumption of goods we might buy from the supermarket, but also of "education, health and the environment". He admits this formula "raises profound difficulties", especially the "challenge of expressing health (including mortality) and environmental quality in terms of income". But he uses it anyway, and discovers that the global disaster that would be unleashed by a rise in temperature of between 5° and 6°, and that is likely to involve widespread famine, is "equivalent to a reduction in consumption" of between 5% and 20%."

So the point is that it is not the lives themselves that are valued but only their economic impact. A person who's economic impact is zero (or close) will effectively not count in the equation. This is important. It means that under the right circumstances, even mass killings will appear as economically profitable. The assertions of those economists who disagree with Stern, saying that preventing a climate catastrophe is not a worthwile enterprise, must also be seen in this context. Monbiot is exactly right when he says: "We have no right to decide that others should die to make us richer."