Pricing human lives!? Let’s see what good ol’ Adolf thinks about this.
“That is inhuman! That is completely inhuman!”
“Typical Americans! No one else would think of pricing human lives!”
“Mein Führer, I am doubtful if that is factually accurate.”
“The Americans are evil! Capitalist! Scum!”
“Mein Führer, let’s look at some examples.”
The Australian government uses A$4.2m as the Value of a Statistical Life (that’s about $3.1m at current exchange rates). A French government working group proposed using €3m (that’s about $3.2m), and the Swedish Traffic Agency uses 22.3m kronor (that’s about $2.6m).
If you’re weeping in utter despair at what this cruel, capitalist world has come to, then why not try this thought experiment?
Say there’s an invention that will double your country’s GDP. But it will also have the unfortunate little side-effect of killing 1 out of 10,000 people every single year. So in the US in particular, this invention would kill about 32,000 people annually.
Question: What do you think about this invention? Should your country adopt it?
If you truly believed that a human life is above all price, then obviously your country shouldn’t adopt this invention. Unfortunately, it turns out that your country has already adopted this invention. Indeed, this invention has been adopted by every country on earth. It is called … the car. Each year, the US has about 32,000 traffic deaths and the world as a whole has over 1.2m.
If human lives were truly priceless, then countries would simply ban the car. Yet this is never done, for the simple reason that the car’s economic benefits outweigh its cost in human lives.
Instead of banning the car, here’s a less drastic alternative. Why not we lower the speed limit to, say, 16kph? True, this would lower GDP, but it would also save many lives. Plus, we’d still get to keep the car. Surely everyone will support this amazing idea, right?
Uh, actually, wrong.
For example, in 1974, the US government imposed a National Maximum Speed Limit of 55mph. Over the next year, the fatality rate fell by 15% (or nearly 10,000 fatalities). But instead of being celebrated, the law was widely condemned.
“The high cost of the 55 mph speed limit!”
In 1987, the law was partially rolled back. And in 1995, it was completely repealed. One study estimated that the 1987 rollback caused an additional 360 fatalities a year. Another study estimated that the 1995 repeal caused an additional 12,545 fatalities over the next 10 years.
If we truly believed that every human life were priceless, then we wouldn’t be willing to compromise at all on safety. We’d lower speed limits to the point where there were zero fatalities every year. But no society or country in the world does this.
Perhaps you’re now thinking, “Well, maybe other people’s lives are not priceless. But as far as I’m concerned, my children are priceless.” So, let’s try another thought experiment.
Suppose your child needs a bike helmet. The standard one costs $20. You’re about to buy it, when you notice a safer helmet.
This safer helmet can eliminate 1 additional micromort. One micromort is a 1 in a million chance of death. So the safer helmet can lower your child’s risk of death by a further 1 in a million. Of course, this safer helmet is also more expensive. So, which helmet do you buy? The cheaper one? Or the safer one?
If your child were truly priceless, then this would be a no-brainer. Of course you’d get the safer helmet, right?
But, what if it costs $20,000? Or, what about $2m? Would you still get it?
Regardless of how priceless you say your child is or how rich you are, there is always some price at which you would not be willing to pay for that safer bike helmet.
When economists talk about the Value of a Statistical Life (or VSL for short), they’re thinking along these lines.
“May I have your attention, please?” VSL estimates are NOT based on GDP per capita. Nor are they based on personal incomes. “I REPEAT.” They are NOT based on GDP per capita and they are NOT based on personal income.
Instead, they are based on experiments like this bike helmet experiment. The idea is to figure out how much you are willing to pay to reduce by a little your probability of death.
Let’s say you’re willing to pay at most $10 more for the safer bike helmet. From this, it may be interpreted that you are willing to pay at most $10 to reduce by 1 in a million your child’s risk of death. This would in turn imply that the VSL you place on your child is $10m. That’s the $10 multiplied by a million.
The name Value of a Statistical Life is a bit misleading. It does not mean that you would be willing to sell your child for $10m.
Instead, VSL estimates are only ever used in cost-benefit analyses that involve small risks of mortality. This makes complete sense, since VSL estimates are derived from experiments that measure how much people are willing to pay in order to make a small reduction in their risk of mortality.
Here’s an example. Say there are two proposed safety measures. The first is to upgrade all cars to Batmobiles and the second is a seatbelt campaign. Either of these safety measures is expected to marginally reduce the accident fatality rate and so save about a thousand lives.
Let’s say the Batmobile proposal will cost $10t and so the cost per life saved will be $10b. This is far above any estimate of the VSL. The Batmobile proposal is thus extremely cost-ineffective and will certainly be rejected.
In contrast, the seatbelt campaign costs only $1b and so the cost per life saved is only $1m. This is less than $9.4m, which is the VSL used by the US Department of Transportation. Thus, this proposal will be deemed cost-effective and will probably be implemented.
Now, isn’t it really cold-blooded to scrimp on Batmobiles, simply because the cost per life saved is $10b? Well, no less so than when we refuse to pay $2m for our child’s bike helmet.
The only difference is that when we refuse to buy our child a $2m bike helmet, we needn’t say that it’s because our child’s safety isn’t worth it. Instead, we can rationalize our decision in other ways. For example, by saying things like “This helmet’s a rip-off!” Or, “Oh-ho-ho-ho! The cheaper helmet is good enough.” We can avoid the psychological discomfort of having to price the priceless.
In contrast, larger organizations do not have the same luxury, because a single decision they make can affect millions of people. It is therefore much more important to do the right thing and perhaps suffer a little bit of psychological discomfort than to make a disastrous mistake.
An example of a disastrous mistake would be to spend $10t on a safety measure that saves only 1,000 lives. $10t for Batmobiles necessarily means $10t less for other things, such as other more cost-effective safety measures that would have saved many more lives.
At work here are the two fundamental economic principles: Scarcity and tradeoffs. Because resources are finite or scarce, we are forced to make tradeoffs. We sometimes have to give up on a safer bike helmet, so that we have money left for other things, such as our children’s education, shelter, and health. Likewise, we sometimes have to give up on one cost-ineffective safety measure, in order that we have money left for other, more cost-effective safety measures.
People often say that a human life is priceless. But in fact nobody is willing to pay the price to eliminate every single health and safety risk. This is simply because resources are scarce and so we have to figure out where best to allocate them.
So, when economists talk about the Value of a Statistical Life, they are not simply being their usual cold-hearted, calculating selves. Rather, they are simply doing what you and I already do on an everyday basis. Namely, figure out which tradeoffs to make in a world of scarcity.
Starring … Parah Salin. Justin Bieber. Donald Trump without his wig. The Greatest German of All Time. The Greatest Swede of All Time. Funny Frenchy. And the Crocodile Hunter.
 $9.4m is the Department of Transportation’s latest 2015 figure: Memorandum, “Guidance on Treatment of the Economic Value of a Statistical Life (VSL) in US Department of Transportation Analyses – 2015 Adjustment”, June 17, 2015. Link [backup].
 For the Food and Drug Administration, I couldn’t find any figures that looked really official, but this 2013 document mentions using $8.1m as the VSL: “Focused Mitigation Strategies to Protect Food Against Intentional Adulteration”, p. 23. PDF Link [backup].
 Australia A$4.2m (2014): Department of the Prime Minister and Cabinet, Office of Best Practice Regulation, “Best Practice Regulation Guidance: Note Value of statistical life”, December 2014, p. 1. PDF link [backup].
 See National Highway Traffic Safety Administration webpage on Fatality Analysis Reporting System Encyclopedia (retrieved on 2015 12 01, backup spreadsheet). In each of 2009 through 2013, there were 33,883, 32,999, 32,479, 33,782, and 32,719 traffic crash fatalities, corresponding to per 100,000 population fatality rates of 11.05, 10.67, 10.42, 10.76, and 10.35.
 World Health Organization Global Status Report on Road Safety 2015, PDF link [backup]: “Over 1.2 million people die each year on the world’s roads” (p. x); “There were 1.25 million road traffic deaths globally in 2013 – a figure that has plateaued since 2007” (p. 2).