Black lives matter. Blue lives matter. All lives matter. When I first started hearing statements like these I was admittedly a bit confused. For one, black lives and blue lives and all lives mattering are not mutually exclusive. So, to my uninformed ears hearing someone respond to “black lives matter” with “blue lives matter” or vice versa made about as much sense as my children arguing “I want to sit on the couch”, “No, I want to sit on the couch!” when there is plenty of room for everyone.
It took a while before it dawned on me what people were really saying. On a surface level, everyone was saying the same thing, that human life mattered. Now you might think that the implication was that certain lives should matter more than others, and indeed some people seem to react negatively to these statements out of that perception, but as I have read about the issue and I have come to the conclusion that very few people actually intend to argue that some lives should matter more.
No, the real disagreement lies under the surface. This is not a disagreement over ideals (at least for most). The undercurrent behind these statements is not about which lives SHOULD matter but about which lives DO matter in practice. When someone says “black lives matter”, they are actually saying that “black lives do not matter” in practice (at least not as much as they should). And, when someone responds with all lives matter, they are often saying “I don’t see the big problem. Things look pretty fair to me.”
As I discussed in previous posts, the assortative structure of our social networks can explain why some feel that racism is a pervasive problem and others can feel like it is not. But, how can one resolve tensions like this? How do you measure which lives matter in practice?
In economics and public policy, an ability to measure the value of a life is vital. If you want to determine whether a potentially life-saving policy is worth the investment, you need to be able to compare the costs (in dollars) to the benefits (in lives saved). The only way to do that is to be able to describe the value of a human life.
Unfortunately, when social scientists first tried to do that it became apparent that just asking people about the value of a human life was not going to work. The answers were all over the map and some people would refuse to answer. And who can blame them? Trying to place a dollar value on someone’s life feels, to use the technical term, “icky”. Others would say that lives were infinitely valuable. However, if life is infinitely valuable, then we should be willing to devote an infinite amount of resources (or at least everything we have) to saving even a single life, but no one lives that way. In other words, there was a disconnect between what people said and how they actually behaved.
In response to this problem, the notion of the value of a statistical life (VSL) was introduced. Instead of asking people about how much they valued life, scientists tried to infer that information based on how people made decisions involving risk. To illustrate this line of thought, let’s thing about an example:
Suppose that you have recently been diagnosed with a rare genetic defect. Most of the time, this defect is benign. However, in 1 out of every 100,000 cases it leads to death. Thankfully, a new 100% effective treatment has been developed that can reduce that risk down to zero. What is the maximum amount you be willing to pay for that treatment?
In this scenario, you must decide between two options:
1. Spend $0. Live with the defect and roll the dice that you are not the 1 out of 100,000
2. Spend x dollars. Treat the defect and live with zero risk (from this particular defect).
In other words, your answer is the amount you would be willing to pay to reduce your risk from 1/100000 to 0/100000.
Notice that if there are 100,000 people with the genetic defect, then without the treatment one of them will die on average, but if they all get the treatment then that person will be saved. So if you take the average amount they were willing to pay and multiply by the number of people, then you arrive at the total amount that these people were willing to spend to save one life. One would expect some variability in the amount people are willing to pay answers to this question, but let’s suppose that the average person is willing to pay $100 (a life is a life, but 99,999 out of 100,000 are pretty good odds). That would mean that the value of a human life is around $100/person x 100,000 people = $10 million.
Now you may notice some problems with this approach. If we change the person who has the defect to a family member or a friend or a stranger, we might get different answers. If we change the risk to 1/1000 would people still be willing to pay $10000 on average (the amount that would lead to the same conclusion)? Would people from different cultures and different belief systems answer the same way? Would our answers change if the person whose life is being saved is 8 years old? What about 80 years old? What if they are in a wheelchair? Or a vegetative state? Should those lives have the same value?
One attempt at addressing these concerns involves assigning value not in terms of lives but in terms of years of life. In that context you can talk about quality adjusted life years (QALY) and assign a value to the number of years saved adjusting for the quality of life. From a moral perspective, that makes me uneasy. Isn’t every life valuable to God? But then again, consider the infamous trolley problem:
A trolley is careening out of control. There is a junction up ahead and on one branch of the track there is an 8-year-old girl blissfully unaware of the danger that approaches. On the other branch is a frail 80-year-old grandmother. Neither has time to get out of the way, but you have a chance to direct the train away from one of them. Who would you save?
I suspect that, discomforting though it may be, most of us would choose to save the young girl. We might even consider that the right choice. This suggests that we place a higher value on the lives of the young and healthy than the old.
These may seem like absurd hypotheticals, but we make analogous decisions all the time. Suppose you want to visit some friends that are 200 miles away (so it takes roughly the same amount of time whether you drive of fly). The risks associated with driving (around 12.5 deaths per billion vehicle miles) are around 10 times higher than they are for flying. Suppose that it costs $100 to fly (including parking). Would you do it? What value of a human life would that correspond to? Well, if you would opt for the cheaper but riskier option (driving), then that means you thought the $100 was not worth it. That means that the value you assigned to your life was less than $33.3 million (the details of the calculation are at the end).
We make similar judgments about risk when deciding whether to get a flu shot, schedule a preventative screening, buy a safer vehicle, take a risky job, etc. By looking at many of these different decisions in aggregate along with hypothetical questions like the one I proposed earlier, one can arrive at an estimate for the value of a statistical life implied by our behavior. These estimates vary somewhat but they tend to be around 5-10 million dollars. The Department of Transportation end Environmental Protection Agency use a value of around $9.6 million when evaluating their own policies (https://www.whitehouse.gov/wp-content/uploads/2017/12/draft_2017_cost_benefit_report.pdf).
In the context of the COVID-19 pandemic, people faced the choice between additional economic hardship caused by shutdowns and additional lives lost due to a virus. The choices that were made reveal something about the value of we assign to human life. Based on a VSL of 10 million, spending $1 trillion dollars to save 100,000 lives or $5 trillion dollars to save 500,000 lives is worth it. But how can we tell how many lives were saved? And how can we determine the economic cost of the shutdowns? These questions will be difficult enough to answer in hindsight, but answering ahead of time when trying to make decisions is incredibly challenging.
Returning to the original topic, the question of which lives matter in practice can be addressed by looking at how people make decisions about risk and amidst uncertainty. In the next couple of posts, I will take a look at the issue of policing to see what clues it might provide about the pervasiveness of discriminatory practices and what that suggests about whether all lives are valued equally.
PS. Your odds of dying on the trip would be around 0.0000025 by car and around 0.00000025 by plane. So, if 10 million people chose to drive then on average 25 would die and if 10 million chose to fly then 2.5 would die for a difference of 22.5 lives. How do the costs compare? It would take around $25 worth of gas to drive. So the total cost would be around $25/person*10 million people=$250 million. For flying the total cost would be $100/person*10 million people=$1 billion. So in that scenario, flying costs $750 million more and saves 22.5 lives which divides out to $33.3 million per life.