Avoiding the trap of sacrificial math

Must we pit the health of the economy against the lives of the many people projected to die if strong measures are not taken against the spread of the coronavirus? Some writers, notably Joy Pullmann in The Federalist, seem to think so. To those who react in horror, alleging that it is indecent to think of economic concerns and money at a time like this, they respond that the economy's health underwrites our own and that we cannot throw the world into a depression to save everyone. At some point we must sacrifice a few in order to save the many. As readers pick up the debate online, it often slides into a competition in ad hoc and speculative utilitarian forecasting that assumes our problem is to determine which group should be sacrificed for the sake of the other.

Such a dialectic threatens to lock our thinking into an inflexible paradigm which constrains our vision to see alternatives and so represents an obstacle to dynamic and proactive problem solving. Both the economy and lives of countless people vulnerable to the virus make demands upon us that we cannot simply ignore. We should not be debating which group to sacrifice. Rather, we should see both the plight of the sick and future jobless as constraints that shape how we tackle the problem.

That the economy is vital to our wellbeing and essential to preserve is a point that has been made persuasively over the past few days. (For its strongest articulation, you can read Esther O'Reilly's "Economic Costs Are Human Costs" at Arc Digital.) I will briefly consider some numbers about the coronavirus as it stands today to emphasize that it is a real threat that must be taken seriously. In the early days of the pandemic, it was popular to compare coronavirus to the flu (another contagious disease that can be deadly; see Agamben here, for instance). But the estimated daily death toll of the flu is 795 to 1,781 per day and the death toll of the flu is not increasing exponentially. The coronavirus daily death toll passed that of the flu on March 23, when 1,873 died of coronavirus. The rate of increase is roughly 13 percent a day and so the number of people who die a day from coronavirus doubles in less than a week.

In recent weeks, many states and municipalities have issued shelter-in-place orders in an effort to "flatten the curve" of new cases and deaths. The exact fatality rate of the coronavirus is not clear and likely varies from location to location based on the quality of care available and the vulnerabilities of the population. On the date of the writing of this essay, March 26, worldometer reports 2,791 deaths in a day. Perhaps the best rationale behind slowing the spread and flattening the curve is that it prevents hospitals from being overwhelmed. A significant percentage of the infected, estimated to be as high as 19 percent of adults and 6 percent of children, have severe or critical conditions. Without intensive care, these cases could easily become fatalities. When the virus overran cities in Italy, doctors had to ration care.

The numbers matter: I am sure we would happily risk a Great Depression with indefinite social distancing if coronavirus's fatality rate were 40 percent, and I'm sure no one would advocate shutting the economy down for two weeks to save four lives. But here we are dealing with forecasts that involve uncertain and even cumulative effects, weeks, months, and years down the road. Both economic and epidemiological forecasts are less precise and less certain the further out into the future we project-all sorts of new developments can change the course. I would be suspicious of anyone claiming to know where the Dow will be one year from now, or prognosticating the exact numbers of coronavirus deaths in the United States. If we think the question is identifying which group of people whose wellbeing we can disregard, then everything hinges on our ability to precisely compare two very uncertain quantities.

Another danger in thinking of the problem as identifying who we can sacrifice is that it encourages us to think that we have solved the problem once we have finished the math. An episode from the popular podcast Radiolab deals with medical triage during times of disaster and includes reports and stories from Pulitzer Prize winner Sheri Fink involving triage during the unrest in the former Yugoslavia, New Orleans during Hurricane Katrina, and Haiti in the aftermath of the earthquake. All the stories are revealing, but the Haitian one provides an powerful example of how a bad ethical theory can limit our ability to evaluate a situation or find a solution.

In the aftermath of the island's earthquake numerous makeshift hospitals and treatment sites were established to provide immediate and emergency care for victims. A problem arose. There was not enough oxygen to go around. The doctors decided to ration the oxygen to those for whom it would do the most good. That meant that patients with preexisting chronic conditions didn't get it. Such decisions are heartbreaking, but that's how triage works.

A woman approaching middle age had a chronic condition, congestive heart failure, and so was not eligible to receive oxygen. Once the doctors made the decision about the rationing of oxygen, it was taken away from her. She was sent to another medical site without it. Gasping for breath, she couldn't understand why she was not being given oxygen. She was slowly suffocating. Tellingly, the reporter interpreted the decision at the first hospital as a decision to let the woman die. As they arrived at the next site, a new doctor came up with a clever solution that involved non-rationed drugs to lower the level of oxygen the woman needed so she could breathe from a discarded oxygen tank. She survived.

Some of the best discussions I have ever presided over in a classroom have been over this case. Students usually argue over which doctor did the right thing and, like the reporter, they usually end up pitting the doctors against one another. They are trying to conform the event into the very limited categories they understand-something like consequentialism versus duty ethics or a sentiment that all life is sacred. They are in effect like Procrustes: They contort the case so that it fits the arguments they have available for it

The reason the students think of this as a conflict between the doctors is that they are thinking of the first doctor's decision not as a one on how to ration oxygen, but as a rationing of life. If you think of the decision by the first group of doctors as the decision that she is to die, then yes, those doctors are indeed in conflict with the doctor who saved the patient. But they need not be so. What is usually missed is that the second doctor respected the distribution of oxygen-he gave the patient a discarded oxygen tank, then treated the patient through other means (diuretics for example) so that she could survive on what was left in the discarded oxygen tank. He was able to both respect the distribution decision and find a way to save her life.

A common effect of corrupting college students (or medical students) by introducing them to the "three ethical theories" (utilitarian, deontological, and something resembling virtue theory) is that they end up locked into a paradigm that distorts reality in order to fit it into the logical space of those three theories. They can't distinguish between rationing a finite resource like oxygen tanks (perfectly reasonable) and rationing life itself (extremely dangerous).

This paradigm limits us by encouraging us to think we have solved the problem once we have finished the math. People who instinctively think in consequentialist terms are prone to this blind spot. Consequentialism does not merely hold that sometimes we make decisions that negatively affect some people for the sake of the greater good, but that the greater good cancels their moral claim on us. If I think of the situation as one in which the sacrifice of eight lives is cancelled out by saving twenty lives, I am unlikely to go looking for alternatives which might save a few of the eight that I have endangered for the sake of the twenty. This could perhaps explain why the doctors who took the oxygen away couldn't see the options available that the second doctor saw.