I’m seeing more posts questioning whether the savings of life from social isolation are worth the cost to the economy. Answering such a question is well beyond my pay grade, but I want to highlight some things to keep in mind while thinking about it.
Let me start by saying that it’s not inherently silly or disrespectful to ask whether the cure is worse than the disease, to use Trump’s framing. Life has great value, but anyone who has contemplated a living will has entertained the idea that in some cases it may not be worth saving a particular life. Those who have living wills have spelled out specific conditions in which they deem that the cure is not, for them, preferable to the disease. More broadly, as a society we make no attempt to save every possible life. Traffic fatalities could be reduced by decreasing speed limits, but we shudder at the increased commute times. Flu deaths could be reduced by making yearly vaccination mandatory, but imagine the pushback if that were tried. There are countless other measures we could take that would reduce deaths, but we don’t, and few argue that we should. This isn’t a matter of which side of the political spectrum one is on. Nobody advocates that our entire economic and political resource should be devoted to saving every single life possible. It’s a matter of balance, of determining which “cures” are worth the cost.
In terms of COVID-19, the main thing to remember is how much we don’t know. I’ve written about the uncertainty in the data we have. Pick your favorite tracker and you’ll see a death count. Whatever your tracker says, it’s wrong. Many people have died from the virus and not been counted. Perhaps they died in their homes or on the street and nobody knew why. Maybe they died early on and their death got counted as flu-related. Some jurisdictions only count COVID deaths if there has been a positive test, many people are never tested. There have also been deaths counted as being from the virus when they weren’t. This virus doesn’t have a unique symptom that clearly distinguishes it. In jurisdictions that don’t require a test for attribution, there have likely been some people who died from other causes whose deaths were listed as COVID-related.
But however uncertain the data is, it’s a tiny uncertainty compared to that involved in estimating what might have been. We can count deaths. We can count unemployment claims. We can look at the stock market numbers, the GDP, the number of small business closings, etc. All these numbers, with all of their uncertainty, still give us a pretty good idea of the scale of life lost and economic harm done. But all together they don’t even begin to address the question of whether the cure is worth the disease. To address that question we need to compare the actual costs of social isolation to what would have happened without it. But here we don’t even have bad data, we have only guesses and imaginings.
How many people would have died without social isolation? We don’t know. We don’t even have a very good guess. Right now, world wide deaths stand at about 110,000, US deaths are above 20,000. There would have been more, a lot more, but we don’t know how much more with any degree of certainty. If the US had not implemented social isolation policies would we have 200,000? 2,000,000? 10,000,000? The last number sounds exorbitant, but some estimates had 70% of the population becoming infected and a fatality rate of over 3% assuming proper medical care. That means 2.1% of the population dying. With a US population of about 330 million that’s already 7 million deaths without factoring additional deaths from an overwhelmed medical system. I’m not, of course, saying that particular estimate was correct. Though it was the considered opinion of at least one reputable medical organization others disagreed. But we didn’t know, and we still don’t know.
This is a new virus and we don’t have good data about its actual spread from which to build solid models of what might have been. Fortunately, nobody has tried the experiment of simply letting the virus run its course without intervention. But there is widespread agreement that letting the virus run its course would have been devastating both in terms of life lost and in terms of economic harm.
How much economic harm would have resulted from lesser responses? What would the effect have been if our current death were, say, 10 times higher than they currently are? In the abstract, we can look at the economic effect of losing 200,000 lives. We can calculate, for example, how much lost income they would represent, how much would be spent on their care funerals, and how much hardship their families may have faced as a result of these factors. Based on the assumption of 200,000 deaths, we can also estimate how many people might have needed hospitalization and what that might have cost. We could do the same for an assumption of 2,000,000 deaths, or any other number of deaths. But these calculations all start with an assumption for which we don’t have much basis as I’ve noted above.
Beyond that, the economic harm isn’t at all limited to just the direct effects of sickness and death. One very important, but difficult to predict, economic variable is consumer confidence. What happens to consumer confidence in a scenario in which 200,000 people die from a pandemic to which the government doesn’t respond? What we’ve seen from the stock market this year does not incline me toward optimism on this point. How many small businesses would have collapsed under such a scenario? How many workers would have lost their jobs with no prospect of rehiring once the peak was past? What kinds of actions might the Fed have taken to minimize damage? Would Congress have passed supporting legislation in the absence of containment and mitigation efforts? The answers to all of these questions are important determiners of what the economic harm might have been, but we can’t do more than make largely unsupported assumptions about them.
Still, we had to make a choice about what to do. Circumstances aren’t always polite about waiting for the evidence to marshall itself before decisions must be made. The worst kinds of decisions are those that have high stakes, low and uncertain information, and yet still must be made immediately. These are the sorts of decisions involved in how to weigh the economic and human effects of responses to a pandemic based on a new virus. If you think you have the answer well in hand, you are at best overconfident. We can’t help having opinions, but we should be extremely humble about them. I am glad that Trump has said that this will be the hardest decision of his life. Whether you take him at his word on that or not, it at least gives lip service to the idea that this is fraught with difficulty.