The Pandemic Imperative
Scenes from the 1918 Spanish Flu pandemic, which infected nearly a third of the world's population and killed some 50 million people.
Important Note: I am not an epidemiologist, virologist, or public health expert of any kind. Nothing here should be interpreted as a perspective from any authority on scientific matters.
Moral calculus changes dramatically during a pandemic. Actions that might have little or no ethical consequence during normal times – taking a vacation, holding a large event, or shaking someone's hand – suddenly take on enormous moral weight. Even the most forgettable action can be lethal. When death grows exponentially through a human network and our institutions are corrupted to inaction, our last line of defense is the moral obligation of every individual to the rest of society.
Kant's Categorical Imperative
In describing the criteria by which we should judge moral actions, Kant gave us the following rule:
Act only on that maxim through which you can at the same time will that it should become a universal law.
In other words, would you want to live in a world in which everyone behaves the way you are right now? If so, you can reasonably call it moral behavior. If not, it is immoral behavior.
Moral living, then, requires ongoing thought experiments in the effect of individual choices at scale and over time. You may, for example, run the decision to take an annual vacation to the Caribbean through the thought experiment.
At the scale of the individual and over short-time horizons, there appears to be little about taking a vacation that's morally questionable. But, implementing Kant's thought experiment, we have to ask: Would we want to live in a world where everyone takes an annual vacation to some sunny locale? We may argue that this would be a fine world. There's even a case to be made that there are morally positive outcomes to everyone hitting the beach – it will help stimulate local economies and maybe de-stress everyone enough that we're all a bit kinder to each other upon our returns. In short, it probably passes the test.
What's important to note here is that the conclusion of the thought experiment relies on projecting the nth-order effects of actions and their relationship with systemic factors beyond the individual.
Small Actions & Major Consequences
Enter stage left: COVID-19.
Let's run a very simple (hypothetical) model. One person is infected with coronavirus. While they're asymptomatic, they infect 3 other people, for a total of 4 cases. Each of those newly infected people infects another 3 people, for a total of 10 cases. Go through this process just 10 times and the result is over 65,000 infected people that can be traced back to those first few transmissions. This is the butterfly effect on steroids.
A virus that infects more than one person for every individual infected will grow exponentially through a population. A little more crude math: as of this writing, we're seeing an average of +15% day-over-day growth of new confirmed cases outside China. Given ~35,000 confirmed cases outside China today, holding that growth rate constant would result in 1 million in two weeks, and more than 2 million by the first week of April.
To demonstrate just how much that rate impacts outcomes, tamping it down to 10% day-over-day growth would lead to "only" a half million cases by the beginning of April. Meanwhile, if we start with our current 15% growth rate but reduce it by just 5% each day (so that the growth rate is 15% today, 14.3% tomorrow, 13.5% the next day, and so on) the curve flattens out by the end of the month.
Modeled growth of confirmed cases outside China through April 8.
"But all bringing the rate down does is slow down the virus, if it's still growing then that doesn't mean there's going to be any fewer infections in the long run!"
Well, perhaps. As public health experts have reminded us, the big difference is that by slowing it down, we increase the odds that our healthcare system can handle the surge of demand from new infections. If the number of infections at any given time overwhelms the hospitals, then it's possible that we'll see a rise in the fatality rate as we run out of doctors, beds, and ventilators.
Ethical Actions
The good news is that the rate of growth is not fixed. The quicker we can bend the growth rate downward (i.e. "flatten the curve"), the better our odds of preventing global tragedy on a scale with which few living people are familiar. In China, where drastic measures were taken to reduce person-to-person contact, the rate of growth fell by roughly 15% per day (three times as fast as the variable growth model shown earlier) from roughly 25% in early February. Whereas early February saw new cases rise day-over-day by as much as 30%, the number of new cases in China is now falling.
Few countries can execute the kind of actions China took. Others, like the United States, appear unprepared to do much of anything. This situation is deeply unfortunate, but the reality it generates distributes moral responsibility for keeping transmission rates low to all of us as individuals.
Of course, viruses have evolved specifically to take advantage of our social nature. They spread because we hate being alone.
Let's return to the vacation thought experiment using Kant's categorical imperative in the context of this new reality. Before, we determined that a vacation was morally justified because the world would be quite alright (at least in theory) if everyone took a vacation.
Now, we must take into account the non-zero possibility that you are unwittingly infected with coronavirus, asymptomatic as you may be. By traveling, you then run the risk of introducing the virus to the population of your destination as well as to the home communities of all your fellow vacationers. While the odds that you are infected today may be low, what world do we create when everyone – including those unwittingly infected – takes this action? Indeed, seemingly innocuous actions at scale produce a world in which the coronavirus growth rate stays high, and more people die.
Due to the deterministic nature of network effects and their blossoming causal chains, the downside risk is colossal irrespective of probabilities. Speeding in a car has a relatively high probability of leading to fatalities, but the downside risk is at least limited to a fairly small number of deaths. The odds that you are infected with coronavirus is relatively low, but the number of people you could effectively kill by being infected and careless with your actions is unbounded. My friend and colleague Josh Liebow-Feeser described this well:
At current growth rates (15% day over day), anyone transmitting the disease to a single person is, in expectation, responsible for 100 infections within 33 days. In expectation, that will result in two deaths, and that's only if we consider the first month. So even if the likelihood that you have the virus is very low, the expected value of the effects of your actions is still pretty serious.
With or without Kant's categorical imperative, this kind of risk assessment at the margins should make clear our moral obligation.
We are a social species, which makes these the measures we have to take uniquely difficult. Of course, viruses have evolved specifically to take advantage of our social nature. They spread because we hate being alone.
This is my reading of moral philosophy during a pandemic, but it's also a personal plea. Our situation makes increased isolation our very duty – to our friends and family, to the elderly and immuno-compromised, and to all of society. Every small action you can take now to limit your contact with others is an action that keeps hospital beds open and ventilators available. The sooner you take action, the greater your impact. If you can work from home, you have a moral imperative to do so. If you can cancel your travel, you have a moral imperative to do so. This extends all the way to cancelling small social gatherings and the frequency with which you wash your hands. Even if you feel no symptoms, you can contribute to keeping this crisis manageable by staying home now. Talk to your friends and family about the seriousness of these actions now, before the infections skyrocket further. They will pay dividends in human lives saved over the coming weeks and months. The best time to limit your contact with other people was yesterday, the second best time is now.
Thanks to Josh Liebow-Feeser and Alex Bitterman for contributing perspective and ideas.