I’ve seen a lot of calls for cost-benefit analysis regarding the COVID-19 pandemic and the economic damage we’re inflicting as we try to control it. I’m not sure it’s possible to do that with the information we have right now — we don’t really know how many people are asymptomatic, for example, and we’ll probably never know what would happen if we did absolutely nothing or took only minimal steps, since no country afflicted by this plague is trying this, uh, creative suggestion.
On top of that, there are countless things we might or might not try to measure and include. When it comes to the benefits of controlling COVID, you don’t just have the lives saved. You also have health-care costs, damage to the economy from people being hospitalized when they could be working (even if from home), and the misery of being intubated. Or not being intubated when you need it because the hospital ran out of ventilators.
And when it comes to the costs, it’s not just the short-term hit to GDP and huge new government spending. There’s the frustration and mental-health problems that might come from being cooped up and/or unemployed, for instance, as well as the future economic damage that comes from letting workers’ and students’ human capital atrophy.
Further, what’s the counterfactual here? Even if the government does nothing, individuals will still react, so we’ll get some of the health benefits and some of the economic damage of social distancing either way.
But hey. We do have a few estimates we can stitch together, and we’re all inside and bored, so why not? NRPLUS members, leave your alternative estimates in the comments!
Deaths. This is probably the biggest uncertainty, because there’s a lot we don’t know about how the virus spreads and kills. An Imperial College study said that COVID could kill 2 million Americans in a do-nothing scenario, and even more if we couldn’t come up with enough ventilators for everyone who needed them, which we almost certainly couldn’t. We can reduce that number by half with mitigation efforts, and more if we suppress the virus more thoroughly. These estimates rely on strong assumptions that will prove wrong in some direction and to some extent. But I think it makes sense to think of this trade-off in terms of a million lives — a nice round number that you can increase or decrease mentally if you prefer a different assumption.
Short-term damage to the economy. This is also not known with precision, but Goldman Sachs is guessing this year’s GDP growth will be about negative 4 percent between a severe contraction now and a rebound later. (When you see claims that the economy will contract 24 percent or more in a single quarter, bear in mind those are “annualized” rates and don’t include the bounce-back.) Without COVID the economy would presumably grow a couple percent, and GDP is around $21 trillion, so call it $1.5 trillion, plus however many trillions the government spends propping everything up by borrowing from future generations. $5 trillion would be the cost of simply losing almost all economic activity for a quarter — no one working at all, no bounce-back later — so it’s probably a reasonable ceiling unless we really do try to lock everything down indefinitely, rather than buying time for a bit while we ramp up production of tests, masks, and ventilators.
Value of a statistical life. This is a number estimated by looking at how much people need to be paid to put their lives at risk. It’s a controversial and unsettled concept, both mathematically and ethically — Regulatory Review hosted a debate on it earlier this year — but the center of gravity of various U.S.-government estimates is around $9 million or so.
Quality-adjusted life years. The above number assumes that all lives are equally valuable, a thought that some might intuitively endorse but that also doesn’t seem quite right. Obviously it’s a lot sadder when a toddler dies in a car accident than when an elderly person with terminal cancer does. So we might want to measure the benefits of stopping COVID in terms of “quality-adjusted life years,” especially because the disease seems to fall heavily on the elderly and hardly at all on children. A decent estimate here is probably $125,000-ish. (You’ll sometimes see it claimed that a year of life could be worth as little as $50,000, but that’s just wrong. The number began being widely used a quarter-century ago, was developed even before that, and has never even been updated for inflation. It’s also ridiculous on its face; it’s less than the median household income, for God’s sake.)
Clearly you can put all this together any way you want. But I find two calculations particularly helpful — both using very round, simplistic numbers so you can scale them upward or downward in your head if you prefer different values.
One is to assume we save a million lives and value everyone’s life equally. In this case we’ve preserved $9 trillion in value, more than 40 percent of a year’s GDP — before we try tallying all the health-care costs of an uncontrolled pandemic and the suffering we avert among nonfatal cases. Starting with these numbers I suspect it would be very, very difficult to make the costs add up to more than the benefits.
The other is to assume we save a million lives, but on average each person only had, say, a decade to live — which is roughly the additional life expectancy of an American male who makes it to 78. The average COVID fatality is significantly younger than that, so we’re probably lowballing here, though many also have other health issues. In this case we’re preserving only $1.25 trillion. I still think the benefits will easily prevail once you account for the other costs of letting this get out of control, which could entail millions more hospitalizations, but obviously it will be a closer call.
Well, if a million fatalities is the right number to start with, which we don’t really know.