"I like it when people are thinking I'm overreacting," the now-ubiquitous Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases, told CNN this week, "because that means we're doing it just right." Fauci's point, as he explained to reporters on Monday, is that "when you're dealing with an emerging infectious disease outbreak, you are always behind where you think you are if you think that today reflects where you really are." That's especially true with a disease like COVID-19, since the vast majority of infections involve mild to nonexistent symptoms. In the absence of widespread testing, the number of known cases therefore creates a misleading picture of the epidemic.
But Fauci's comments raise the obvious question of how we can know when the government is overreacting. Although The New York Times calls that "a taboo question," it is one we need to grapple with, given the potentially enormous economic impact of aggressive COVID-19 control measures.
Fauci was referring to the guidelines that the Trump administration issued on Monday, which recommend that elderly Americans and people with serious pre-existing medical conditions remain at home. They also advise the general public to avoid bars, restaurants, unnecessary travel, and groups of more than 10 people. But local governments have gone much further than that, in some cases ordering businesses to close and confining millions of people to their homes except for "essential" purposes.
Some commentators think such draconian measures should be imposed at the national level. Cornell law professor Michael Dorf argues that the federal government needs to "lock us down" and suspend the writ of habeas corpus to "save the nation."
Dorf glides over the constitutional issues raised by such a policy. "What about civil liberties?" he writes. "In normal times, the government may not confine people for the public safety absent 'clear and convincing evidence' that they pose a danger to themselves or others. One would hope that during a pandemic the courts would construe that standard on a population basis rather than one by one."
Dorf also hopes the courts will accede to a creative interpretation of the Suspension Clause, which says "the privilege of the writ of habeas corpus shall not be suspended, unless when in cases of rebellion or invasion the public safety may require it." Does the international spread of the COVID-19 virus constitute an "invasion"? Maybe! "The short answer is no one knows, because Congress has only ever suspended habeas in wartime," Dorf says. "But there is reason to think that the courts would dismiss a habeas case following nearly any congressional suspension."
Civil liberties aside, the combination of travel restrictions, business closures, and lockdowns will pack an economic wallop. Official numbers from China, the country that has imposed the most sweeping restrictions in response to COVID-19, indicate that industrial production dropped by 13.5 percent over two months, the services index fell by 13 percent, exports dropped by 16.5 percent, retail sales by 20.5 percent, and fixed asset investment by 24.5 percent. By one measure, unemployment rose above 6 percent.
What about the United States? The Associated Press notes that "the vast changes deemed necessary to defeat the virus—people and companies no longer engaging with each other—are bringing everyday business to a halt and likely delivering a death blow to the longest economic expansion on record." A.P. cites an estimate by Gregory Daco, chief U.S. economist at Oxford Economics, who "expects the American economy to shrink at a staggering 12% annual rate in the April-June quarter," which "would be the most dismal quarter on record dating back to 1947." Daco "thinks the economy will post zero growth for 2020 as a whole."
Oxford University economist Ian Goldin is even more pessimistic, suggesting that the response to COVID-19 could lead to an "economic crisis" in the United States that "perhaps even eclips[es] that of 2008," which cost an estimated $22 trillion. McKinsey & Company likewise sketches a scenario in which "the global economic impact is severe, approaching the global financial crisis of 2008–09."
While such projections are highly uncertain, so are the benefits of extreme measures like those recommended by Michael Dorf. But we can roughly estimate the maximum possible value of such interventions, which gives us some idea of how much economic pain can be rationally justified.
In the worst-case scenario imagined by the U.S. Centers for Disease Control and Prevention—a projection that assumes efforts at containment and suppression are largely ineffectual—COVID-19 causes 1.7 million deaths in the United States. Applying the "value of a statistical life" (VSL) used by the Environmental Protection Agency (EPA) to measure the cost-effectiveness of regulations (about $8 million in current dollars), the cost of that outcome would be huge: $13.6 trillion. Using the somewhat higher VSL calculated by Vanderbilt University economist W. Kip Viscusi based on labor market data ($9.6 million), the cost would be $16.3 trillion.
But even assuming that extreme measures are 100 percent effective at preventing that loss (which they certainly will not be), the economic cost, if it is similar to what happened during the Great Recession, would be larger. Those VSL-derived numbers do not include the medical and economic costs associated with nonfatal COVID-19 cases that might be prevented by aggressive intervention. But it is still probably a generous estimate of potential benefits from curtailing the epidemic, because no intervention will be completely effective. Furthermore, the EPA and Viscusi VSLs are arguably excessive in this case, since COVID-19 deaths are heavily concentrated among the elderly, meaning fewer years of life lost on average.
The cost-benefit analysis looks better if you boost the number of deaths expected in the worst-case scenario. Researchers at Imperial College London, for instance, project a maximum of 2.2 million COVID-19 deaths in the United States, which would be a loss similar to the cost of the Great Recession based on Viscusi's VSL. But that projection assumes the "absence of any control measures or spontaneous changes in individual behaviour," which is hardly realistic.
We can scale back both the expected number of deaths without drastic action and the economic impact of trying to prevent them, but we still seem to be left with a big imbalance between costs and benefits. Something like a nationwide lockdown makes sense only if you combine a low estimate of the economic cost with high estimates of the policy's effectiveness and the number of deaths that would otherwise occur.
"You are past the time of monetizing these decisions," New York Gov. Andrew Cuomo told The New York Times today. "You are at a point of deciding: How many people are going to live, how many people are going to die?"
That formulation implies a much higher level of certainty about the efficacy of COVID-19 control measures than anyone can claim, and it completely ignores the tradeoffs they entail. When government agencies consider imposing regulations aimed at protecting health or safety, they routinely take into account not only the number of deaths that might be prevented but also the cost of doing so. That makes sense, because finite resources spent to reduce one kind of risk, depending on the payoff, might better be spent or invested elsewhere, possibly in ways that would save more lives. But in the face of the COVID-19 epidemic, politicians seem to be proceeding on the dangerous assumption that cost-effectiveness does not matter.