When Scientists Admit Mistakes, That Means Science is Working

When the first cases of COVID-19 were discovered in a hospital in Wuhan, I had already studied the U.S. public health response to seven epidemics and pandemics. I am no stranger to the role of politics in pandemic response. Still, the recent Senate hearings grilling Dr. Anthony Fauci alarmed me.

Fauci is a physician and immunologist who served as the director of the National Institute of Allergy and Infectious Diseases and as chief medical advisor to President Trump during the period of Covid-19—who has subsequently become a scapegoat for those who didn’t like the restrictions that came with the pandemic. As Senator Rand Paul put it, based on the evidence-free conspiracy theory that Dr. Fauci funded research that led to the pandemic: “One million Americans died from COVID and not even an apology from Dr. Anthony Fauci.”

While we might recognize such claims as preposterous, it would be a mistake to treat these hearings as mere political theater. Those who are treating errors in the understanding of Covid-19 as criminal threaten one of the most important public goods we have during an infectious disease crisis: the willingness of experts to publicly admit mistakes.

Dr. Fauci has openly acknowledged missteps, communication errors, and evolving judgments in responding to the COVID-19 pandemic—an honesty that is not matched by his critics. Instead, his honesty has been leveraged by them to sow doubt in science itself.

But the reason we turn to experts during a crisis is not because they are omniscient. Of course they aren’t. It is because they learn faster and better than non-experts. Learning, however, is a naturally error-prone process. And learning when a novel infectious disease is fueling a global pandemic is particularly so. One can hope that scientists responding to such a crisis will be able to reduce uncertainties in a linear and predictable fashion. But that is not the norm.

As a social scientist, if I discover something new, when I present my findings to my colleagues, the most likely result is that they will tell me all the ways they think I might be wrong. My job is to prove that my findings hold even after considering their counterarguments. Skepticism is essential in the process of separating fact from fiction. We have more confidence in results that have survived such pressure-testing.

Experts at the Centers for Disease Control and Prevention (CDC) during the pandemic received just this type of skepticism from other experts. And yet much of the political commentary surrounding these normal and essential controversies cited the evidence of controversy as proof the CDC was flailing.

This approach treats science as though the discovery of new facts unfolds like a treasure hunt where the first person to find the treasure opens the lid and everyone can see the treasure with their own eyes. In a treasure hunt, discovery and validation are essentially the same thing.

But science does not work this way. Take the discovery of HIV. Dr. Luc Montaignier identified the virus under an electron microscope in February 1983. Uncertainty surrounding whether the virus was causative or opportunistic led to more than a year of additional research, culminating in four papers published by Dr. Robert Gallo and his colleagues in 1984, that affirmed Montaignier’s finding. According to Montaignier, it was one of these publications, in particular, that “convinced the scientific community that this new retrolentivirus was the cause of AIDS.”

This example shows something that is true across the sciences. New discoveries that eventually become accepted as scientific facts can linger for an extended period as mere “claims” while uncertainties are addressed.

Critics argue that the scientific uncertainty around the discovery of the AIDS virus was unnecessarily prolonged and cost lives. If moving quickly on yet-to-be-validated evidence is the right approach during a pandemic, note that Fauci and his public health colleagues are currently being attacked for doing exactly that. Public health experts believed at the outset that COVID-19 was like SARS, with transmission most likely to occur late in the course of infection. Hence the guidance that directed the public and health care workers to wear masks when caring for patients already showing symptoms. Mounting evidence of asymptomatic infection eventually proved that COVID-19 transmission was very different from SARS, leading to new masking guidance.

When should public health experts act on early, but unvalidated, evidence versus taking the time to iron out uncertainties? While it may look obvious in hindsight, there is no magic formula that can sort out in advance when the better course will have been to slow down and validate versus act on plausible, but still uncertain, claims.

Climate scientists would almost certainly never argue their experience in the public eye has been a picnic. But they had one big advantage over those trying to guide the response to a pandemic driven by a novel infectious disease. Climate scientists had decades—as opposed to only weeks or days—to conduct research before they were asked to explain what they knew to the public. In fact, climate scientists did not always expect warming to result from burning fossil fuels. Up until the 1950s, the prevailing wisdom was that the oceans would absorb excess carbon dioxide so that there would be minimal if any warming. Roger Revelle and Hans Seuss, who together discovered the error, both received awards for their research.

When a novel infectious disease comes to town, the first cases detected are rarely the actual start of the outbreak. There are no tests to make a positive diagnosis. Some of the suspected cases turn out to have other causes. People who are infected are overlooked. Data collected at the outset is patchy and error-prone. As the public health community pivots to emergency response footing, they begin to collect data more systematically. Even so, new data can rapidly overturn early assumptions. The case definition for COVID-19 was updated almost weekly during the first 10 weeks of the pandemic.

Monday morning quarterbacking is easy. Try advising countries on how to manage a growing pandemic while the ground is moving beneath your feet.

As the United States’ performance during the COVID-19 pandemic is reviewed, even those who are not engaging in political grandstanding tend to reduce the effort to a scorecard that tallies the times experts were right versus the times they were wrong. Even if the reversals are few in number, they are barely tolerated. We usually give examinations to people after they have had time to learn and master the material. What is the right metric to use for public health experts who are trying to inform us when the answers are still being discovered?

Speaking of treasure, I think we were lucky during COVID-19 to have public health experts who were willing to admit mistakes as soon as they were discovered. Things will go markedly worse for us if, during the next pandemic, they are less forthcoming.

An earlier version of this article first appeared on MS NOW.