Shutterstock / NIH / The Atlantic

Reopening is a mess. Photographs of crowds jostling outside bars, patrons returning to casinos, and a tightly packed, largely maskless audience listening to President Donald Trump’s speech at Mount Rushmore all show the U.S. careening back to pre-coronavirus norms. Meanwhile, those of us watching at home are like the audience of a horror movie, yelling “Get out of there!” at our screens. As despair rises, the temptation to shame people who fail at social distancing becomes difficult to resist.

But Americans’ disgust should be aimed at governments and institutions, not at one another. Individuals are being asked to decide for themselves what chances they should take, but a century of research on human cognition shows that people are bad at assessing risk in complex situations. During a disease outbreak, vague guidance and ambivalent behavioral norms will lead to thoroughly flawed thinking. If a business is open but you would be foolish to visit it, that is a failure of leadership.

Since March, Americans have lived under a simple instruction: Stay home. Now, even as case counts spike in states such as Arizona, Florida, and Texas, many other states continue to ease restrictions on businesses, and suddenly the burden is on individuals to engage in some of the most frustrating and confounding cost-benefit analyses of their life. Pandemic decision making implicates at least two complex cognitive tasks: moral reasoning and risk evaluation.

My academic subspecialty is the psychology of judgment and decision making. The foundational experiment in this discipline began with the prompt: “Imagine that the United States is preparing for an outbreak of an unusual Asian disease.” (The glibly xenophobic use of “Asian” as a shortcut to inducing fear and confusion is a subject for another article.) The experiment asked participants to choose between two public-health policies: In option A, one-third of the population survives for sure, but no one else makes it; in option B, there is a one-third chance that all survive, but a two-thirds chance that none do. For some participants, these options were described in terms of how many lives would be saved; for others, how many would die. Participants consistently chose option A, which offered certainty, if they were thinking in terms of potential gains (saving lives) but option B, which involved more risk, if they were thinking about potential losses (dying). A weighty decision was swayed dramatically by the semantic framing. (This observation earned one of the experimenters the Nobel Prize for economics.)

The cognitive-science canon is replete with uncanny predictions relevant to the coronavirus era. Researchers have studied the human tendency to discount preventable harms that arise from nature and to overreact to harms that arise from human action. The literature predicts that people will take comfort when a coronavirus fatality is attributed to “underlying conditions”—for instance, a patient’s age or chronic maladies—that they do not share, and they will be tempted by the quick dopamine hit associated with shaming those who fail at social distancing. Cognitive scientists even have experiments to explain the “declining marginal disutility” that people associate with others’ deaths—the feeling that the difference between no deaths and one death is really bad, but the difference between 110,000 and 111,000 deaths is negligible. Evocatively termed “psychophysical numbing,” this confounding juxtaposition of the mathematical and the existential is where Americans live now.

As states gradually reopen, seemingly simple judgments are likely to grow more fraught. What does six feet between people look like? The literature suggests that I am more confident I’m six feet away from a friend than from a stranger, that I’m more likely to blame people not of my race for standing too close, that I overestimate my compliance with public-health guidance but underestimate yours. Humans have difficulty calculating exponents, which is particularly crucial to understanding the speed of disease spread. They struggle to estimate the correct answer to a problem without drifting toward the answer that best serves their own interest. With more freedom of movement, Americans also have more opportunities to make judgments of others—who always seem to be doing it wrong. How can people be sitting in groups, chatting, at an outdoor bar? Who would take their kid to swim in a public pool? Are you inviting those people inside your house?

Even when shamers have the risk calculus right, social-distancing shaming is still useless or even harmful to society. Each judgment is a chance not just to get the math wrong, but to let indignation outstrip empathy. Living in a dense, diverse city, I know that I place moral and practical value on playgrounds, parks, and, indeed, protest marches that I might have viewed as indulgences were I still living in my hometown in rural Maine. Individual citizens—citizens facing a range of permissible options, receiving confusing public-health messaging, triaging competing ethical commitments—are not the best targets of our practical and moral concern. Even within academic psychology, scholars are prone to focusing on individuals who make suboptimal choices—workers who do not save, or employees who choose bad retirement investments. In the pandemic, this urge is a red herring; it is too easy to focus on people making bad choices rather than on people having bad choices. People should practice humility regarding the former and voice outrage about the latter.

At the least, government agencies must promulgate clear, explicit norms and rules to facilitate cooperative choices. Most people congregating in tight spaces are telling themselves a story about why what they are doing is okay. Such stories flourish under confusing or ambivalent norms. People are not irrevocably chaotic decision makers; the level of clarity in human thinking depends on how hard a problem is. I know with certainty whether I’m staying home, but the confidence interval around “I am being careful” is really wide. Concrete guidance makes challenges easier to resolve. If masks work, states and communities should require them unequivocally. Cognitive biases are the reason to mark off six-foot spaces on the supermarket floor or circles in the grass at a park.

For social-distancing shaming to be a valuable public-health tool, average citizens should reserve it for overt defiance of clear official directives—failure to wear a mask when one is required—rather than mere cases of flawed judgment. In the meantime, money and power are located in public and private institutions that have access to public-health experts and the ability to propose specific behavioral norms. The bad judgments that really deserve shaming include the failure to facilitate testing, failure to protect essential workers, failure to release larger numbers of prisoners from facilities that have become COVID-19 hot spots, and failure to create the material conditions that permit strict isolation. America’s half-hearted reopening is a psychological morass, a setup for defeat that will be easy to blame on irresponsible individuals while culpable institutions evade scrutiny.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.